#help-with-audio

1 messages · Page 2 of 1

glacial spruce
#

I like to think of it as "wire it up and it runs itself based on simple inputs" vs "controlled by something else".

unborn drum
#

I asked in the main cp help channel, but here might be better.

When I attempt to download a file while playing audio, the audio glitches.

I got some advice last night from todbot here: #help-with-circuitpython message

But I’m having trouble implementing it.

#

This is using a pi pico w, cp8 and PWMAudioOut, btw.

I’m generating a dial tone, which sounds great until the request starts downloading. Then it glitches out, and once the downloading is over the dial tone is normal again.

unborn drum
#

relevant code:

audio = AudioOut(board.GP0)
mixer = audiomixer.Mixer(buffer_size=256, voice_count=2, sample_rate=SAMPLERATE, channel_count=1, bits_per_sample=16, samples_signed=True)
audio.play(mixer)            
mixer.voice[0].level = .25    
mixer.voice[0].play(sine440, loop=True)
while mixer.playing:
  time.sleep(5) #the tone sounds fine here
  response = requests.request("POST", url, json=data, headers=headers) #glitches during this
  print('done') #and sounds fine again here```
weary gyro
#

oh good so that works! now try setting buffer_size=2048 and reload

unborn drum
#

Same glitch!

weary gyro
#

what is your SAMPLERATE?

#

You could try increasing buffer_size more and more and see if the glitching gets less. I'm no too experienced with how the WiFi in Pico W works. It may just be turning off all interrupts entirely when doing a network request, so no amount of buffer can save us. I have a Pico W around here somewhere, I'll see if I can recreate what you're experiencing

queen cradle
#

[not useful, but it's a cool sounding glitch]

unborn drum
unborn drum
weary gyro
#

ahah this is the kind of stuff I like. And I've been wanting to actually use my Pico W for networking stuff (guiltily I have not even installed CircuitPython on it yet)

queen cradle
weary gyro
#

CircuitPython audio is definitely a good source for glitch music fans

unborn drum
#

To answer an earlier Q: sample rate is 22050 right now, but I’ve tried down to 8khz

tall mango
#

I'm doing some research to try and design an analog audio mixing device, does anyone know if there's a device or circuit that can add reverb or echo to an audio signal without the use of digital tools?

#

Also without needing to record it in a large room with lots of reverb

weary gyro
tall mango
#

thanks! I was also starting to look into reverb tanks but they're a little expensive

weary gyro
unborn drum
# weary gyro Okay so I was able to replicate what you're seeing. Here's a gist of the comple...

Amazing. It works!

I’m curious how you discovered that a second request didn’t glitch.

Now, for a curious plot twist: it seems like the url has to be the same for the dummy POST and the real POST. If I hit httpbin first as the dummy request, then play audio, then do a request to the real URL, it glitches again.

Maybe that points to the glitch originating from a dns lookup or something.

#

Thanks again for all your help!

I should probably collect this info and post it to the circuitpython GitHub

weary gyro
glacial spruce
weary gyro
#

oh yeah the PT2399 echo chips. not a good reverb but if you're looking for janky karaoke "reverb", it's the thing

glacial spruce
#

I was thinking the Reticon SAD1024

weary gyro
#

oh! are those available again?

glacial spruce
#

I have some NOS ones, but I've heard they may have failed in the box.

glacial spruce
boreal pilot
# 
    from audioio import AudioOut
except ImportError:
    try:
        from audiopwmio import PWMAudioOut as AudioOut
    except ImportError:
        pass  # not always ssupported by every board!

##Creating the MP3 Decoder
mp3 = open("Redboy_ready.mp3", "rb")
decoder = MP3Decoder(mp3)
audio = audiobusio.I2SOut(board.A1, board.A2, board.A0,left_justified=False)


def IntroSound():

    if intro_voice == True:
        decoder.file = open("Redboy_Dialogue_Y_5.mp3", "rb")
        audio.play(decoder)
        while audio.playing:
            pass

I have this audio succesfully working, but it's not super loud. Is there a code-side thing I can do to louden it?

weary gyro
boreal pilot
#

It needs to be like, toy volume. So I’m not sure if I’m using the correct hardware or not.

weary gyro
# boreal pilot Yeah I’m powering it with 5v

What kind of speaker are you using then? A 4 ohm speaker will generally be louder than a 8 ohm speaker. If you're using one of those really high-impedance 500 speakers then it will be very quiet

boreal pilot
#

But I popped one out of the case and hooked it up to the board.

weary gyro
#

I don't have experience with those speakers, but I believe the case is tuned to make the speaker louder. (acoustics is a dark art)

boreal pilot
#

Oh, interesting okay.

weary gyro
#

The other problem could be your 5V power supply not being able to put out 3W

boreal pilot
#

Oh super!

lapis acorn
#

I'd like to make the audio output from a ScoutMakes FM Radio board available on my Network. I have it working with a USB capture dongle, but I'd like to eliminate USB to reduce the size of my project. The ADS 1115 doesn't look like it has an appropriate sample rate for audio... Any suggestions for an appropriate audio encoder?

glacial spruce
#

Many small microcontrollers include analog inputs with sufficient bandwidth, and of course there are handy outboard chips with a variety of interfaces (the MAX1242 fits in an 8-pin package and provides and SPI interface, for example)

ashen sinew
#

when it comes to the audio soundboard, when i am using it through the uart will the trigger pins still trigger it as well or does it have to be one or the other?

glacial spruce
ashen sinew
#

ok thank you.

stable iris
#

can I ask an headset repair question in here?

odd oasis
stable iris
#

So I have an headset since 2020 from cyber acoustics a company Ive come to trust over the years. But this one has recently started to have this problem where the left side (the one without a wire) randomly goes off/mute and I cant hear anything from that side for a few minutes/seconds while listening to music or playing a game. I tried it on several different computer including my work PC, and the balance in the config panel but I believe it's an hardware / software-on-chip problem

#

the first issue I have is being able to open it at all

#

it seem sealed and the pads seems sewed in, I don't see any screws either, it seem I would have to rip the textile to open it which kinda kill the point? I was hoping to take a look and see if something simple needs to be replaced maybe a bad solder like madbodger said or a cap

glacial spruce
#

Look around it to see if the seam looks different somewhere. In many cases, they sew most of the seam with the cover inside-out, then invert the cover around the unit, and sew the last bit. In cases like that, you can pick open the final seam with a seam ripper, then peel the cover off to get access, then reverse the process to put it back together. However, in other cases, glue is used, and you may need to do some more investigation to figure out the best way to get access without harming something.

stable iris
#

Let's see how fast my patience runs out 🤣

#

I probed it with the jimmy tool but around the blue ring slit it seems sealed with plastic except I can slid a knife inside around 1/10 of the circle

#

but I guess there is value in seeing the insides as Ill probably buy the same one again

#

so it's the cost of learning how to open the next one to repair it if all else fails...

odd oasis
stable iris
#

the plug has a fastener so it cant really be pulled it pushed in

odd oasis
stable iris
#

and yeah I tried different usb plugs and pushing the plug as far as possible doesnt seem to help

odd oasis
#

Is it 3.5mm or usb?

stable iris
#

usb

odd oasis
#

Ohhhhh

stable iris
#

it's a cyber acoustics ac-5008 if that helps but I already tried to find schematics or a tutorial to open it

#

I tried a jimmy / butter knife / steak knife to open it (gently) by putting it in the gap between the blue ring and the pads but it seem Id have to break something inside to pry it

#

It's a 25$ headset so not a big loss but it could also just be a 1$ cap to change or a 20 seconds resoldering job...

#

Last time I did this it was for a pc speaker and a wire inside had slightly come off the board so all I do was solder it back and put electrical tape on it to fix it to the side inside so it would stop jaangling around

valid aurora
#

on the ItsyBitsy M0, the A0 pin is a DAC right? anything special to do to take advantage of that, or is just using analogWrite() enough?

#

tried to output a sine wave and got this weird, clipped, ring mod looking thing

image.png
#

here's the code fwiw. it's meant to be a basic audio processor framework w/ buffered audio, and basic wavetable oscs to test it with. i definitely cut corners in a couple places, but nothing i did should be causing that ring mod-like behavior

📎 DuinoOsc.ino
glacial spruce
valid aurora
# glacial spruce Yes, analogWrite() is sufficient. You've got a few things going on. I think wh...

when you set _amp, it's setting a weird phase jump instead of amplitude.
do you mean in the AudioProcessor constructor? The 1.f / (i + 1) just calculates the coefficients for a sawtooth's harmonic series, so that when you have multiple oscillators, they build a saw

I think your wavetable is scaled wrong
so it should be 0-1023 instead of 0-255?

rewriting your phase code to just count by 1s
so have the phase counter be from 0 to wavetable size - 1? that would still require using floats for the phase step if i want to tune things properly

glacial spruce
#

It is apparent to me that I don't really understand what you're trying to do.

valid aurora
#

it’s mostly just a test. the oscillator is only there to have something to output. my main goal was to have a fixed-rate output, swapping between two buffers so i can write to one while reading from the other

glacial spruce
valid aurora
#

browsing through, this looks interrupt-driven? if so it could definitely be a good reference

#

i wanted to do interrupts at first but favored quick implementation over being the most correct

glacial spruce
#

You don't need to use interrupts, but perhaps the rest of the code might be useful.

valid aurora
#

i think interrupts are better in this case for pseudo-multitasking

#

right now my processBlock() function is polling the timer to keep from blocking the audio output, with interrupts i can have it do only processing

inner lava
#

Methinks, is there any project out there for making either a spectrograph or a bare bones XY scope?

#

I saw tubetime made some driver for CRT tubes, but I don't know how to modify it to make it suitable

glacial spruce
# inner lava Methinks, is there any project out there for making either a spectrograph or a b...

That approach is a linear regulator variant of Jim William's switching regulator design, which is in turn based on a CCFL power supply. I ended going back and reading Williams' document and deciding I liked the switching supply approach better, but the special transformers are kind of hard to get. Since it's basically a modified CCFL supply, and it's possible to get CCFL supplies cheaply as surplus, I split the circuit into two parts and just put the CCFL supply between them.

modular-crt-power-supply.jpg
#

It does actually work

vector-crt-workbench.jpg
inner lava
lapis adder
#

Ik there’s Df player mini for arduino to play music and sound fx

#

But is there any Bluetooth options similar to that

#

LIEK I can play music from my Phone and it’ll connect to the arduino?

glacial spruce
inner lava
glacial spruce
#

That looks like a flat CRT design, which is somewhat more difficult to convert (those normally use magnetic deflection for one axis, and electrostatic deflection for the other).

inner lava
#

they say to keep one hand in your back pocket while working on those... I keep both of them!

glacial spruce
#

We all start as beginners, and those flat screen CRTs are a creative design that not a lot of people know about. I'm a CRT hobbyist, so I've accumulated a bunch of knowledge on CRTs, but these days, not a lot of people know about them.

karmic needle
#

can anybody recommend a good small speaker/amp (~ 3inch/8cm) that accepts 3.5mm and sounds fairly decent?

hasty crater
#

@icy karma @fickle leaf Here's what the circuit looks like. Now you guys are making me feel like I'm going to start a fire. 😅

fickle leaf
#

Just disconnect the wire from the 3w amp’s negative output pad leaving the capacitor positive wire connected to the positive out. The capacitor negative should remain connected to ground.

#

In effect, you were just shorting the amp’s neg output to ground.

hasty crater
#

If I disconnect the negative of the 3w amp there probably won't be any audio.

#

It's all soldered into the protoboard, kind of a pain to experiment with now. I should probably move it to a breadboard.

fickle leaf
#

Both the pos and neg outputs of the amp are actively jumping between the power voltage and ground, out of phase with each other.

hasty crater
#

ah i see

fickle leaf
#

Your solution really just operates from the positive output since the negative signal is shorted to ground.

icy karma
#

I would just add 100k series resistor and get rid of the capacitor.

#

What is the benefit of shorting the negative output to ground?

fickle leaf
#

Basically, the outputs are PWM signals that are out of phase with each other and that require a filter to eliminate the PWM noise.

#

Just like with a DC motor controller, the motor is the PWM filter. For a class D audio amp, the speaker itself is the PWM filter.

hasty crater
#

Something like that with the blue negative wire disconnected?

fickle leaf
#

Yes, and with the blue wire connected to gnd

hasty crater
#

GND of what? The 3W module gnd?

fickle leaf
#

(As it likely is since the cap neg is connected to ground)

#

Yes, the common signal/power ground between the two amps.

#

Both amps share a ground, correct?

hasty crater
#

There is no common ground between the two amps. the 3w is digital, 20w analog.

fickle leaf
#

The power supplies don’t share a ground?

hasty crater
#

i tried a common ground, sounded horrible because it's a dgnd and agnd fighting each other.

#

the i2s is using the feathers power and gnd

#

there is no agnd on the feather

fickle leaf
#

Ah. My assumption was that the power sources shared a ground.

terse maple
#

wait, are you using an I2S class-D amp as a preamp for a class-D analog amp?

hasty crater
fickle leaf
#

If the power supplies are isolated, then that explains why your circuit worked. Egads.

#

Are you using wall warts or batteries?

icy karma
#

Do you need the 3 W I2S amp? Why not use pwm audio on your microcontroller and connect that directly to the 20 W amp?

hasty crater
#

i would have used fritzing to lay it out but last i checked fritzing is no longer free

#

because i prefer i2s?

#

i don't think pwm sounds as good

#

at least with the stuff i've done with it in the past

#

started on an esp32-s3 but esp32's don't have mp3 support

#

so had to move to the rp2040 feather

fickle leaf
#

Please take all my previous observations with a grain of salt. Sorry about the distraction.

hasty crater
#

any effort to help improve the audio quality i will gladly accept

fickle leaf
#

The only issue you may encounter will be if the final power supply situation creates a common ground.

hasty crater
#

it all actually works pretty well and even at about 80% amplification sounds good

#

but i do not want to start a fire

#

🙂 will be keeping a closer eye on it now, you got me paranoid.

fickle leaf
#

Ideally, the polarized cap should be replaced with non-polarized. Placing two caps in series in an anti-polar configuration would work.

hasty crater
#

tried that made no audible difference though i have no way of testing out the electrical characteristics in detail, no scope.

#

oh you mean two non polarized caps? i don't think i have any of those.

fickle leaf
#

Using an anti-polar config protects the capacitor from being damaged.

hasty crater
#

and didn't even know those were a thing

#

or do you mean swapping + and - on one of the caps in line? that doesn't sound good.

fickle leaf
#

They are a thing, but hard to find and afford at > 10uF. Here’s the alternative.

IMG_1319.png
hasty crater
#

you can do that?

#

had no idea

fickle leaf
#

The schematic is from a commercial guitar pedal, so yes. The effective capacitance value is the average of the two.

hasty crater
#

yeah i need to tear the protoboard down and move to a breadboard

#

that is really neat.

fickle leaf
#

The voltage rating should be greater than the signal voltage.

hasty crater
#

hmm 16v caps

#

should i be using ones above 20v then?

#

just to be safe?

#

though it's on the input line for the amp so probably only needs to be above 3.3v

fickle leaf
#

What’s the power supply to the 3w amp?

hasty crater
#

the feather

fickle leaf
#

5v usb or batt or 3.3?

hasty crater
#

powered via usb cable

#

oh you mean the 3w duh, 3v3 out

#

i2s breadboard wires are all going to feather gpio's

fickle leaf
#

So 6v rating minimum would be good

hasty crater
#

currently using 16v 1000uF caps. maybe that's why i got lucky with it?

#

just slapped in the biggest cap i had 😅

fickle leaf
#

Yes, they may be more tolerant of a 3v negative voltage, at least for a while

hasty crater
#

a while? are we talking potential fire hazard down the road or something? for a while sounds ominous.

fickle leaf
#

The cap would probably just dry out more quickly than normal. It might leak some dielectric. Two caps in the anti-polar series config will protect against that.

#

At least for 30 to 40 years, then all bets are off.

hasty crater
#

looks like i'm researching op amps today

fickle leaf
#

You don’t need one for your circuit, just another cap in series with the one you already have.

hasty crater
#

i'm ok with some unoptimal discharge considering the wonky setup. as long as there aren't fireworks i'm ok with that.

#

i did find a 60w I2S amplifier out there, one more than 3W does exist. i just have to go with some obscure company i've never heard of before.

#

i think i have everything i need for now. will get to the breadboarding tonight maybe tomorrow. there's no rush on it.

#

i appreciate your advice, maybe some other time for sure 👍 gonna go lay down for a while i'm beat.

#

and my brain hurts

fickle leaf
#

@hasty crater Don’t want you to misinterpret the circuit example I attached. It was only to show the two caps, not to imply that any other circuitry was needed.

hasty crater
#

Ive been watching some youtube videos on op amps and clearly i have no idea what I’m doing and got very lucky i didn’t blow my little amp up.

wicked sonnet
small plover
glacial spruce
pearl trout
#

Help~! I am trying to develop an audio application using CircuitPython rather than Micropython, because I found an article (https://learn.adafruit.com/adafruit-pdm-microphone-breakout/circuitpython) mentioned CircuitPython supports PDM microphone which my board (M5Stack Atom Echo) uses. But when I ran the source code in the article, I got error said "NotImplementedError: PDMIn not available".

Adafruit Learning System
Adafruit PDM Microphone Breakout

From Sound Pressure to Pulse Density Modulation

solemn flint
#

Support would vary depending on the MCU, most likely, since the PDM input would make use of particular chip-specific peripherals.

pearl trout
#

OK. So it seems unfortunately my board is not supported, right? 😞

wild parrot
pearl trout
#

Thanks for the suggestion. I found the proof from the CircuitPython document saying that "Not available on Espressif.". My board happens to use ESP32 chip. 😂

old sierra
# hasty crater i would have used fritzing to lay it out but last i checked fritzing is no longe...

I use https://github.com/schreibfaul1/ESP32-audioI2S on ESP32 (classic) and ESP32-S3 to play MP3s and for me it worked much better than the I2S on the RP2040. The RP2040 worked ok with the MAX98357A, but there were still random dropouts. I never got the RP2040 to produce good sound with the NAU8822a. But using the same audio board, the ESP produces great audio...

GitHub
GitHub - schreibfaul1/ESP32-audioI2S: Play mp3 files from SD via I2S

Play mp3 files from SD via I2S. Contribute to schreibfaul1/ESP32-audioI2S development by creating an account on GitHub.

hasty crater
#

@old sierra that's for arduino and I'm using circuit python. circuit python doesn't have mp3 support for the ESP32's yet. no i can't switch over to arduino either as there are many other aspects of the project that rely on circuit python. i definitely appreciate the advice though, had I been using Arduino I definitely would have gone with that!

old sierra
regal lava
#

anyone know if audioIO on Python is outdated?

stable iris
#

Really n00b question but I don't really know what words to search for. But how do I know if I can output sound with a board if I can't see a big speaker on it ? As far as I know it would need a DAC, some sort of codec software and some sort of speaker ?

glacial spruce
#

It's a somewhat subtle question. An Arduino can create simple waveforms, even without a DAC, and you can hook those to an amplifier and speaker. With a DAC, you have the opportunity for more wave shapes. You only need a codec if you're demodulated encoded audio (MP3, for instance).

stable iris
#

so without a DAC I'd have to do some work to create an amplifier circuit and possibly a pair of wire with the same resistor and cap for stereo sound ?

#

Or I still have to do that even with a DAC ? And when you say wave shapes does this mean Id have to use a smaller wav file possibly with less information per second in it too or lower frequency ?

glacial spruce
#

You'll need an amplifier to drive a speaker in any case.

#

An Arduino can just create PWM waveforms with rectangular shapes: normally beeps and boops. If you want to play back music files, you'll need a DAC.

unborn drum
haughty zenith
glacial spruce
#

You can get low volume from an ordinary op-amp (one of the AdaFruit boards does this with a TL072, I think). And of course there are simple power op-amps like the LM386.

pastel grotto
glacial spruce
#

Could be a data formatting problem

unborn drum
pastel grotto
#

The click is there not just at the beginning or at the end, but the whole time during the playback. I will check the link anyway, thanks.

unborn drum
pastel grotto
unborn drum
pastel grotto
#

And I added the Audiomixer, to control the volume

#

But the result is always the same

unborn drum
#

Maybe post the code you are using, and a short clip of the audio output?

pastel grotto
#

Okay, will do. I might need half an hour or so

unborn drum
pastel grotto
# 
"""
Audio pin: board.A1
Audio pin: board.A2
Audio pin: board.A3
Audio pin: board.D0
Audio pin: board.D1
Audio pin: board.D4
Audio pin: board.D5
Audio pin: board.D6
Audio pin: board.D9
Audio pin: board.D10
Audio pin: board.D11
Audio pin: board.D12
Audio pin: board.D24
Audio pin: board.D25
Audio pin: board.MISO
Audio pin: board.MOSI
Audio pin: board.SCK
Audio pin: board.SCL
Audio pin: board.SDA
"""

import board
import audiomp3
import audiopwmio
import audiomixer

audio = audiopwmio.PWMAudioOut(board.A1)
mixer = audiomixer.Mixer(
    voice_count=1,
    sample_rate=16000,
    channel_count=2,
    bits_per_sample=16,
    samples_signed=True
)
decoder = audiomp3.MP3Decoder(open("slow.mp3", "rb"))

audio.play(mixer)

mixer.voice[0].level = 0.1
mixer.voice[0].play(decoder)

while mixer.playing:
    pass

print("Done playing!")

I tested the last code from this comment https://github.com/adafruit/circuitpython/issues/5136#issuecomment-1434157983 and there was just the beep sound and no clack sound.

unborn drum
#

One more thing: in your code, your volume is set to 0.1, which is quite low. When you amplify it later back up to the desired output volume, you’ll hear more noise from the processor internals. It would be less noisy to play the mp3 at volume=1 and then turn down the output of your amplifier.

pastel grotto
#

Thank you very much, I had to set it to 3096, for whatever reason and the clack is gone. Just during start and end there is one to hear, just like in the issue you posted above.

Btw: are there alternatives to play mp3 via rp2040 with higher bit rate files and less issues?

glacial spruce
median hemlock
pastel grotto
#

I already tried the dfplayer mini with arduino, but it... slow. It needs a few seconds to be useable and thats the reason why I tried the PAM8302.
But thanks for the suggestions
btw:

NOTE: This is not currently working for audio playback of files. Only sine wave test currently works. The problem is that pure Python code is currently too slow to keep up with feeding data to the VS1053 fast enough.

stable iris
# odd oasis Ohhhhh

so follow up about it, completely embedded in textile under it there is 4 screws, the textiles even go inside and is sewed hard

#

so I couldn't have removed it without tearing it, also there is streaks of caked yellow glue so it was also glued to the plastic of the headset

#

Im opening it but I expect an epoxy tomb with unmarked chips

unborn drum
#

Id like to use synthio.Synthesizer. The docs say its experimental, so i tried the 8.1.0 beta but got the same results:AttributeError: 'module' object has no attribute 'Synthesizer'

>>> import synthio
>>> dir(synthio)
['__class__', '__name__', 'MidiTrack', 'from_file']```
random bone
unborn drum
#

Thanks for the quick reply. Absolute newest- I feel so cutting edge!

unborn drum
#

Hi again, audio channel 🙂 I'm generating two short (one cycle) tones and using audiomixer to play both tones at once. When i play either tone by itself, the sound and the visualized waveform is smooth. However, when I play both tones, I get this regular glitch near the zero-crossing.

glitch.jpeg
random bone
#

if "less visible" tone was about 1/5 the frequency of the other one?

unborn drum
# random bone if "less visible" tone was about 1/5 the frequency of the other one?

They are different tones (440 and 350hz), together they make dial tone. When I combine them in protools the resulting waveform is smooth, same result if I graph it with matplotlib. Here's the code that results in the output with periodic glitch.

    samples = sampleRate // frequency
    buffer = array.array("h", [0] * samples)
    for i in range(samples):
        angle = 2 * math.pi * i / samples
        buffer[i] = int(math.sin(angle) * 32767)    
    tone = RawSample(buffer, sample_rate=sampleRate)
    return tone

def dialTone():
    if not audio.playing:
        audio.play(mixer)            
        mixer.voice[0].level = .25    
        mixer.voice[1].level = .25
        mixer.voice[0].play(tone440, loop=True)    
        mixer.voice[1].play(tone350, loop=True)

sampleRate = 32000

mixer = audiomixer.Mixer(buffer_size=1024, voice_count=2, sample_rate=sampleRate,
                        channel_count=1, bits_per_sample=16, samples_signed=True)
tone440 = generateTone(440)
tone350 = generateTone(350)
dialTone()```
#

I have a feeling that I might be generating one too few samples to complete the cycle of one of the tones?

unborn drum
#

I love a good plot twist. Changing the sample rate to 31000 smoothes it right out

noglitch.jpeg
random bone
#

If you could open an issue with the minimum reproducible case, that would be great. The person who is most familiar with this is away for a few weeks but it would be good to get it written down in an issue

#

thank you

unborn drum
random bone
#

it may be some kind of granularity issue

unborn drum
#

Either way I'm happy enough with my solution. I'm sounding like Ma Bell over here now

#

Btw I tried using the new Synthesizer library in the latest build to do this, and it was fun and easy. The output wasn't clean enough for this purpose, but it was musically exciting 🙂

random bone
#

glad to hear that, and we welcome suggestions, since it's brand new

dull basalt
#

Hi, I'm trying to record my voice with a PDM MEMS microphone (like this one https://www.adafruit.com/product/3492) , record it in a array , and finally play it on a I2S AMP (like this one https://www.adafruit.com/product/3006) . Micro and I2S amp are plugged into a rasperry pi pico. it works but the sound that i heard is faster than my recording voice, so the sound is more acute. Anyone as an idea to solve this problem ? thank you. I use

glacial spruce
#

You may need to adjust one or both sampling rates to so they match, or (worst case) resample the audio to a different rate in-between

dull basalt
#

Thank you but everything is adjust to the same sample rate

dull basalt
#

ok I double sample rate from the PDM MEMS, it seems to works

random bone
tall mango
#

Not necessarily hardware related but I'm just wondering how I can reduce noise on my microphone. I have a Blue Yeti and I've been trying to record samples of each key on my grandma's old upright piano to use in Ableton since it's too broken and out of tune to play normally. I'm planning to retune the piano samples digitally but the problem is that there's a lot of noise coming from behind the microphone at my computer tower and laptop. I've tried the Stereo and Cardioid modes on the mic which seem like they'd be the better options, but they both have similar noise levels.

lucid anchor
#

Is it actual ambient noise or is it electrical noise? Microphones can record noise even in a perfectly silent room if there are electrical sources that are contributing to it.

#

You want whatever mode only records in front of the mic. I can't remember which on the Blue Yeti that is.

#

(I'm mostly a noob with audio, but have been learning as I go. I also have a meeting in 5 minutes. So I'll help as I can, but you'll probably have to wait on others with more knowledge than me to get more specific assistance.)

tall mango
#

I think cardioid is the mode for that. I think it's mostly noise from my CPU and GPU fans in my computer.

lucid anchor
#

Ahh.

#

Then you could put some kind of baffle between the mic and the sound source.

late arrow
#

The Blue Yeti is so sensitive, it picks up everything. :/

tall mango
lucid anchor
#

Would need to be big enough to block around the mic though.

lucid anchor
#

Try turning down the gain in that case, perhaps?

#

My piano is pretty loud relative to ambient noise, so I feel like you could get away with that.

tall mango
#

I think i left the gain about 1/3 turned up? I'm not at my computer right now but i remember setting the gain to be just loud enough for the piano to be about the same volume as it would be to my ears.

#

If that makes sense

lucid anchor
#

That does make sense.

tall mango
#

If only I could just steal a wall and put it in front of my computer lol

lucid anchor
#

Hehe. Is everything in the same room then?

#

How difficult would it be to move your computer? (Moving a piano is pretty much out.) Basically, get a long XLR cable or whatever the Blue takes, and run it out of the room to your computer while recording.

#

A lot of shenanigans to do that though.

tall mango
#

Yeah my computer, piano, and mic are all downstairs in the "basement" (looks like a basement, but isn't actually one) which already has awful echoing

lucid anchor
#

Oof.

tall mango
lucid anchor
#

Yes! That might work.

#

XLR cables can be pretty long before there's audio degradation. So no worries there.

tall mango
#

I'm definitely gonna get a USB extension cable since i somehow lack them

lucid anchor
#

Ahhh yes those are super helpful.

tall mango
#

The standard blue yeti isn't XLR sadly :(

lucid anchor
#

Ah ok.

#

USB can also get pretty long without issues.

#

I like Amazon Basics USB extensions, fwiw.

tall mango
#

Yeah i had a few of their cables at some point and they're really nice

#

Affordable too

lucid anchor
#

Yep!

#

Well, that's what I would suggest. Give that a try. Removing the noise in the room might be your only option with the Blue Yeti being so sensitive.

#

Good luck! Meeting time!

tall mango
#

Thanks! Good luck on your meeting.

#

Might actually just buy some sticky foam at the local hardware store and 3D print a noise shield to stick it to because I'm broke right now.

tall mango
tall mango
#

46 hour print time 😬

#

At 0.3mm layer height too

glacial spruce
late arrow
fickle leaf
#

For some rooms, diffusion can work better than absorption (foam) for reducing echo. It’s easier, too. Bookcases full of unaligned books make for a wonderful diffusion panel. They help eliminate reflections from parallel surfaces. Room clutter is surprisingly effective, too.

#

My spouse’s library/office became a go-to as a preferred vocalist booth for a few clients in spite of the non-carpeted floor.

tall mango
#

I'm actually just trying to reduce noise around my microphone, echo isn't really that much of a concern for what I'm doing.

shrewd sequoia
#

anyone knows of a good but cheap usb audio card for a pc daw setup --- thanks

random bone
#

what's your budget?

#

read the reviews carefully to find any gotchas

shrewd sequoia
#

ordered a pci-e card, mostly cause of the cost-$23.00 canadian. it's a weastlinks card with a cmi8736 chip, an old one but seems good, does any one have it? i'll let you know if it's any good when i get it.

glacial spruce
#

You had asked for a "USB card", and PCIe is a different kind of interface, but if your host machine supports it, should be fine.

sleek salmon
#

I’ve used the sound fx board with amp connected to a trinket pro to build this but the audio is extremely low, like can hardly hear it even when it’s quiet in the room. Any way to make it louder? Oh and I have tried sending the + command to increase the volume and it’s at the max already

glacial spruce
#

You should be able to get reasonable volume with that setup. I'm guessing either your audio file is recorded at a low level, or you have something connected only, so it's not working as effectively as it could.

south plaza
#

Are versions of the MAX9814 board sold that don't include the microphone?

#

I plan on wiring my own up

#

My end goal is to record some audio from a microphone and save it to an SD card

#

And beyond the automatic gain levels, are the MAX4466 and the MAX9814 basically swappable?

#

Should I just desolder the built-in mic?

#

There's this third party board but I don't know why this has so much more on it compared to the official one

image.png
glacial spruce
#

It doesn't look like there's a lot there, other than a local power supply with a linear regulator (a good idea with low-level signals like microphones) and some jumpers, along with the IC itself

south plaza
#

Though the official MAX9814 board is pretty minimal too

#

So should I buy one like the official one (I'll probably get it off aliexpress) and desolder or should I get the bigger one that doesn't include the mic?

south plaza
#

I've looked into it some more and I can't really find more info about the 3rd party DIYMORE design

#

But I also don't know if desoldering the other one is as straight forward

ivory turtle
#

Hello there!
I'm new to audio on a Pico. I have a Pico running the official MicroPython 1.20 and want it to play MP3 sound files. Initially I wanted to buy the PAM8302 from Adafruit, but it wasn't awailable, so I bought the MAX98357A. But I think I might switch in the near future, when I find a PAM.

Anyways: Is CircuitPython needed in any case to run any of these two boards? I couldn't find any downloads for the libs, that are recommended by the official tutorial on the adafruit site (audiomp3, audiopwmio, audiocore, audiobusio).

random bone
#

The audio... modules are built in to CircuitPython. The MAX98357A is an I2S amplifier, so as long as whatever software you have can do I2S, then it will not work. CircuitPython, MicroPython, the pico-sdk, and the RP2040 Arduino all can do I2S.

The PAM 8302 is an analog amplifier. You'll need to use PWM audio or analog audio (but there is no DAC on the RP2040) to use it.

ivory turtle
random bone
#

The audio... modules are part of the CIrcuitPython firmware. They are written in C.

I had not heard of picozero. CircuitPython includes most of that functionality as libraries or built in, in a different form, of course. We have an IR receiving library. It would not be hard to convert your shift register library; there is also a library :https://github.com/adafruit/Adafruit_CircuitPython_74HC595

ivory turtle
#

Ok, cool!
So I'll copy my old version and try to translate it using CP then 🙂

Can I come back here, if I face problems during that process?

random bone
ivory turtle
loud scroll
#

Does a Low pass filter improve pwm audio

#

if so, what values do i use?

harsh gust
#

I have been trying to build this circuit successfully since last year

glacial spruce
# loud scroll Does a Low pass filter improve pwm audio

It depends on what you mean by "improve". It can smooth out the digital pulses to a (somewhat) more analog voltage. The values to use depend on the PWM frequency, what kind of response speed you need, how much ripple/harmonics you can tolerate, etc.

harsh gust
#

I kept failing over and over again. By now I have made 6 copies of em in total. 4 of them are working, and 2 of them are not working properly.

#

But the rest of the 4 don't seem like they are amplifying. Instead, it seems like they are actually reducing the volume.

#

it sounds louder when I plug in the piezo directly.

#

Is there any reason for this?

glacial spruce
#

It's a slightly odd circuit, with some frequency peaking circuitry, so it appears to be aimed at amplifying some frequencies, but not others. It also acts as an impedance converter, so it will show more apparent gain into a low impedance load than a high impedance load.

harsh gust
#

could you elaborate more on "showing more gain into a low impedance load than a high impedance load" pls?

#

I cannot get it

glacial spruce
#

The piezo microphone is a "high impedance" signal source, meaning it cannot provide much current. If you run its signal into a high impedance load, it will deliver its full voltage, but into a low impedance load, the load will soak up the current which will keep the voltage from getting as high. The op-amp circuit depicted provides current gain, so it can drive full voltage into a lower impedance load.

harsh gust
harsh gust
#

If you are not busy and have spare time, do you mind skimming through mine and Richard's comment

glacial spruce
#

It seems that his use case (hydrophones) is different from yours (monitoring ants, IIRC)

harsh gust
#

It's almost at the very end

#

ah-ha

harsh gust
#

for my use

glacial spruce
#

His comments indicate he's having the same thought that I was, that you could be losing signal to parasitic oscillations. These low noise high impedance amplifiers are very sensitive to circuit configuration, wire length and placement, etc.

#

As for "better without it", that depends on if you're getting sufficient signal without it. If you are, yes, you're better off.

#

His comment about capacitance and pin 2 is a little confusing, as there's a big capacitor hooked to pin 2 (decoupled by some resistance). Come to think of it, I might be tempted to try something like a transimpedance amplifer instead of the voltage follower in the circuit depicted.

harsh gust
#

But I would try with proper setup tomorrow and see if it's helpful to use it or not

harsh gust
glacial spruce
#

He said "You are asking for trouble doing that, you don’t want any more capacitance to ground from pin 2 than you have to."

harsh gust
#

yeah right

glacial spruce
#

But there's C1, connected to pin 2 (via R4)

screenshot_1554.png
harsh gust
#

yeah

glacial spruce
#

C1's reactance, along with R4 and R5, will give a varying frequency response, with more gain at some frequencies than others. This is one of the effects I was referring to earlier.

loud scroll
glacial spruce
#

Sorry about that. I wasn't quite sure what sort of an answer you were looking for. A low pass filter can sort of "smooth out" a digital PWM waveform. Whether this is an improvement depends on what you're doing with it.

loud scroll
#

What about PWM Audio?

#

it sounds pretty horrible

#

and recently i found this page like " USE LOW PASS FILTER TO SMOOTHEN OUT PWM!1!!!!!11!!1!1!1"

glacial spruce
#

A lot of microcontroller PWM audio does sound pretty terrible. However, there are fancy forms of it and special techniques that can sound less-terrible to quite good.

loud scroll
#

Can you give me a few methods to make it sound less terrible?

glacial spruce
#

It rather depends on what you're trying to accomplish, and what you're using to do it. Are you trying to play back music or speech, or generate video game tunes, or just beeps and boops? Are you using an Arduino, or RP2040 or a Teensy 4?

loud scroll
loud scroll
#

can i read an analog pin and write to a screen WHILE i play audio?

glacial spruce
#

That I don't know offhand. Since the RP2040 is dual core and has PIO functionality, it seems like it might be possible, but could be pretty complicated.

loud scroll
#

Dual core isn't yet supported by circuitpython, i think that is coming soon in r9 or r10

#

as for PIO i'll need to set up PIO with SPI

#

which im not the best at, i just found out what PIO is

glacial spruce
#

I'm not good at PIO either, I found it sort of mystifying until a recent article in Elektor magazine showed up with several well-explained examples.

odd oasis
#

I believe the audiomp3 functionality in circuitpython is non-blocking, but I’m not sure how it’ll affect the execution of other commands.

loud scroll
#

it doesn't, but everything slows down

echo hedge
#

We'll likely never run python on both cores. Instead we'll use the second core for supporting C code. The PicoDVI support is the first to do it.

harsh gust
#

It's kinda feeling empty and depressing that I have devoted so many hours on building this circuit since last year fall,

#

now I have 2 days left for my school expo exhibition

#

and just figured out it's not a proper circuit for amplifying the sound of ants

#

💀💀💀

glacial spruce
#

Yes, you can use your existing op-amp with a high value feedback resistor to build a transimpedance amplifier

#

I hate having to redesign with 2 days left

glacial spruce
#

A transimpedance amplifier is pretty simple.

screenshot_1557.png
harsh gust
#

so I have replaced R7 with 100k pot

#

and have LED and 2k resistor parallel to the circuit

glacial spruce
#

You'd connect pin 3 (the non-inverting input) to the junction of R2, R3, and C2. Then you'd connect the piezo sensor to pin 2 (the inverting input). Disconnect R4, and replace R5 with a high value resistor (or pot). The higher the resistor, the higher the gain.

harsh gust
glacial spruce
#

That just biases the output around ground.

#

Which could be part of your problem, depending on what you're connecting it to.

harsh gust
#

then what should I do with R7(100k pot)?

#

ahhhh Idk what to do rn🥲

tall mango
#

What would be the simplest way to connect my turntable, cassette deck, etc. to my speaker set without having to swap cables or use janky splitter cables? I've seen a few of those splitter boxes online but they're usually like $50 and have only like four outputs.

weary gyro
# tall mango What would be the simplest way to connect my turntable, cassette deck, etc. to m...

you need a preamp with a "phono input" for your turntable (turntables require special inputs) and a "line level input" preamp for devices like a cassette deck. Then these preamp outputs go to a mixer and the output of the mixer goes to an amp to drive the speakers. While you can have all those functions be separate boxes, most people get a "home stereo amp" that combines all those functions into one device. For instance, this device has all those features (Not recommending it, just an example) https://www.amazon.com/Bluetooth-Amplifier-Multi-Channel-Amplifier-Remote-Pyle-PDA4BU/dp/B081QGYJ9C/

Screenshot_2023-05-02_at_5.34.24_PM_copy.jpg
tall mango
#

My turntable has a built in preamp, but I just need an option to basically 'combine' two already line level signals.

glacial spruce
#

You can build a passive combiner out of resistors, or a simple mixer with op-amps

unborn drum
tall mango
#

Oh i guess that would work

unborn drum
#

Or maybe an input switcher

image0.jpg
loud scroll
#

Can i use a Circuitpython device as a USB Microphone?

echo hedge
#

No, CircuitPython can't stream audio in

dusty rivet
echo hedge
#

we haven't implemented it

#

it'd be similar to streaming audio out

dusty rivet
#

Gotcha

radiant axle
#

Heya, i'm working with a Adafruit PDM MEMS microphone and i get this audio, do you know what can i do to change that ?

glacial spruce
#

Sounds like AC hum, possibly a power supply, ground loop, or noise coupling issue

kind cedar
radiant axle
#

If needed I can send it here

glacial spruce
#

I doubt it's code related. It sounds like a bad power supply to me.

radiant axle
#

Maybe i'm not using the good pins ?

glacial spruce
#

It could be a bad connection somewhere. For example, if the ground isn't connected, it'll end up grounding itself the long way, which can bring in tons of noise

radiant axle
radiant axle
#

Update: Beside the unwanted noise, when i snap my fingers i can hear the sound when listening to my file (but can’t hear my voice), idk if it can help solve the problem ?

dull basalt
#

hello, does somebody explored pitch shifting with circuit python ? I'm trying to record a signal with a MEMS microphone , then transforme this signal in each pitch (A, A#, G, etc...)

unborn drum
rare torrent
#

Figure I'll ask here first but anyone have any recommendations for a 25(ish) key midi keyboard. Basically to use as potential input to CP projects. Extra points if the keyboard can take input to play project output too.

late arrow
#

The Arturia and Akai seem to be good value for the money in my research and I got a couple of people who liked the Akai, including BlitzCityDIY. Maybe check out the Akai mini, comes in a 25 key too

rare torrent
#

Which version is the 37 key? or do they just have one (which I can find I'm sure).

late arrow
#

I bought the Akai MPK mini plus

rare torrent
#

I was looking at the Akai ones earlier today. I have my big 88-key but a little less convenient to move about!

#

Thanks

late arrow
#

ha, me too. I told my wife I wasn't moving our electric piano into my office

rare torrent
#

Ha yeah i get that. Funny now how my keyboard still works well it's 15 years old and how tech has changed so much.

late arrow
#

I only bought the mini plus because it had CV / Gate inputs in case I really get into this and go modular, but that's a long ways out

rare torrent
#

It has the midi ports too I notice, which is something I could see being useful (for me)

late arrow
#

yeah, I really wanted MIDI too, Been building todbot's picostepseq this weekend. 🙂

rare torrent
#

Ha so much talk on synths lately I decided to look what's about.

late arrow
#

great minds think alike!

glacial spruce
#

I went for an M-Audio keyboard, seems to work well for me.

fickle leaf
#

I’m a fan of the M-Audio semi weighted controller. For quick modular input, been relying on the Arturia key step.

sour bronze
#

Could I hook these up to an arduino or esp32?

rn_image_picker_lib_temp_eb11ee5f-900b-45eb-ab92-a449450e9aca.jpg
odd oasis
#

I suppose you could, but the pinouts aren’t public knowledge so you’d have to kinda figure out how to use it properly.

#

Not sure which pads connect to VIN or its output, let alone what type of interface it uses…

glacial spruce
#

You might be able to use transistors to switch the signals, but you'd have to work out the polarity and so forth.

south plaza
#

So now that I'm experiencing noise in my microphone setup, I did some tinkering and figured out that it's caused by this code:

    int r = analogRead(A0);
    // If audio is detected...
    if (r < silenceMin || r > silenceMax) {
        if (isRecording) {
            silentChainSamplesCount = 0;
        }
    } else {
        if (isRecording) {
            silentChainSamplesCount++;
            if (silentChainSamplesCount > 6000) {
                Serial.println(silentChainSamplesCount);
                stopRecording();
            }
        }
    }

When I comment this out, the audio noise I get from my microphone is negligible, but when I add it back, the noise does as well. I suspect it's related to analogRead putting the noise on the ADC, which is picked up by the preamp. How should I limit this noise (or stop it all together if possible)?

I'm hoping to do it with parts on hand but I'm not sure where to begin. My only reference is a video which mentioned some capacitors and a low dropout regulator (whatever that means).

I also found this SO post which mentions some capacitors, but that presumes some external voltage regulator (which I don't think applies in my case)
https://arduino.stackexchange.com/questions/69133/noisy-analogread

glacial spruce
#

Noise can come from the power supply, noise pickup (capacitive, inductive, and radiative), coupling from nearby digital circuitry, ground loops, and more. You can filter the power supply with additional capacitance or even a decoupling network. You can also add capacitance to the signal, which will absorb higher frequencies, which can help if the noise is high frequency and your signal is low frequency.

unborn drum
#

I'm trying to build a super simple audio amplifier, more or less like the image, except my source is a Pi Pico. When i eliminate C1 and R1, aka connect the audio pin directly to the base of my transistor, i can hear my audio fine. But if I attempt to follow the diagram and connect R1 between the collector and base, and put C1 in series with my audio pin, i hear only a faint hissing. I've tried putting the circuit together several ways, and im relatively certain I have the pins correct. Can anybody shine some light on what I'm missing?

Screenshot-2020-10-10-at-5.png
dusty rivet
#

Have you tried something like 1uF?

#

I also suspect you might also want a 10k between base and emitter

unborn drum
#

same outcome.

dusty rivet
#

A quick google search turns this up

#

Seems like you’re pretty close

random bone
# unborn drum ive tried a variety of caps, including 1uF

C1 is polarized, but the positive end is an audio signal that is probably less than 9v. Try a non-polarized cap. The idea is to superimpose an AC signal onto a biased DC voltage

Normally I would expect two bias resistors forming a voltage divider to allow the amp to operate in a linear region

unborn drum
#

for the sake of clarity here are the circuits im actually using 😄

Untitled.jpeg
random bone
# unborn drum I'm trying to build a super simple audio amplifier, more or less like the image,...

Neither circuit is great. I don't know where you got the original schematic, but I would expect more components, as I mentioned (and so did skerr). Search for "class A transistor audio amplifier". Find a textbook or project book. Random schematics off the internet are often dubious.
Alternatively you can use an IC audio amp, inexpensive and more foolproof. LM386 is classic but there are more modern ones

#

Adafruit sells several audio amp breakouts. Also note the PWM audio has a carrier frequency superimposed on it and you may want some low-pass filtering

unborn drum
random bone
glacial spruce
#

I'm also not sure about returning the bias resistor to the load instead of the supply.

unborn drum
#

No audio through the cap

#

And just for fun, here’s the bare c1815 with nothing else 💁‍♀️

glacial spruce
#

This "Gain without Gamble" article from the May 1974 issue of Elementary Electronics explains some of what's going on. It describes essentially the amplifier you posted as the "Russian Roulette Special" and covers why.

📎 Transistor-gain-without-gamble.pdf
unborn drum
cursive beacon
#

I really don't know much about audio, but in an effort to eliminate the pop that starts when playing WAV or MP3 files in CircuitPython I've switched to playing files using a single channel with audiomixer. The initial pop/tick seems to have gone, but I'm hearing an intermittent tick - sometimes during playback, sometimes when no playback is happening. You can hear it in this video clip about 1 sec (with no audio yet playing) and 4 sec (with audio playing). Right now I have a sloppy 3.5 mm plug alligator clipped to audio in and ground pins in the build. I'm waiting for breadboard friendly stereo jacks to arrive so I can add this for a more secure "plug-in" build. Should I expect that tick sound to go away? It even ticks occasionally with nothing playing, which is annoying. Is there something else I can do to eliminate the tick & make sound seem more smooth? For context: the button is an MQTT publisher & the build with the off-brand neopixel strands & audio is the MQTT subscriber. Both use Raspberry Pi Pico Ws. The music is played from an microSD card / reader wired onto the subscriber. Thanks for advice. Again, I'm really not skilled in audio, so terms may be new to me. Thx for newbie-friendly explanations.

cursive beacon
#

Update - cleaning the home office/lab & ran across a breadboard-friendly stereo jack. Added it, but still hearing the random ticks. Any idea how to eliminate these? Thx!

glacial spruce
#

It may be worth looking into the possibility that it's power supply noise being coupled into the audio signal

random bone
cursive beacon
# glacial spruce It may be worth looking into the possibility that it's power supply noise being ...

Thx. The speaker is powered separately - it's plugged into an iPhone-style wall charger. I've also tried some small hamburger-style speakers, both plugged into USB chargers & after being charged & running off the battery. Same ticks. And the ticks only seems to occur when the code is running, even if nothing is playing through mixer. @random bone I stop playing with: mixer.voice[0].stop() I'm only using one voice. I do start the audio mixer with this code & then I use the stop line above if we start and nothing should be playing:

mixer = audiomixer.Mixer(voice_count=num_voices, sample_rate=22050, channel_count=1, bits_per_sample=16, samples_signed=True)
audio.play(mixer)```
Whenever I need to begin playing a song I use the code below reading in the wave fresh so that I start the file from the beginning. This could play one of several files, or begin playing the current or last file from the beginning, as well.
```p
def play_voice(filename):
    path = "/sd/disco_songs_wav/"
    if mixer.voice[0].playing:
        mixer.voice[0].stop()
    # read in all beats & simultaneously play them at audio sound .level = 0 (no volume)
    print(f"About to play: {path+filename}")
    try:
        wave = audiocore.WaveFile(open(path+filename,"rb"))
        mixer.voice[0].play(wave, loop=False )
        mixer.voice[0].level = 1.0
    except Exception as e:
        print(f"There was a problem playing audio: {e}")```
#

Ticks don't seem to be in the same place.

random bone
#

i have heard random noise that does seem to be electrical, but I've also heard random noise that seems like intermittent software issues. The audio code has been "gotten to work". It could use a cleanup and improvement: https://github.com/adafruit/circuitpython/issues/5136 and many other issues. The mixer stuff is a workaround.

cursive beacon
# random bone i have heard random noise that does seem to be electrical, but I've also heard r...

I'm currently using audiomixer because I found that the mp3 playback had the consistent pop at the start of each track. Audiomixer seems to get rid of the pop, as reported above, but I've still got those intermittent pop/click/crackles that happen. I've tried four speakers so far. Two hamburger style, two larger USB-only. All have their own power source. Is there anything else I might do to counteract this? I have a higher quality Bose speaker from one of those maker kits the company sold a while back. It's at my school office, but I can give that a shot. I also am using a breadboard. I'll eventually solder things to a protoboard, but unsure if that really has any impact. Thx for helping me think through issues. I'm also seeing @tardy jetty 's Breakbeat Breakbord linked below seems to have some additional components between the audio jack & the microcontroller, which I don't have. He's got two resistors & a capacitor. I haven't tried this, but will give it a shot & see if I notice any improvement. https://learn.adafruit.com/breakbeat-breadboard/build-the-breakbeat-breadboard

Adafruit Learning System
Breakbeat Breadboard

Build this simple button board to play synced wav files using CircuitPython's audio mixer

random bone
cursive beacon
random bone
# cursive beacon Thx - I'm using PWMAudioOut. I've not used I2S audio before. Relying on the spea...

PWMAudio is a significant compromise -- it loses audio fidelity. Eventually try an I2S breakout like https://www.adafruit.com/product/3006 and see how it works (that one is mono -- the stereo I2S breakout is discontinued unfortunately because of a discontinued part). The upcoming Feather RP2040 Prop-Maker has an on-board mono I2S amp.

Adafruit I2S 3W Class D Amplifier Breakout - MAX98357A

Listen to this good news - we now have an all in one digital audio amp breakout board that works incredibly well with the Raspberry Pi! If you're looking for an easy and low cost ...

cursive beacon
#

Ah! I have one of these. I tend to buy a bunch of parts when attempting something new, so I've got this but haven't worked with it. What I'm hoping to do is still have a 3.5 mm audio jack so I can plug in an external speaker. The learn guide has a lot of terms I don't understand, but I do see the section for "Speaker Output" has connections labeled - & +. Could I wire those to the ground & output pins on the audio jack I'm using, completing the rest of the wiring as described in the learn guide? I see the section has this language, but frankly, I don't really have the engineering knowledge to understand this:
"This amplifier is designed to drive moving coil loudpeakers only. Speaker impedence must be 4Ω or more. The output signal is a 330KHz PWM square wave with a duty cycle proportional to the audio signal. The inductance of the speaker coil serves as a low-pass filter to average out the high-frequency components. Do not try to use this as a pre-amplifier."
Also, all the speakers I'm using have stereo audio plugs & my jack is a stereo jack (even though these are single speakers). I notice these parts say "mono" Amp. As long as I use a single speaker and have a mono file, would this still work?

cursive beacon
#

And while I'm trying to work with the current build I'm using, I've noticed that the ticks won't start at all until I audio.play (again, they don't start right away after .play, but do begin seemingly randomly after I start a .play), so I changed my code to create the Mixer, but don't audio.play(mixer) right away. The speaker remains quiet without random ticks until I .play. But I need to include an audio.play(mixer) when I want to play a song & this creates that same initial pop that I had when playing mp3 or wav files without audiomixer.
I thought about deiniting & recreating the mixer, but couldn't get this to work without a glorious, ear-piercing crash, and even if I did get this deinit/recreate to work each time I wanted to play a song, that wouldn't solve the problem of getting a tick at the start of the first audio.play.
I've tried stopping play with both mixer.voice[0].stop() and mixer.stop_voice(0). These seem to result in the same thing. The music play stops, but I can still hear a random tick every once in a while.

cursive beacon
#

So I tried to see if I could mimic @tardy jetty 's wiring with the 1ohm resistor & the 2.2mf capacitor. https://learn.adafruit.com/breakbeat-breadboard/build-the-breakbeat-breadboard I only had a 10mf capacitor, but I figured if it worked, that'd be a good sign. When I tried this, I could hear the sound was muffled, which I assume was correctly filtering out the high frequency portion of the sound, BUT there was NO CHANGE in the ticks. They still occurred, still sounded high, even thokugh the capacitor was used. SO I assume the additional circuitry does't help.
If it's useful for Adafruit folks, I'm going to start with a barebones pico & audio output. I'll play mp3 and wav using PWMAudioOut, then show audiomixer, and illustrate where these ticks/pops occur. Haven't tried the amp and I2S yet.

Adafruit Learning System
Breakbeat Breadboard

Build this simple button board to play synced wav files using CircuitPython's audio mixer

random bone
#

did you mean 1 ohm or 1kohm?

cursive beacon
#

Yes, 1k ohm, mistyped. Sorry. I always check with a multimeter since my vision is so poor - I can never read the bands properly. I think I constructed the circuit properly because when I played back the sound was muffled - I'm assuming this is filtering out the high frequency. I didn't have a 2.2mf capacitor so I used a 10mf one & the muffling was definitely noticeable. I assume if this worked as hoped, then I wouldn't have heard the intermittent ticks, but they sounded the same, no muffling - still high pitched. Still high-pitched. I'm prepping up a Pico with a .wav and .mp3 I can share and I'll try the three options - mp3 & wave without audiomixer & with audiomixer to demonstrate when the ticks show. I'll record the playback on video. I'll be sure to share the small code samples & sound files I use. I'll open a CircuitPython issue once this is documented.

random bone
#

My color vision is non-standard; I have to use a multimeter too

cursive beacon
# random bone Intermittent ticks are definitely glitches, not just PWM artifacts.

Strange thing. I just downloaded the two .mp3 files from CIrcuitPython Essentials - happy.mp3 & slow.mp3. These are much smaller files than the ones that I'm using, but when I play back these mp3s, even repeatedly, I get the same tick/pop when the file first starts, as expected, but I don't get any ticks/pops randomly when the files are playing. These files seem to play at a lower frequency (not sure if I'm using that term right) - they sound more "bassy" than any of the Disco tunes I'm playing back. But if there were an issue with my .mp3 files, I would think the tick would appear at the same spot in the file, but it doesn't. Still random.

random bone
cursive beacon
#

Hmm. I didn't change the bit-rate, it remains at 32-bit, but before you responded above I tried lowering the project rate in Audacity from 44100 Hz to 22050 Hz, which I know is what we need for wav files, but mp3s say 44100 is at the top of the range. Quick playback of the first file that I converted to the lower sample rate and I had no pops other than the start. Will continue to experiment with these other files to see what's up. Never worry about telling me something I know. I actually don't really understand the difference between the bit format & sample rate.

glacial spruce
#

MP3 doesn't support 32-bit encoding that I know of.

random bone
#

the kbps rate determines how much computation needs to get done to decode the MP3.

#

This is separate from the "sample rate", which is about the sampling rate of the original uncompressed audio, or the playback sampling rate.

glacial spruce
#

Also different from the sample depth, which is the number of bits per sample. Most formats use 8 or 16 bits per sample.

cursive beacon
#

I tried lowering the bit rate & didn't see some benefit, but when I changed the circled value from the original 44100 to 22050, the random playback clicks were eliminated (except the click at the start). I then tried 32000 & this seems to work as well.

image.png
random bone
#

It is possible there are artifacts being generated somewhere in the conversion process, and it's not a bug per se. If you do file an issue supply us with the original files and we can try to duplicate.

cursive beacon
#

@random bone and @glacial spruce I just tried updating the wav files, which we originally 22050, but used Audacity to move down to 16000. Changed the sample rate when setting up my Mixer in CircuitPython to the same 16000, and now I don't hear any of the background ticks, neither when no sound is playing, nor when the songs are playing. I now only hear a tiny tick when the code first starts & I set up the mixer when the code starts, like this:

mixer = audiomixer.Mixer(voice_count=num_voices, sample_rate=16000, channel_count=1, bits_per_sample=16, samples_signed=True)
audio.play(mixer)```
And that's not really noticeable since it only happens when the code first runs.
So the audiomixer with wav seems like the best solution since this gets rid of the pop at the start of each sound play. I've saved those original files & will share. I bought these files via iTunes over the weekend, then used Audacity to convert to WAV & MP3 files, but something strange about those higher sample rates must have caused the ticking issue - and it's so weird because it was both random, and it showed up when no file was actively playing in AudioMixer, but . It occurred across all four files - "Boogie Wonderland - Earth Wind & Fire", "Disco Inferno - The Trammps", "Play That Funky Music - Wild Cherry", and "Stayin' Alive - the Bee Gees".  
BUT here's something to watch for. If I change the sample_rate in the line above, even BEFORE I actually read in any files & play it on voice[0], I get the random tick/crackles. So all that executes in my code are those two lines above. No files are played on any voices. I noticed this without changing any of the files back to the larger format, I just changed 16000 in the line above to 22050. No sounds playing, but those random ticks occur. Weird.
glacial spruce
#

I wonder if it's clocking out a blank buffer at that speed when nothing is playing

severe bison
#

Thanks for the discussion @cursive beacon et al & for pointing me here @late arrow. Was having some related problems and this likely saved me a lot of time!

cursive beacon
# severe bison Thanks for the discussion <@217065986384330763> et al & for pointing me here <@2...

Ah - so glad it helped. Didn't see @late arrow 's post, but he's a very good human who loves music. Hopefully in the next few weeks I'll update my tutorials on YouTube for some info on using audiomixer for playback without pops/ticks. I wish I knew enough re: why the higher sample rates were causing random ticks, which is so weird. My limited-knowledge brain suggests the ticks would happen at the same place with a sample rate difference. I clearly have much to learn re: audio. Cheers!

severe bison
fair forge
#

Hi y'all!
I have the max98357A and the link to the learning guide is 404 error.

Anyone have an alternative link or other documentation on this little thing?
Thanks!

desert fog
fair forge
#

Thank you!

severe bison
#

Testing modified code from John Park's Breakbeat Breadboard. It works nicely as far as the popping, however, it continuously loops through clips. Is there a way to set this up so it starts at the beginning of each clip when it starts playing?

#

OK, I think I figured it out - code is rough/modified, but seems to ~work. Basically add the mixer.voice play thing to the handle_mixer function:

image.png
#

(maybe it will help someone else)

#

Still pretty rough code at this point

severe bison
#

Wonder if increasing the RP2040 clock speed would affect anything.

weary gyro
#

Hi @cursive beacon & @severe bison, you can also reduce clicking by increasing buffer_size in AudioMixer. It’s default is pretty small and I usually set it to 2048 or 4096 without introducing too much latency

cursive beacon
severe bison
#

So in my experimentation, I've been sending 3.3V outputs directly from my RPi Pico to a guitar amp. Seems to work fine, but is there any reason I shouldn't be doing this?

#

My limited research on line level has been slightly confusing TBH.

desert fog
#

Depends on the amp I imagine. If it's just digital inputs it shouldn't be a big problem assuming it has TTL thresholds and doesn't draw more than a few mA.

severe bison
#

Basic question here, but CircuitPython et al has to do a lot more processing work to output sound from an MP3 file than an uncompressed WAV, correct? Obviously size is the big drawback.

echo hedge
earnest dune
#

ive been messing with baremetal audio on the rpi, just .incbin'ing a raw audio file (not even wav headers) into my binary, got things working well, but my bootloader is limited to a 12mb binary

#

need to try and embed a codec into things

bold vault
#

Has anyone did fft with a pdm mic. Most fft code crashes my board or doesn’t work.

#

I got the ardino fft library to work I’ve it doesn’t do it very good

desert fog
#

FFT is pretty tough to do on most microcontrollers.

bold vault
#

Yea I can see. Do you have any recommendations?

#

I saw the circuit playground did it but I probably have to reverse engineer the library

small niche
#

You can try this library. It's meant for Arduino boards but the underlying code is board-agnostic as far as I can see. They're just calculations. All you really need to do to is to supply the proper samples either through DMA or whatnot.

stable iris
#

is it true that audio equipment still uses a resistor / caps to produce sound or a more expensive setup with transistors to amps the voltage between each stages ?

#

ie: "Some amplifiers are capacitively coupled (there’s a capacitor between each stage), while others are directly or resistively coupled. The direct
and resistive styles are also called DC coupling because the voltages on one stage get passed to the next. It’s a tougher type of circuit to design, but it results
in superior sound. So, most good audio gear works that way, and I expected to see that kind of circuitry here."

glacial spruce
#

There's a third option as well: transformer coupling. However, the "superior sound" claim is fairly bogus, it's possible to get top-quality sound with any of those coupling methods. The main problems with DC coupling are low frequency problems ("motorboating") and bias issues.

desert fog
bold vault
bold vault
#

I’ll try it again tonight, thanks though

glacial spruce
desert fog
glacial spruce
#

In the case of differential signals, there are three different kinds of DC bias. There's asymmetrical bias, which works similarly to DC bias in a single-ended circuit, common mode bias, which tends to be cancelled out, and symmetrical bias, which has effects similar to asymmetrical bias. For interstage coupling, this can affect operating points, but is usually only a big deal when there's a large amount of amplification involved, as the bias gets amplified along with the signal. When driving a speaker, common mode bias has no effect, the speaker just floats at a different level. However, asymmetrical bias will push a DC voltage through the speaker, offsetting the speaker cone, generating heat, etc. Some speakers can be damaged by a DC bias.

unborn drum
#

I'm making a synth using the new synthio library. I'm also recording samples using analogbufio and RawSample. both are working independantly.
However, I'd like to play both the synth and the sample at the same time, so i made a mixer object.
The issue is that synthio seems to give the mixer signed values while analogbufio can only fill an unsigned buffer ("H") . So i get "ValueError: The sample's signedness does not match the mixer's" either way i set the mixer's signed_samples argument.

I don't think i have enough memory to convert the analogbufio array to signed after the fact. Am I out of luck?

random bone
#

@wispy mortar you may have a comment ^^

wispy mortar
#

There is a way to change the buffer from analogbufio into 'h' without using extra storage though it's a bit underhanded. ```py

a = array.array('H', [0, 32768, 65535])
u = ulab.numpy.frombuffer(a, dtype=ulab.numpy.int16)
u += 16384
u += 16384
u
array([-32768, 0, 32767], dtype=int16)
a
array('H', [32768, 0, 32767])

You can create a ulab array that views onto the same memory as the array but with a different type. The pair of add operations have the net effect of shifting the array values by 32768, which turns the smallest possible value of 'H' (0) into the smallest possible value of 'h' (-32768), and so on.
#

I did not verify for certain that the u += 16384 works without allocating intermediate storage, though.

#

for some reason, directly adding -32768 (to do the manipulation in a single operation) is not accepted.

#

I think adding format conversion to audiomixer would potentially be an acceptable feature if someone contributed it; there's nothing mathematically hard or processing intensive about converting all samples to 16-bit signed inside audiomixer, but it does increase code size.

echo hedge
#

We have memoryview.cast() as well.

unborn drum
stable iris
#

About this: "The direct and resistive styles are also called DC coupling because the voltages on one stage get passed to the next. It’s a tougher type of circuit to design" Why boost the voltage in several stages with transistors instead of say doing 3.3V=>600V in one step ?

#

Instead of says microvolts=>3.3V=>5V=>24V=>96V etc ? Is the signal easier to clean/amplify/condition this way ?

desert fog
#

A lot of it has to do with frequency response (Miller effect for example). Also, any noise (both DC and AC) will be amplified along with the signal unless you take steps to deal with it.

#

Then of course there's the physical and electrical properties of semiconductors.

glacial spruce
#

You can easily boost digital signals more, or analog signals if you're not particular about the waveform (that's how class C amplifiers work)

unborn drum
#

are there any clever techniques for playing two audio samples with different sample rates at the same time? I thought i might make a separate audiopwm pin, but i ran into RuntimeError: All timers in use

#

my use case is playing samples over a beat. I want to be able to change the playback speed of the samples. Easy enough to do by changing their sample rate, but that doesn't work inside a mixer that's already playing something

glacial spruce
#

Arduino, CircuitPython, or something else? Playing to a PWM, DAC, I2S, or other kind of output? In general, I'll usually see a resampling technique to bring the data streams to a common sampling frequency, but there can be exceptions, depending on the environment and fidelity required.

unborn drum
severe bison
#

Pico et al sound output

autumn elm
#

Does anybody have a project I could look at that is doing decibel conversion in circuit python from a sound level sensor? I'm using a KY-037 or KY-038 (not sure exactly which one, but it looks like those) for my sensor, but I can't seem to get any results that seem remotely right. If anyone has some sample code I could look at (and rip-off) I'd really appreciate it. It's my first time posting, so if this question should be asked elsewhere please let me know. Thanks in advance!

harsh gust
#

Amp circuit 1 https://www.electronics-lab.com/project/piezo-element-amplifier-for-guitar-and-violin/
Amp circuit 2 http://www.richardmudhar.com/piezo-contact-microphone-hi-z-amplifier-low-noise-version/

How can I tell which one amplifies more?
Circuit 1 says it can amplify up to +20dB but I cannot see how much Circuit 2 can amplify.

Electronics-Lab.com
Piezo Element Amplifier for Guitar and Violin - Electronics-Lab.com

The board presented here is a preamplifier for contact microphones that are used for amplifying the...

Richard Mudhar
richard
Piezo contact microphone hi-Z amplifier – low noise version

This is the purist approach solution to the tinny contact mic sound problem if you want the lowest electrical noise, or the lowest distortion at higher levels. Piezo mics aren’t usually considered …

glacial spruce
harsh gust
#

ohhh

harsh gust
# glacial spruce They're basically both 10X gain as-is, but can be adjusted for more. However, t...

I have worked on the second one (Richard's circuit) since last year and haven't been successful adjusting the gain boost (since there is no potentiometer knob in the original circuit)

I tried your idea first, it was not working well as I boost the gain, it sounded tinny and distorted both in the breadboard and perfboard.

So I asked Richard directly on that post, he commented that your approach is not recommended so I tried what he suggested, and the overall volume decreased, not increased.

I don't know if you remember my project, it's called Anthesizer, hearing an ant colony by installing piezo elements into their nests.

glacial spruce
#

I remember the project.

#

I'm curious why he doesn't recommend a potentiometer. Does he prefer fixed gain, or just doesn't want more than 10X gain?

#

"Tinny" usually refers to lack of low frequency response, there are ways to deal with that.

#

"Distorted" is more complicated, but often happens when you run into clipping.

harsh gust
harsh gust
#

this was his response to my answer

glacial spruce
#

Ah yes, that response. That doesn't make engineering sense to me, so either Richard or I (or both) are missing some important details.

harsh gust
#

hmm it could have been easier if I can summon you in my place but I can only wish

harsh gust
glacial spruce
#

You realize that if you told a sound engineer something was seriously messed up, she'd come right back with "messed up HOW? Crossover distortion? Intermodulation distortion? Phase distortion? Equalization problems? Hum pickup? Clipping? Asymmetry? WHAT?"

harsh gust
#

right

glacial spruce
#

Her next question would usually be "where in the signal chain does the problem show up?" And that question is pertinent to your problem too.

harsh gust
#

or would this be a better choice than amp circuit to boost the signal from piezo

glacial spruce
#

No, that circuit is designed to generate clipping distortion (by using diodes in the feedback loop of an op-amp). It would likely do the opposite of what you want.

harsh gust
#

aha, I got you.

glacial spruce
#

Learning the basics is slow going, but it will pay off

harsh gust
#

I am thinking of using Daisy Seed to implement MaxMSP gen~ patch

Screenshot_20230613_075148_OneNote.jpg
#

but this time I really need that amp circuit to function properly to achieve my dream lol

harsh gust
#

I am thinking of using Daisy Seed to implement MaxMSP gen~ patch

glacial spruce
#

I don't know what any of that is

harsh gust
#

For my previous exhibition, I had to just bump up the gain on my laptop without using the amplifier. But my goal is to make my exhibition independent from laptop

#

so my goal is to create "piezo-piezo amp-daisy seed-speaker amp- speaker" chain for each nest

#

and for my next exhibition, I am planning to use 4 nests combined all together like that

glacial spruce
#

You may need two stages of amplification between the piezo and the Daisy, but that should be doable if required.

harsh gust
stable igloo
# harsh gust what do you mean by two stages of amplification? does that mean it requires two ...

As someone with some audio background, I can answer that. Think of your piezo mic like an electric guitar with active pickups or electret condenser (ribbon mic would also be a good example). You need an initial mic amplifier to get the tiny signal up to something that conventional preamps can deal with. If the signal is not hot enough for the Daisy, you need something similar to a mic preamp AFTER the piezo amp to get up to line level.

harsh gust
#

so daisy's audio input is not gonna be the same as my apogee duet input

stable igloo
#

Daisy is equivalent to your laptop, in your use case, correct?

harsh gust
#

yes I guess so!

#

ohhh I am such a dumb perso n

#

I don't know why I have assumed Daisy's audio input would have all that audio to line level amplification thingy

#

So I might need sth like this right?

#

https://www.amazon.com/DC1-8-12V-TDA2822M-Amplifier-Channels-Receiver/dp/B07DB46Z8B/ref=asc_df_B07DB46Z8B/?tag=hyprod-20&linkCode=df0&hvadid=416723174073&hvpos=&hvnetw=g&hvrand=16446595650882164969&hvpone=&hvptwo=&hvqmt=&hvdev=c&hvdvcmdl=&hvlocint=&hvlocphy=9031179&hvtargid=pla-875835833490&psc=1&tag=&ref=&adgrpid=100759323984&hvpone=&hvptwo=&hvadid=416723174073&hvpos=&hvnetw=g&hvrand=16446595650882164969&hvqmt=&hvdev=c&hvdvcmdl=&hvlocint=&hvlocphy=9031179&hvtargid=pla-875835833490

DC1.8-12V TDA2822M Audio Amplifier Board Module 2 Channels Stereo 3...

Features: 1, Small size, high power, no noise, portable. 2, Volume control potentiometer is convenient for user to adjust volume. 3, Wiring terminals allow to wire with only a screwdriver, not soldering iron. 4, Using high quality IC board, which helps a better performance of wiring. 5, Simple ci...

stable igloo
#

Think of the piezo amp as a part of your piezo mic, rather than a separate component in your signal chain. Like electrets, piezo mics and pickups tend to require amplification before they can be treated as a "regular" audio signal.

harsh gust
#

right

stable igloo
harsh gust
#

but in my previous exp, when I hooked up the piezo directly to my audio interface input, or to the random cheap guitar amp, it was pretty loud enough.

#

and when I was using piezo amp circuit I built, the gain was significantly lower, as if I am using an attenuator instead of amplifier.

#

ofc, there must be sth wrong with the amp circuit I built

stable igloo
#

Piezo amps, in my experience, tend not to be terribly high gain. Their job is usually to give usable signal.

harsh gust
#

Maybe I might need that preamp and poweramp both. that would be really helpful for my use case.

#

I am picking up ants' sound with the piezo element

stable igloo
#

So, they likely compensate for the frequency response curve characteristics of the piezo mic/pickup.

#

Yeah. That sounds like a good course of action. You could probably play with a cheap LM386 board or the like after the piezo amp and get good enough signal.

#

Also, on the side of identifying "tinnyness" in your signal, you likely have a frequency analyzer tool in your DAW. That is extremely useful there as it will show if you're clipping at a particular frequency and give you a direction to go down for troubleshooting.

harsh gust
#

when the pot is in the lowest level, the amp was working normal just as the original circuit. But when I turn it up a tiny bit, it started going insane like that

#

I thought I def did something wrong I tried in different breadboard, different power supply, changed the opamp to new one, etc

#

but it was not working properly in my case.

#

Ok I googled a lil bit

#

Would this be a good try?

#

Using this one as a piezo amp
https://www.amazon.com/dp/B00LNACGTY/ref=sspa_dk_detail_0?psc=1&pd_rd_i=B00LNACGTY&pd_rd_w=u0X56&content-id=amzn1.sym.f734d1a2-0bf9-4a26-ad34-2e1b969a5a75&pf_rd_p=f734d1a2-0bf9-4a26-ad34-2e1b969a5a75&pf_rd_r=HGAACKT5NDZHPPQB7Y9J&pd_rd_wg=OUp99&pd_rd_r=ec6f2295-88d1-421f-931a-7c052800e7d0&s=electronics&sp_csd=d2lkZ2V0TmFtZT1zcF9kZXRhaWw

HiLetgo 5pcs LM386 Mono Audio Amplifier Module 200 Times AMP Solo 5...

HiLetgo 5pcs LM386 Mono Audio Amplifier Module 200 Times AMP Solo 5V-12V 10K Adjustable

#

and using this one as mic level-to-line level amp
https://www.amazon.com/Universal-Mono-Pre-Amplifier-Kit-Intermediate/dp/B0002Z82XU/ref=asc_df_B0002Z82XU/?tag=hyprod-20&linkCode=df0&hvadid=644617763239&hvpos=&hvnetw=g&hvrand=6603306314332749532&hvpone=&hvptwo=&hvqmt=&hvdev=c&hvdvcmdl=&hvlocint=&hvlocphy=9031179&hvtargid=pla-430966926371&psc=1&gclid=CjwKCAjwp6CkBhB_EiwAlQVyxVxRgwgjcppP4YQH6vK75AEDE35N_t3vn6TmmoaEQmmtEb4sWz8TFhoCPhwQAvD_BwE

Velleman K1803 Universal Mono Pre-Amplifier Kit, Multi-Colour

Velleman kit was developed as pre-module for a number of audio applications where the input signal is too weak. APPLICATIONS as a microphone pre-amplifier or for level correction. Power supply: 10-30vdc/10mA. Cable length: 50 cm.

#

And then it goes to Daisy Seed
https://www.electro-smith.com/daisy/daisy

Electro-Smith
Daisy Seed — Electro-Smith

New batch coming at the end of June! Daisy is an embedded platform for music. It features everything you need for creating high fidelity audio hardware devices. Just plug in a USB cable and start making sound! Programming the Daisy is a breeze with support for a number of languages including C++,

#

And this for speakers
https://www.amazon.com/PAM8406-Amplifier-Digital-Channel-Amplify/dp/B08RDN58SZ/ref=sr_1_9?crid=SYNAV91ETAC7&keywords=speaker+amp+circuit&qid=1686683498&sprefix=speaker+amp+ci%2Caps%2C697&sr=8-9

2 Pieces Audio Amplifier Board Pam8406 Mini Amplifier Board 5v, 5w+...

Features: Amplifier type can be connected to select functions. Type D for high power, type AB for low power, factory default is high power. External shutdown can be added, SD: the default is high power level, and at low power level, chip will shut down. MUTE: low level mute, factory default high...

stable igloo
#

That seems like it should likely do the trick, yes.

harsh gust
#

yayy thank you @glacial spruce and thank you @stable igloo

#

thank you so muchh

median hemlock
#

FWIW Getting a clean signal from a microphone/transducer is a rabbit hole that turns into an all consuming black hole. Frequency response, gain staging, noise rejection, etc. If things are off and you can’t figure out why, find an oscilloscope and trace the signal through all the parts. I love the idea of “Ant Music”.

harsh gust
#

I totally agree 200%

#

I have spent quite a lot of money on components and such for this.

harsh gust
#

I will try with those for now I guess

stable iris
#

Is there a guide out there on what is legit and what is a scam in the audio field ? Might be a cynic but it all sound like scams nowadays. Especially when they pull out bandwidth as if singers sings 400x faster than in the 1980 and you need more bandwitch than about 200kb/s like on a rca cable

desert fog
#

The speed of the singing won't affect bandwidth.

#

Frequency content of the music will.

#

Also classical is some of the widest band music out there.

stable iris
#

but needing like 1000 mbps cables instead of 200kb/s like the old copper wires ?

#

I mean I assume the normal home user here not studios that must record on tape master with as much data as they can afford

#

What I often see sales people trying to sell which doesnt make sense to me is gold cables/power cord with a very high bandwidth at 20-100x the prices of normal rca cables

desert fog
#

That's just technology improvements.

desert fog
stable iris
#

or it looks like an RCA cables but with "shielding" all over, sleeves of different materials (probably usually for the beauty of the cable(? 🤣 )) etc

desert fog
#

If you need the bandwidth (for instance in a studio with 48 audio channels from various microphones). Even then you'll have dedicated equipment for it, and gold coated cable won't make any difference.

#

The power cable also doesn't make much of a difference, although sometimes adding a ferrite bead can help.

stable iris
#

but wouldnt that need to be measured ?

#

Sounds like premature optimization like another field says

desert fog
#

But they somehow have a lot of money, and marketing is more than happy to sell them placebos.

stable iris
#

but I can safely ignore what most salespeople says unless they actually heard the sound system and measured something right so it's not based on pure air / speculation ?

desert fog
#

Correct

#

Generally speaking outside of a studio 192kbps is more than enough.

stable iris
#

Im asking that because Im starting to consider getting 2 legit speakers that can connect to computer and other normal audio systems including ham/etc. But Im more familiar with studio stuff because I used to book for a friend and give them moral support when recording 🤣 But that studio stuff is way out of my price range and considering that if you get stuff like audiotechnika or sennheiser etc you probably pay half of it for the brand. So my approach would be more around finding what is inside / specs

desert fog
#

Not saying it isn't for the brand in this case, but a lot of the higher end companies are doing actual R&D work.

stable iris
#

but if you can aonly afford their entry level stuff do you still get these benfits ?

#

Im not an audiophile but Id often argue with my friend and his other friends about studio audio vs gaming speakers/headphones I saw nothing wrong and I remember hearing a world of difference between the two 🤣

desert fog
#

"Gaming" means that they dropped the quality into oblivion and added lots of rainbowBarf

stable iris
#

I mean just from seeing what is inside when it's not wireless (look like 1940 radio kits from popular electronics) I dont ever want to spend more than 30$ for them 🤣

#

In 90% of the case there isn't even a chip inside or anything related to improving the sound...

stable igloo
#

When it comes to headphones, sometimes you do get what you're paying for. For example, Beyerdynamic headphones are pretty pricy and do look on the inside like old radio kits but they are made in Germany and just about everything is replaceable by anyone who knows basic soldering and how to use a Phillips screwdriver. On top of that, they sound great and are incredibly comfortable.

Price-wise, a set of DT-990s cost about $160, which is about $8 in 1940.

Largely, if you're looking for good audio equipment, once you get to a baseline of quality, you hit a plateau of subjective characteristics, which are largely dependent upon your use case and potentially less clear cut stuff like preference and personal ethics (ex. I'm pro-labor, so, buy things that are ethically produced when I'm able to afford it).

stable iris
#

I meant stuff like most of the teardowns Ivee seen everyone uses the same small magnet no magtter if they arte 20$ or 200$ when tha magnet affect sound quality a lot in headphones. The board inside like about 2mmx2mm and seems to have no IC just directly connecting from the voltage line of the sound card without any filter/op-amp /dsp to make the sound better etc

#

the dt-990 pro I saw before seem to be quality stuff and I like the test points stuff (hurray for right to repair)

dfbfbf.jpg
desert fog
#

The test points are likely used during manufacturing.

stable iris
#

yeah but normally they hide them or epoxy over them

desert fog
#

Well if it's another manufacturing step it might not be worth the cost.

#

And they probably don't care that much.

stable igloo
#

Sometimes, simpler can be better. Though this is absolutely subjective. I prefer minimal processing, personally, so, I generally use a pretty flat pre and headphones that don't have any ICs that modify the signal. To me, they are a negative, when present. But, if you like your bass boosted, it's worth choosing something that sounds right.

The driver magnet type, size, and shape can indeed have a major impact on quality and color of the since produced. A good way to hear this in an exaggerated fashion is to listen to samples on the Jensen Loudspeakers website (guitar speakers). Neodymium magnets are very powerful for their size, so, they are typically smaller and lighter than AlNiCo for the same power. This influences how the speaker cover vibrates. Ceramic are pretty weak and inefficient, causing them to need to be much heavier for the same power. This also impacts the vibrations of the cone and gives a "vintage" color to the sound. AlNiCo are somewhere in between.

stable igloo
stable iris
#

yeah maybe I have outdated knowledghe

#

but saw a coupl of videos and they seem to say they even got help from the company to repair it (but might have changed)

stable igloo
#

Yeah. Their support is pretty good from my experience.

#

For your use case, I'd definitely look at the specs for what your sound card can output and use that as your target. No point in getting speakers that can handle a bajillion kHz sample rate digital signal if your machine is just able to output 48kHz or 96kHz. I'd be surprised if there were a perceptible difference in most audio past a 96kHz sample rate.

When looking at interconnects (RCA, phono, SPDIF, TOS link), it's, like most audio stuff, very subjective. If the audio is just to be listened to, like music, any of them will work as well as the rest. If for video, you want to try to meet or beat the transmission speed of the video. If for games, I'd probably go with TOS link, if it is an option. This is because, while the digital signal does need to be decoded to create the analog signal to move the speaker, it is transmitted at the speed of light, which can help reduce perceived latency when syncing it with the video output.

#

Basically, I'd look at the specs of your source and destination, try to find a good fit, and try to get a listen in person because what sounds "right" to you might not for someone else.

Also, any frequency response advertised outside of the 20Hz to 20kHz range is pure marketing.

desert fog
glacial spruce
#

Right: while the advertised frequency response may exist, there is argument whether that capability has any audible (or other sensory) effect.

stable igloo
glacial spruce
#

For good or ill, my hearing loss isn't "normal"

young pawn
#

The MAX98357 says "Output Power: 3.2W at 4Ω, 10% THD, 1.8W at 8Ω, 10% THD, with 5V supply". https://www.adafruit.com/product/3006
Now, some speakers only say "4 Ohm" or "8 Ohm".
But for other speakers you can get these diagrams showing Frequency Response (red) and Impedance (blue). I notice that on "X Ohm" speakers "X Ohm" is just the minimum. And it goes up to >30 Ohm in one spot. 30 is obviously way more than 8. I'm just curious, what effect does that "high-ohm spot" have? Does it even have any effect? I guess it is already "calculated in" into the "1,8W at 8 Ohm" spec of the MAX98357

frs8_8_fs.gif
Adafruit I2S 3W Class D Amplifier Breakout - MAX98357A

Listen to this good news - we now have an all in one digital audio amp breakout board that works incredibly well with the Raspberry Pi! If you're looking for an easy and low cost ...

glacial spruce
#

Yes, speakers have a "nominal" impedance, but it varies a lot due to mechanical resonances and the like.

#

Normally amplifiers have a low enough output impedance to keep control of the speaker, but it can have some effects on the sound.

young pawn
#

Thank you.
And directly affecting my project:
Just try the speaker out? Or is there a way to tell from the datasheet that for example a 30 ohm impedance at one frequency is definitely too high and will noticeably degrade the audio? I didn't find any such thing. So I think I will just try it out unless you stop me 😄

glacial spruce
#

Like I said, with a reasonable low-impedance amplifier (most are), it should be just fine. If you're building a custom cabinet, I'd avoid having a resonance at that frequency, as the amplifier will have less control over cone motion there, but otherwise it shouldn't be a problem unless you're going for extra sonic detail.

young pawn
#

Thank you 😄

solemn frost
#

Hello all, i am working with a walkie talkie using arduino and nrf24, i am able to transmit audio successfully but I want to amplify the audio output. so i was looking for an amplifier circuit. there are 2 terminals for the speaker. if i connect the output to + of my speaker then where should i connect the - of the speaker. If i connect - of the speaker to ground, i dont listen anything. I am taking input from arduino's digital pins 9&10. This is my 1st audio project. Please help me.

amp.png
glacial spruce
#

That op-amp, while it can amplify, doesn't have much drive capability. I'd suggest using a speaker amplifier (LM386, PAM8302, etc.) for speaker driving duties.

solemn frost
#

Okay

harsh gust
#

I have several questions..!!

#

1> This is a pre-built preamp circuit I bought from Amazon. Is this the right way to connnect components?

20230625_172416.jpg
#

2> This says Mono Amp but why does this have 2 inputs?

20230625_172443.jpg
#

3> What would be the best way to power these 4 components efficiently?

20230625_172619.jpg
glacial spruce
harsh gust
stable iris
#

why mono thought is stereo much much much more difficult ?

tranquil sinew
#

Stereo is almost double the work as you need two amplifier channels and not always needed

median obsidian
#

I have this old stylophone I built a year back
Lying in the cuboard

#

Not sure what the issue is
It's not working any ideas on how to test ?
Also how long to 9V BATTRIES last ?

glacial spruce
#

I'd reach for a voltmeter first and check some likely points. As for 9V battery life, there's shelf life, and life when plugged into something. For shelf life, it ranges from about a year (for the cheapest ones) to 10 years or so (for the fancy ones, or even more for the lithium ones, but those are uncommon). For a battery plugged into something, it depends on what the quiescent current draw is, but is generally fairly short, as 9V batteries simply do not have a lot of capacity.

median obsidian
#

OK ok got it working
I think the volume pot was having issues
And changed the 555 timer on the board
The batteries were showing 9.2V so I got worried on of the caps had leaked

median obsidian
cinder briar
#

and also because you're moving the stylus

#

moving across a conductive surface will sound scratchy

median obsidian
#

Also do you think it would be practical to add something like an arduino and have this interface with a computer ?
(Would also like to be able to power this with a USB connector instead of the current 9V battery set up

cinder briar
median obsidian
#

Would the arduino it self be able to do all this ?

cinder briar
#

Actually yes

#

Except for driving the speaker

#

But that just requires a simple amplifier circuit or module

glacial spruce
compact nexus
#

I think I have it set to full volume in the code

young pawn
#

Is that a MAXsomething I2S DAC+Amp?

compact nexus
#

Ack I forget which one. Lemme check my purchase history

young pawn
#

lol I'm basically building the same thing currently. Except I want to play random tracks form an SD card that are like seagulls, ocean waves, sea shanties 😆

#

Except "I'm building [...] currently" = "I haven't actually started with the actual code." because I first had to make other people fix all the bugs in the I2S library (and fix some myself). Then fix the arduino-audio-tools RP2040 driver, and now I think I need to improve the performance of volume control 😆

compact nexus
#

Adafruit Audio FX Sound Board + 2x2W Amp - WAV/OGG Trigger -16MB PID: 2217

young pawn
#

Huh, not sure if my experience then applies. Still sharing it because maybe it's still useful as a comparison. I'm using one of these MAX98357A I2S amps. It's 3W with 4Ohm speaker. I'm using a 6cm 4 Ohm speaker and at full volume its way too loud. Like way louder than how I normally watch videos or listen to music

glacial spruce
tall mango
#

What would be the best way to get a headphone-level audio signal from a feather? I noticed Adafruit no longer has the I2S headphone amp which I tried once and it was extremely loud (it might've been line level, but I'm not sure and can't find it now)

small niche
tall mango
#

Just found that the same time you posted it lol, do you know if it can be used just as a DAC? (As in the feather just passes I2S or other audio data to it and it spits out the analog signal of it without decoding the mp3/wav/ogg itself)

#

Actually nevermind, I think I'll just use it for decoding too

glacial spruce
weary gyro
tall mango
glacial spruce
#

You can always add an L-pad, but most line level signals don't have the low impedance to drive headphones very loud.

weary gyro
#

In the case of the PCM5102 boards I've played with, they are strong enough to drive headphones (but I've only tried it with earpods and sony mdr7506s)

tall mango
#

Ok, I'll consider getting one then

random bone
#

interestingly, Pimoroni used to sell a Pi accessory that used this chip, but then discontinued it and now have PCM5100 one. I'm not sure why. Maybe there was a parts shortage of the '5102.

weary gyro
#

Replying to myself: ah, the difference is in their dynamic range & noise. Not really an issue for most uses by us hackers

Screenshot_2023-07-01_at_12.53.21_PM.png
glacial spruce
#

To get that full range and low noise, you'd need to use careful filtering and shielding anyway. If you're not doing that, it's not going to matter.

gritty basin
#

What types of capacitors would you guys recommend for audio? I've read that PP generates less distortion then electrolytic, but others more generally recommend electrolytic, only thing I've seen people agree on is to stay away from mylar

#

Okay my options got chosen automatically, can only find electrolytics at 2200uf on tme

glacial spruce
#

Right, it depends on the value and use case. For large values, electrolytics are the way to go.

gritty basin
glacial spruce
#

It depends on the use case. For coupling capacitors, you want low leakage current and low distortion (I usually use film capacitors, but not always PP). For power supply filtering, you want low ESR and high ripple capacity, leakage isn't critical. For general purpose decoupling, it's not critical, I usually use ceramics in that position (but X7R or better dielectric).

desert fog
#

Mixing capacitor types and values can also be beneficial. Keep in mind that all capacitors behave like capacitors up to their self-resonance frequency, at which point the ESR and lead inductance will dominate.

oblique radish
#

how i can measure speaker impedance with arduino or any other module

glacial spruce
#

Most speakers are fairly low impedance, which is tricky with an Arduino. Also, do you just want the DC impedance, or are you looking for something fancier like an impedance versus frequency sweep? There are some Arduino based component analyzers out there with an open source design, you could look at how they work for ideas.

oblique radish
glacial spruce
# oblique radish you are right , impedance vs frequency sweep analysis

An Arduino by itself would be fairly limited in the sort of analog analysis it can do, but with some outboard hardware, it could make a dandy controller. I had been thinking you could start with a DDS sine wave generator like AD9833 and some output measuring equipment, but a little looking around showed there are some chips like AD5933 that integrate a lot of the functionality of an entire network analyzer (even better than just an impedence sweeper). A web search for "AD5933 network analyzer" yields some ready built modules for under US$50 that would make building something like this much easier. There's an Arduino library available that includes examples for frequency sweeping https://github.com/mjmeli/arduino-ad5933 and the wiring is fairly easy, as it's an I2C controlled chip.

drowsy parcel
#

can someone help me to fix something about adafruit breakout pdm mic? it
it is emerge
I am trying to convert buffer array to standard wav file to save voice with adafruit breakout pdm mic

#

i am using rp2040 with pdm mic

#

and also using circuitpython i do not know how to implement with c

glacial spruce
drowsy parcel
#

okey

#

i am doing it already but there is problem about sample rate. when i take raw sample values to wav, i am choosig 16k sample rate which i choose but the sound which conversed is so fast. I mean if i record audio for 10 second, after conversion i got 2 second audio.

#

if i explain all of my algorithm from beginning

#

this my circuit python code:

#

it is basically record samples for 1 seconds and it stores in a sd card

#

i am recording samples as .npy format for 1 second

#

after that i unplugged my sdcard from my circuit then plug my computer

#

i am taking all .npy files

#

when i try to convert it to wav file, there is a problem about sample rate

#

my sample rate is 16k but when choose it into soundfile library which on my computer

#

the sound is faster than i record

#

i mean if i record 10 seconds it would be 2 second after convert it

#

when i find sample rate with try and fail, it is approximately 7.3khz

#

how can it be possible

glacial spruce
#

It could be the settings of your recording from the PDM microphone, or a bandwidth issue

drowsy parcel
#

it is same with documantation

#

sample_buffer_size = 200
sample_rate = 16000
mic = audiobusio.PDMIn(board.GP3, board.GP2, sample_rate=sample_rate, bit_depth=16)

young pawn
#

maybe a stereo vs mono issue?

drowsy parcel
#

when i choose mono

#

the sound is exploding

#

i can choose only stereo

#

i can send you

#

with wave.open('sine_wave.wav', 'w') as wav_file:
samplerate = 16000
nchannels = 2
sampwidth = 2
framerate = len(data)
nframes = 'NONE'
comptype = 'not compressed'
wav_file.setparams((nchannels, sampwidth, samplerate, framerate, nframes, comptype))
wav_file.writeframes(data)

#

do not use headphones

#

Write data to a wav file

with wave.open('sine_wave.wav', 'w') as wav_file:
samplerate = 16000
nchannels = 1
sampwidth = 2
framerate = len(data)
nframes = 'NONE'
comptype = 'not compressed'
wav_file.setparams((nchannels, sampwidth, samplerate, framerate, nframes, comptype))
wav_file.writeframes(data)

#

when i change nchannel to 1

#

with wave.open('sine_wave.wav', 'w') as wav_file:
samplerate = 7350
nchannels = 1
sampwidth = 2
framerate = len(data)
nframes = 'NONE'
comptype = 'not compressed'
wav_file.setparams((nchannels, sampwidth, samplerate, framerate, nframes, comptype))
wav_file.writeframes(data)

#

when i choose sample rate to 7350

#

it is more clear

solemn flint
#

I don't know this library, but setting "framerate" to the length of the data looks suspicious at first glance.

drowsy parcel
#

okey i will be focused on it

#

i changed it to possible values like buffer size 200 or sample rate but result is samew

#

same*

solemn flint
#

Actually they may be correct, just named weird. You have nframes = 'NONE', but that's being passed in as the slot for the comptype parameter, so it makes sense for that, for instance. And the original len(data) would work for nframes, but you just had that parameter labeled as framerate instead, etc.

drowsy parcel
#

okey i will check parameters and i will tell you the result.

#

i am using 'wave' library at this time

#

btw

#

but just i wanted to highlight something. when i change the sample rate, the length of sound is changing also. sample rate should be number of samples for a time. So, when i use a sample rate which i used in microcontroller, it should be same. Because

#

because it is constant you know. I did not understand this topic. I think i had a mistake on my algortihm on raspberry pi pico software

#

am i wrong?

#

there is a problem on sample recording

#

but also there is thing that i should tell you, there is not lost on sound which i recorded. It is full at the begining until end

#

is there a auto decimation step when i record samples on my sdcard xd

#

i will return when i will check my parameters on wave library thanks

twin compass
#

Hello! Does anyone here know a way to connect the audio output from my Soundboks 3 speaker via a 3,5mm audio cable to my esp32? I would like to read the audio signal of what is currently playing through the speaker, but realized after just plugging the Tip Ring and Sleve directly into the esp32 that the signal was not what I expected. Haven’t quite figured out entirely what comes out from each of the connection points other than that the tip and ring is often right and left channel respectively, oscillating around 0V, so it goes to negative voltages which as I understand the ADC does not support.

tall mango
#

I just got a used 7" reel to reel player/recorder, and I'm trying to figure out ways to get rid of the large amounts of noise. I suspect the tape I'm using is a little older and therefore noisy due to being written to so much. I also wanted to get a demagnetizer to demagnetize my tape heads, but they're rather expensive for a tool I'll use probably only a few times. Is there a way I can DIY a demagnetizer without damaging my tape heads?

#

Here's an example, the first photo is a snippet of the audio I recorded onto the tape, and the second is the same section but recorded back onto my computer. There's also not very great fidelity in the mids and highs.

image.png image.png
#

I did test another tape by the way, and while it had less noise it was still somewhat present.

random bone
tall mango
#

I only have two rolls of tape, and they're pretty old (one of them I'm estimating was recorded in the 70s) but I plan to buy a new reel today.

random bone
#

sounds good

unborn drum
#

That’s really weird noise to see on the spectrogram. It’s so even and broadband. Makes me wonder if something else is going on in the signal path

tall mango
#

I also did some careful cleaning of the tape heads with some alcohol and qtips but it didn't make much of a difference.

tall mango
unborn drum
#

Could you upload audio clips?

tall mango
#

Yeah one sec

random bone
#

if you could generate sine wave or something similar and then record and play it back, that would also be informative

#

you can do that in Audacity

tall mango
#

Yeah I did that with the last tape deck I had but I will do it with this one too later

tall mango
#

Ill record the sine wave once the motors cool down (it doesnt have any cooling so I assume its normal??)

random bone
#

maybe record a long chirp too to check the frequency response

tall mango
#

Would a frequency sweep work better? I made a test signal that goes from 60Hz to 22050Hz over the course of ten seconds.

unborn drum
#

Interesting. Sounds like tape hiss alright. New tape may be the answer. How are your audio levels? If you have room, you could turn the level up, and the underlying hiss will be lower.

#

Also what speed are you at?

tall mango
#

The hiss gets louder with the volume level so I don't think its the amplifier or anything like that. Probably the tape or tape heads.

tall mango
unborn drum
tall mango
#

The hiss comes from the tape deck line out, which is what im recording from

random bone
#

if you are not playing, do you mean there's residual hiss?

tall mango
#

Theres a little hiss when its not playing, but it gets a lot louder when playing

random bone
tall mango
#

Also heres the results from the sine and frequency response test (it has some annoyingly high pitch sounds so be careful)

#

Seems like it struggles with higher frequencies but it can still do it

random bone
#

yeah, that just sounds like tape hiss to me. Does the tape deck have any noise reduction circuitry like Dolby?

#

are you recording at 3 3/4 or 7 1/2 ips?

tall mango
#

Im not sure, this ones old but I can check the maintenance booklet

tall mango
random bone
#

it would be on the panel; it would be things you can turn on and off

#

there would be Dolby button, etc.

tall mango
#

Oh, then no

random bone
#

what is the brand and model?

tall mango
#

EICO RP100

random bone
#

a lot of the noise reduction stuff was invented to make cassette tapes palatable at all: Dolby, fancy tape forumlations, etc.

#

EICO is not a high end brand. This is tubes, right? Tubes can get noisy. But frankly, this sounds like the tape decks I encountered in the 60's as a kid

tall mango
#

I think it does use tubes

random bone
#

you are not going to get CD quality out of this

tall mango
#

Ah, ok

#

And theres no improvements I can make to it to get it to sound better?

random bone
#

fancy reel-to-reel decks used for mastering at that time used higher tape speeds

#

tape hiss comes with the territory

#

you can record at as high a level as possible without flat-topping. The more signal the less hiss

#

browse around in https://audiokarma.org/ and < https://gearspace.com and read people's experiences. You can touch up the deck. But it's not going be what you're used to, no matter what. ... I've lived through this. CD's and digital audio are a big improvement over what came before, despite the current fascination with vinyl

tall mango
#

Unfortunately, I'm stuck with 7 1/2 ips. At least until I can get a reel to reel that's more reliable.

random bone
#

My father used to record the concerts of our local amateur orchestra. I would operate the tape deck. We also recorded TV programs off the air, etc. He had a reasonable deck. there was hiss

#

what is your goal, or is it just to fool around?

tall mango
#

At least it was like $10, the guy said it didn't work but it works just fine. Unless "works" means good quality.

tall mango
#

I've been familiar with vinyl and regular audiocassettes for a while but never really focused on tapes as much.

random bone
#

and not record it digitally and go back to analog?

unborn drum
#

Do the meters on the front of the recorder work?

tall mango
#

I don't think so

unborn drum
#

How are you setting the level?

tall mango
#

I tuned it until the signal was just about the same as my original audio clip but through the DAC

unborn drum
#

“It” being? A gain knob on the recorder?

So the setup is DAC-> Tape and then Tape -> DAC?

random bone
tall mango
#

So I'm feeding a line level signal from a DAC connected to my computer into the line in port on the tape deck, then playing the recording back through my ADC connected to my laptop.

unborn drum
#

Can we see your machine? 🙂

tall mango
#

Sure

unborn drum
#

Looks lovely! Those knobs are something else, huh?

Are the mic levels down for both channels?

tall mango
#

Yes

unborn drum
#

And on the PLAY SEL knob, the eq is 7.5?

tall mango
#

Oh, no, it was set to TK1

#

The only tape I have is mono

#

The knob doesn't actually turn past TK2

unborn drum
#

Is that for the upper part of the knob or lower part?

tall mango
#

The upper part, the lower part doesn't move

unborn drum
#

Huh. Is it stuck? It’s label seems to indicate that the lower part selects the eq

tall mango
#

Oh I think the knob was misaligned

#

Nevermind yeah the knob is a little stuck

#

Should I add oil to it?

unborn drum
#

Eh, I can’t answer that one! If it were me, I’d keep trying to wiggle it. Also I was just looking at the meters, and their labels begin at -7db. That’s really pretty high. If your signal is low, you might not see them moving.

I still have this suspicion that the tape can handle more volume than you’re giving it. So, try turning up the laptop or turning up the LINE LEV knob and turning down the level on the ADC in. See if you can get those meters to move at all before the tape distortion becomes audible.

If you’re recording as hot as possible and still have that hiss, it may be inherent to the machine.

tall mango
#

I turned it too much and the indents on the knob that turn the potentiometer actually just broke... oops

unborn drum
#

Drat!

tall mango
#

I might be able to find some pliers to loosen the potentiometer to the point where it isnt as stuck anymore

random bone
#

which knob are we talking about?

tall mango
#

play sel

random bone
#

i think the the upper knob has only four positions. The lower knob is the EQ selection on the right, at 1 o'clock and 2 o'clock ?

#

(I looked for a manul for this but they are all for-pay downloads)

tall mango
#

Oh, i think youre right

random bone
#

is the EQ part of PLAY SEL (the lower knob) a two position switch or a potentiometer?

tall mango
#

I tried maxing out my volume on everything and it did some weird stuff to the frequencies

image.png
#

I do notice less noise though

random bone
#

the way to set the level is to feed the signal and then adjust the LINE LEV knobs to keep the level just below something (maybe 0db here??)

#

did you see the meters move at the higher level?

#

I can't tell if there are red markings on the meter, but below that

unborn drum
#

The traditional calibration level is -20db in digital to 0db analog

tall mango
#

Oh, I forgot to check the meters

unborn drum
#

Assuming they work 😉

tall mango
#

Yeah, the meters won't budge

random bone
#

if you adjust the LINE LEV controls does it make any difference? They're not moving at all?

tall mango
#

Yeah, even all the way up

random bone
#

are the meters completely immobile? I've had very old meters get stuck: sometimes they are dirty. Unfortunately I cannot find any schematics for this to theorize on what might be going on. If you have the equipment you could measure voltage at the meter and see if it's varying.

#

but having both not move is suspicious...

tall mango
#

I would, but sadly my multimeter battery is toast and I don't have any extra 9Vs

#

Also I just looked at the back of the machine, theres bias adjust and meter adjust knobs. Maybe that's the issue?

unborn drum
#

It’s also not clear what those meters measure. Signal going to record head? Signal coming from play head?

unborn drum
#

Bias seems fine based on your screenshot

tall mango
#

No difference :(

#

I turned the knob all the way clockwise and counter-clockwise and it stayed at the bottom

#

Maybe whoever assembled this missed something

random bone
#

take a picture or mark the current settings of the meter adjust knobs 🙂

tall mango
#

I set it back to what it was before

unborn drum
#

Bummer. I agree with danh that they could be stuck.

Also like danh i’m old enough to remember tape, and I think this level of hiss is within the realm of normal for pre-Dolby home recorders

glacial spruce
#

For physical reasons involving tape speed, particle size, head gap, etc., there's a roll-off of higher frequencies, along with some intrinsic hiss. Some equalization schemes try to recover the higher frequencies, but tend to boost the hiss (which is mostly higher frequencies) in the process. The Dolby ones use a pre-emphasis/de-emphasis technique to minimize this, and can work pretty well, but only when things are properly adjusted.

random bone
# tall mango I set it back to what it was before

i don't know if the meter covers would snap off, but I would be very careful. A service manual would be very helpful. You might ask in the forums I posted above if anyone can give guidance. It's a project and a learning experience! And if this one doesn't work out, another one might.

tall mango
#

It actually came with the instruction booklet and assembly booklet

random bone
#

the schematic may have voltage readings at various points

tall mango
#

Yeah, but I have no way to read them right now sadly

random bone
tall mango
#

Unless I can find that other multimeter

random bone
#

time to sleep on it, maybe. this is a multi-day or night project

#

same idea as video: lots of search results for "stuck vu meter"

glacial spruce
#

Thanks for those pointers, I have a stuck meter for which it would be difficult to source a replacement, so if I can unstick it, that would be helpful.

random bone
#

i have a dual-pointer SWR meter with one sluggish pointer. this may give me some ideas

glacial spruce
#

Oh yeah, those dual meters are presumably speciality items too

random bone
#

fer sure: it's a quality Daiwa meter, but it's 40 yo

random bone
# glacial spruce Thanks for those pointers, I have a stuck meter for which it would be difficult ...

i did turn the center "screw" on one of the movements on my Daiwa meter, and maybe it loosened it a bit. But equally important was that there was a static charge on the printed meter face under the plastic cover. Moving anything (tools, fingers) near the face affected both pointers. Blowing breath on it to add moisture did not help, but a a wipe with a dryer sheet cleared up the charge completely. I also wiped and cleaned the outer protective face And this is with ambient humidity about 50%. Then tested with another SWR meter in line, and it's looking good.

glacial spruce
#

Nice! That static can really mess with sensitive meters.

#

If you didn't see it, you might enjoy this writeup of a baffling/amusing incident I had with an analog meter a while back #help-with-projects message

random bone
glacial spruce
#

Those are workhorses too.

tall mango
#

I'm looking for some tape to get for my reel to reel, and I don't want to spend too much on really nice mastering quality tape because it's not a very high end machine. Would it be okay to get some old stock tape from eBay? I mean from around the 1960-70s, or would it have degraded by now?

#

My mom used to work with tape at the radio station she worked at and she's worried the tape might be brittle and unusable by now.

oblique radish
desert fog
#

It's 4 ohms at DC, right? Speakers are mostly inductive, so by the time you reach into the kHz it likely has considerably higher impedance.

oblique radish
#

found this circuit, it is hard to found out lmc6482. can i use 2 lm358 for the circuit?

Screenshot_2023-07-11-07-27-48-261_com.android.chrome2.jpg
desert fog
#

Probably not, but I'd have to read through the datasheets, and I'm not going to do that from my phone.

oblique radish
oblique radish
oblique radish
#

specs of test speaker.

Screenshot_2023-07-11-07-35-52-466_com.whatsapp.w4b2.jpg
glacial spruce
#

In any case, you'll need some additional circuitry

desert fog
#

As madbodger said, not directly.

#

Also, impedance is not a single value. It's a frequency-dependent curve. When you see a single number describing impedance, that value only applies at a single frequency (or within a band of frequencies).

glacial spruce
#

Right, they want to know how to do an impedance vs frequency sweep

oblique radish
drowsy parcel
#

hey, can someone explain my how to get binary data from my PDM Mic with circuit python as shown below?

image.png
#

I am trying to make a PCM to PDM process from scratch

glacial spruce
#

Hopefully CircuitPython includes functions to do this sort of thing, as native Python isn't particularly good at bitwise manipulations itself. Maybe ask in #help-with-circuitpython ?

random bone
unborn drum
# tall mango My mom used to work with tape at the radio station she worked at and she's worri...

I think this often depends on storage conditions. My last job using my reel to reel was digitizing some ancient UNICEF radio tapes. They had been properly stored since the 50’s and 60’s and the tape was in fine shape.

You could also look into baking your tape. Like, in the oven: https://www.radioworld.com/industry/baking-magnetic-recording-tape#:~:text=Place the tape in a,to working with the tapes.

Radio World
Baking Magnetic Recording Tape

Baking Magnetic Recording Tape Jan 7, 2002 12:00 PM Over time, magnetic recording tape becomes unplayable because the binder used to adhere the magnetic material to the backing or a chemical added to the binder becomes unstable. Tapes in this condition will leave a gooey residue on the tape transport. This residue is comprised mostly […]

oblique radish
#

@glacial spruce @desert fog
used ad9833 for generating sinewave of 300Hz and sweep 20Hz - 20000Hz. can find impedance

desert fog
#

Nice!

drowsy parcel
#

will i take the binary datas?

glacial spruce
#

Presumably it would be binary data

desert fog
#

Those would just be signed 16 bit integers.

drowsy parcel
#

i understood but it does not look prety true, logically i mean. if i show you the values of which i take from PDMIn class: array('H', [33297, 33288, 33298, 33307, 33270, 33290, 33294, 33294, 33342, 33373, 33327, 33279, 33265, 33283, 33293, 33260, 33415, 33300, 33167, 33299]) and its just first 20 values. their mean aproximantely 32K when i convert it to wav file it gave really bad sound only

#

and i develop simple algorithm for that

#

but it is not also effective

#

if i tell you my algorithm: 1. take samples mean 2. subtract mean from all sample's value 3. map every sample array to 32K again 4. convert wav data

#

i mean by 3. step map every data between -32k to 32k

#

let's say before process if data looks like that: [33297, 33288, 33298, 33307, 33270, 33290, 33294, 33294, 33342, 33373, 33327, 33279, 33265, 33283, 33293, 33260, 33415, 33300, 33167, 33299] after process: [ 4540 2792 874 1172 2195 -1215 1726 -1343 191 1385 2067 -362
1385 -447 405 -1428 2792 -1726 -106 788]

#

after that process sound can be understandable but it is not clear at all

#

what do you prefer for me pls, i need so much help

random bone
drowsy parcel
#

what do you prefer for me, my job is record with pdm mic and save it to sd card as sound file

#

which path i should fallow

random bone
#

this is not something the CircuitPython is very good at

#

due to hardware limitations

#

the sample to record has to fit in RAM, and then there will be an interruption while you write it out

drowsy parcel
#

yes, i meet with this trouble. I just recorded samples as .npy file in ulab's numpy and i just limited with 1 second

#

for example, i record 10 seconds as 10 ".npy" files

random bone
#

we don't have I2SIn. If we did I would be recommending an I2S input mechanism

drowsy parcel
#

so, sir can i record sound's as 1 second file

#

it is also suit for me

#

i can merge them after process

#

just i need clear and understandable sound

#

https://www.youtube.com/watch?v=5lH-tQw0tlU i wanted to fallow these steps but i was confused how i can get the pdm data by PDMIn class

YouTube
STMicroelectronics
STM32 Microphone Audio Acquisition: Part 3, PDM to PCM Conversion

Find out more information: http://bit.ly/AN5027-AppNote
http://www.st.com/memsmics
http://www.st.com/STM32

The STM32 Microphone Audio Acquisition video series is intended to provide information, suggestions and practical solutions for interfacing with PDM digital MEMS microphones using STM32 microcontrollers.
Starting from the general concepts...

▶ Play video
#

because it does not look like true, logically i mean. just 33K values.

random bone
#

around 32k is the zero-crossing point, so you have to remove the DC bias to center it around zero. That is what you are doing already.

drowsy parcel
#

oh

#

okey

random bone
#

try recording silence vs a loud noise

drowsy parcel
#

okey i will try and observe

#

now i want to ask a last question. should i subtract samples with 32767?

random bone
drowsy parcel
#

to remnove the dc bias