#help-with-hw-design

1 messages · Page 30 of 1

woven bluff
#

I once went to a museum in Germany where they fire a at least 5-meter tall Marx generator in front of people. The EMP from that thing can easily fry audiences' cell phone.

#

even a 30cm Marx generator I made, 2000pF caps I think, fried my PC in the next room

tough matrix
#

what is a Marx generator?

distant raven
#

Developed by Erwin Otto Marx in the 1920s to generate high voltage electrical pulses

tough matrix
#

until now, I only knew one Marx...

woven bluff
#

only way to make a decent lighting

supple pollen
#

Lots of good ways to make lightning. A Marx generator is a clever voltage converter that works by charging up a bunch of capacitors in parallel and discharging them in series.

distant raven
tough matrix
#

Unlike most people here, I studied Das Kapital in college.
Or at least was supposed to :)

rough smelt
#

has anyone used the rpi pico before?

#

on the adafruit pinout it says pin 32 is gnd

#

but on this site it says the pin is ADC Gnd

#

i currently have the a pdm mic soldered in and the gnd wire is in pin 32

#

its not working and the only thing i could think of is it can be this. Is anyone familiar with it? Is there a difference between ADC Gnd and regular gnd?

distant raven
rough smelt
supple pollen
unique patio
#

AGND is still connected to ground, according to the schematic, and I verified with an ohmmeter. It's just likely to be less noisy.

rough smelt
rough smelt
#

I just don’t see why else this won’t be working

unique patio
rough smelt
#

I have it soldered in rn which is my mistake for not checking it first. Thought it would be fine looking at the pinout.

#

I’m just wondering is there a way I have to instruct it to use digital not analog for those pins

unique patio
#

the pins can be either digital, and 32 (GP27) can also be analog. ... Note that the pin number is NOT the same as the GPIO number.

#

pin 32 is GP27, pin 31 is GP26, etc.

rough smelt
unique patio
#

is the other board you are using also an RP2040 board?

rough smelt
#

Yep

#

Diagram I made. The lights work so I know it’s not the board

unique patio
#

that looks ok to me a first glance. Is it arduino or CircuitPython code?

#

on the other board are you using the same GP pins, in the same order?

rough smelt
rough smelt
unique patio
#

which GP pins are you using on the itsy bitsy

rough smelt
#

Here’s my diagram. The only change was I used analog 1 for the data instead on what’s it’s in now

microcontroller-diagram.png
#

A0 sorry

#

Not 5!

unique patio
rough smelt
unique patio
# rough smelt Yep

what library are you using for PDM on the RP2040? Our ZeroPDM library is not for RP2040 at all

#

and it's not for nRF either

rough smelt
#

The one that comes with the rpi

unique patio
#

what did you use for nRF?

rough smelt
#

The one that came with it

#

Just import pdm.h

#

They both have their own library

unique patio
#

please give me links to both libraries

rough smelt
#

That’s the itsy bitsy one

#

Yea

#

Rpi has one that’s also very similar

unique patio
#

so I need a link for that

#

is it Arduino's library or is it in the philhower core?

rough smelt
#

Wait no it is the same reference

rough smelt
#

This one

#

More specifically the pdm part

unique patio
# rough smelt Later

The rp2040 uses PIO, so the pins need to be consecutive, probably. It looks like they are in the right order, but I'm not sure. There's no error checking, though, that vets that, it apperrs. I don't know what to say other than that you might ask in the RPI forums if anyone has gotten it to work. You could also try reversing the data and clock pins, but as I said, at first glance I think they are in the right order. Maybe you can find some samples uses of that library in other RP2040 code to confirm that.

#

are you confident about your soldering skills?

rough smelt
ocean coyote
#

hello, reverse engineering this 3.56mhz nfc tag powered led circuit. Trying to understand why the capacitors were chosen so i can calculate what was used.

image.png
supple pollen
#

That drawing doesn't look quite right, with both parallel and series capacitors on the left

ocean coyote
#

one sec

#

ill get a better image

#

slightly different layout because its an older revision

#

the big red pads are where it attaches to the tag

#

i know the resistor, C2 C3 and C4 values but the other 2 caps seem to be in a range my meter cant test

#

i desoldered them to test ^

supple pollen
#

Ah, I see, C1 and C6 are resonating capacitors

#

They'll form a tank circuit with the inductance of the tag to resonate at the NFC frequency.

#

C2 is a coupling capacitor, I suspect its value isn't critical. The other two are filter capacitors, which also shouldn't be critical.

woven bluff
#

C2 looks like a DC blocker

#

C3 is rectifier tank, if too small the LED would flicker, 4.7uF should do, C4 is additional filter for ripples but seems unnecessary for LED

ocean coyote
#

So did I draw my circuit diagram correctly?

ocean coyote
#

C2 and c3 are 100nf caps and c4 is 1nf

ocean coyote
supple pollen
#

In a sense. If the frequency doesn't match, you'll get almost no voltage.

low anchor
#

If I were to plug in a TRRS breakout to a modern-ish smartphone, how would it likely detect that? Androids have system events for when the headphones are connected or disconnected, so I'm thinking I could wire a reed switch to the headphone jack and that way get an input from the switch.

The buttons on the cable could also be an option but I'm not sure those would work well for holding down. (The application is that I'm mounting an old smartphone to my door, and then by using the reed switch and a magnet on the deadbolt I can let the smartphone identify if the door is locked or not.)

low anchor
#

Some sources would suggest that there's physical switches inside the jack, but others make it seem like the connection been the ground & microphone pins is what makes it detect or not. What would be the easiest way to tell? I'd imagine if it's the latter I could use a TRRS cable and plug it in with nothing connected on the other end and see if that's detected

supple pollen
#

It depends on the phone. I suspect some have support for things like inline mute switches that you might be able to leverage.

latent jungle
#

plug it in with nothing connected on the other end and see if that's detected

That's what I would suggest. I would imagine most smart phones got rid of the mechnical detection method and instead relied on the mic detection method (to save cost).

frosty schooner
#

I've hooked up a 38KHz IR receiver to my mcu and the signal it generates is... upside down. Its output is supposed to be 0V when idle then rise to the supplied voltage when it receives data. I'm supplying 3.3V, but when it receives data the voltage drops from 0 to about -1.35V. If I disconnect the output pin from my mcu then it drops to around -3.3V.

Is there something in the datasheet I'm not getting? It says in the absolute maximum ratings that the output voltage VO is from -0.3 to VS + 0.3, but figure 1 shows VOH and VOL below VO, but they're not define anywhere. I have the resistor (330 Ohm) and capacitor (0.15 uF) as in the application circuit, but I get the same results with VS connected straight to pin 3 and GND straight to pin 2. Is anyone familiar with this receiver or a similar one that could help me out?

https://www.pololu.com/file/0J892/tsop384.pdf

supple pollen
#

I don't see how it could provide a negative voltage if it's not supplied with one. Something confusing is happening and I can only guess what. One guess is that something you think is zero volts isn't really (broken wire, bad connection, breadboard bus that isn't continuous, etc.). Another guess is that you have some sort of high frequency noise that is confusing the voltmeter.

frosty schooner
#

I have no idea. Here's the expected result I saw on a yt video, though idk exactly what model of receiver he used, vs my result. Here I isolated it so GND and VS are connected straight to the power supply and OUT is connected straight to the oscilloscope.

Screenshot_2024-05-29_204000.png ir_receiver_isolated.png
supple pollen
#

Ah, you have an oscilloscope, that's useful for eliminating high frequency noise. Try probing the ground pin of the receiver.

frosty schooner
#

When I first turn on the PSU ground drops to -1.5V then it steadily climbs back up to 0V, though idk if that's a result of how the oscilloscope works (I'm really new to this) because VS jumps to 1.6V then steadily goes down to 0V when I turn on the PSU.

#

GND and VS. I also have the ground lead on the probe connected to the GND binding post of the PSU.

SDS00030.png SDS00031.png
supple pollen
#

Something isn't right there. If the PSU ground is connected to the scope ground, there shouldn't be any voltage between them.

frosty schooner
#

Oh wait sorry I was calling two different things ground. The PSU's negative is connected to the IR receiver, and there's a third binding post GND that I assume is connected straight to earth ground, and that's the one my probe's ground lead is connected to. It looks like the PSU is just making the 3.3V difference between positive and negative by making negative about 1.5v lower than earth and positive 1.5v higher than it.

supple pollen
#

I'd suggest tying the PSU's earth ground to the negative output terminal.

distant raven
#

😅

supple pollen
#

In any case, I'd tie the scope probe's ground reference to the same thing as the IR receiver's zero volt lead.

frosty schooner
# supple pollen In any case, I'd tie the scope probe's ground reference to the same thing as the...

I was just thinking that. I've been wondering why there's always a wave in the voltage, I thought it had something to do with the microcontroller messing with it, but it just occurred to me when writing my last message that it must be how the PSU works. I just remembered hearing somewhere, probably in a keysight tutorial video, that the purpose of the ground lead is to reduce noise and it should be connected to earth ground and not the circuit. I started to think that it also serves as a reference for the probe's voltage like.... ground 🤯

Now the output is idle at 3.3V and drops to 0V when it sees a signal 🎉 So my software wasn't working cause I assumed it would be the other way around.

SDS00032.png
#

Thank you for the help

swift aspen
#

it's weirdly hard to find JST-SH 1.0mm male connectors pre-wired! Easy to find male/female wired-up connector pairs in every larger size, from 1.5mm on up. The 1.0mm ones seem to come only in SMD component form?

#

Just trying to connect up a wee tiny linear servo without hacking the JST-SH female connector it comes with...

distant raven
swift aspen
#

I'm gathering that. Still, the 1.5mm spacing version has wired male connectors...

distant raven
#

You could buy an extender and cut the female end off right at the connector

swift aspen
#

So the only good option is just to cut the connector off and try to crimp/solder/wire up a new one.

#

all the 2-connector wire bundles I've seen are F-to-F so far

distant raven
#

I wonder if they even make a cable plug for male at that size

#

Probably not

swift aspen
#

Probably best to just be more careful what linear micro-servos I buy going forward!

distant raven
#

Indeed 🙂

#

You could make a small board with header holes or pads on one side for the wire, and a male plug

#

Or depending on which pin count, polulu sells some premade ones

swift aspen
#

luckily I think I only got a couple of these micro-servos, out of curiosity, so in future perhaps I can find ones with 1.25mm or larger ones

distant raven
swift aspen
#

yeah, that's why I'm looking for a male one...

#

so I can plug it into something!

distant raven
#

You could just use a 3 pin jst sh to 2.54mm board and solder header pins to it

swift aspen
#

but since there's no hope of finding a prewired one, I'll just cut the connector off

distant raven
#

That’s an easy fix

#

Don’t cut it

swift aspen
#

well it's quicker than ordering something

#

these little spring-probes are fiddly at the best of times... good enough to confirm it works, but that's it

#

I see the data sheet specifically classes this as a wire-to-board connector

#

but the ZH series is available in both types. It's size discrimination.

distant raven
#

doesn't hurt to make a few and keep them around IMO

tough matrix
#

the problem with sh is that it is too small for through hole version, so they only make smd header - and this is hard to incorporate in wire-to-wire assembly without producing an adapter board as sketched by skerr

#

but I personally would cut off the sh female connector and just crimped either a standard 3pin "DuPont" female header, or a 3pin jst ph connector

frosty schooner
#

I have another issue following this morning's. When I connect the IR receiver's output back to the mcu, the output voltage only goes up to about 1.4V instead of 3.3V. Why might this be happening? I want to make the output voltage go from 0V to 3.3V, and I think the simplest solution would be to use a mosfet like in this schematic I found online.

I would connect the gate to the 1.4V from the receiver, the drain to 3.3V from the power rail, then the source to ground. A lot of common mosfets like the IRF510 say the gate threshold is 2-4V so I don't think those would work, but the 2N7000's datasheet lists 0.8-3V so that one would work for my application. Does this sound right or is there something else I'm missing? The fact that it inverts the signal is fine, that'd actually make it more intuitive.

mosfet-voltageshift.png
cursive sentinel
frosty schooner
#

Ohhh right. I'm struggling to find any mosfets with lower thresholds, the 2N7000 is the only one that comes close. I also just learned about BJT transistors and how by connecting the base to ground you can amplify the voltage, but I'm still trying to figure out how I'd use that for what I want to do.

cursive sentinel
#

BSS138 is what they use in things like the I2C level translator.

frosty schooner
#

One that one's datasheet the threshold is still 0.8-1.5 and the on-resistance shoots vertically right at around 2V on figure 4 think_pepe idk if that would work either

cursive sentinel
#

On my phone now, so I can't really look at the datasheets.

frosty schooner
#

Ideally I'd like the receiver's output voltage to not be so low in the first place. When it's not connected to the mcu it matches the input voltage at 3.3V. When the mcu uses gpio pins as an input they become high impedance but I thought that only affected current.

cursive sentinel
frosty schooner
#

yes, both at 3.3V

cursive sentinel
#

And you're just trying to level shift it to 5V?

#

If so, I'd recommend just using a 74AHCT1G125.

#

SOT-23-6 buffer that accepts TTL thresholds at the input and spits out 5V CMOS on the output.

frosty schooner
#

No I mean the receiver's output is 1.4V for some reason and I want to level shift it to 3.3V so the mcu can see it rise and fall since it's supposed to be a digital signal between 0V and 3.3V. I found the SN74LV1T125 which looks similar to the one you suggested, and it says I can translate up from 1.8V to 3.3V when VCC is 3.3V. Later in its datasheet it says the high-level input voltage must be at least 1.37V when VCC is from 3V to 3.3V so it miiiight work.

SDS00033.png
knotty tiger
supple pollen
#

BSH103 looks like it can conduct nearly half an amp at 1.4V on the gate. Still, I think there's some issue with the IR receiver, its wiring, or (as argonblue points out) the MCU it's connected to.

Screenshot_2024-05-30_at_5.23.47_PM.png
frosty schooner
frosty schooner
knotty tiger
supple pollen
#

Those SOT-23 packages are actually reasonably easy to solder to perfboard (or an adapter board), but I agree finding the root cause should come first.

inland jungle
#

BJT could work if it can source a bit of current

swift aspen
misty notch
#

Hello all,

well, I have a RPI PICO W project with a keyboard, a buzzer and other attached devices (a I2C display among others).
Unfortunately the keyboard is such a size that I run out of pins and I need an extra output pin for the key matrix to be ideal, so I'm thinking about sharing one of the key matrix's row pins, that are input pins, to drive the buzzer.

Here you can see a very simplified view of the setup

I know that the simple solution is using a I2C expander as MCP23017 but as I ideally want to keep it as simple as it could be, I've tried to share a pin and activate/deactivate the keypad scanning when appropiate via software.

It works almost good... because for some reason it looks like the buzzers put some current/voltage/whatever while the keypad is active into the pin so it is detected as a keypress.

I've made a simple test putting a LED between the pin and the buzzer (in this picture, LED goes from 24 to buzzer +) and it partially works, because it prevents the _buzzer-inducted-keypresses- but also the buzzer doesn't beep, despite the led is lit when it it supposed that the buzzer beeps. So I guess the LED causes a voltage drop too high for the buzzer to operate. But I'm not an electronics guy so:

Could be any solution for this? A diode? drive it through a transistor? octocouplers? whatever?

Thank you very much in advance

image.png
supple pollen
misty notch
#

Thanks, I'll try what you say. However, the diagram is, as I said, very simplified, in fact all the GPIO pins are occupied: 23 for the keyboard matrix, 2 for a i2C display and one for the buzzer. That's why I need and extra pin because my keyboard matrix should be ideally 12x12 so I miss one pin, so no ADC approach is possible. Your second approach looks very nice. I'll give a try and come back

Thanks a lot!!!

frosty schooner
#

I found solution online to yesterday's issue (and drew my first schematic to demonstrate it 🎉). The receiver's output is still 1.4V, that I couldn't figure out how to fix, but I added a pullup resistor and diode to the circuit. Now the voltage goes between 2.7V and 0.6V, not perfect but it's enough to cross the MCU's digital logic threshold.

IR_Receiver_Schematic.png
knotty tiger
frosty schooner
knotty tiger
#

but if set to floating digital input, it shouldn’t sink enough current to pull it down that low (typical impedance is in the megohm range)

frosty schooner
knotty tiger
frosty schooner
# knotty tiger yeah, try setting it to floating input or internal pullup. you might not need th...

🤯 facepalm I just disabled both resistors and now it works fine without the external resistor. Now I'm still wondering why the pulldown resistor cuts the voltage basically in half outside the mcu. Isn't it essentially in parallel with whatever TTL inside the mcu the pin is connected to so it shouldn't be dropping any voltage? (I have no idea what's going on here, the idea of pullup and pulldown resistors is still confusing to me)

supple pollen
#

A pulldown resistor will "pull down" the voltage whether it's inside the MCU or not.

knotty tiger
frosty schooner
#

ohhhhhhhh I see. Now I have to learn how voltage dividers work, thank you.

supple pollen
#

It's a basic concept that underlies a surprising amount of circuitry, well worth learning.

swift aspen
#

(I know I can have them made online, but instead I think I'll just buy other micro-servos going forward!)

limpid hinge
#

I'm playing around with the Adafruit ADS1015 daughtercard interfacing to my RASPI. All I am trying to do is test sampling voltages from 0V to +3.5V using a 10K potentiometer as a voltage divider for the input.

After failing to get an accurate reading going directly from the potentiometer to the A0 pin on the ADS1015, I tried to stick a LM258N voltage follower buffer in front of A0 in case the output impedence of my circuit was too high. The current state of my test circuit is shown in the diagram below. However, even with the buffer in the path, I'm still not able to get able to get an accurate reading from the ADS1015 at a +3.5V input as measured with my multimeter. Instead, I'm only getting +2.8V with the ADS1015 reading (see voltage measurements shown in the diagram).

I'm thinking I'm missing something very basic which is not surprising given how new I am to all of this stuff. Any pointers are appreciated.

image0.jpg
limpid hinge
#

Got a tip from someone and figured it out. I had the MUX register misconfigured on the ADS1015.

distant raven
limpid nest
#

Trying to understand the limitations of TWAI and how I can have CAN FD with an esp 32 s3. It seems I need a spi or i2c transceiver?

supple pollen
limpid nest
#

OK I am working on an open source design. If someone wants to rejig it to have FD they CAN (see what I did there?)

supple pollen
#

Ah, if it's a design you're going to release, you have no knowledge nor control of what it might be connected to, so if I were in your place, I'd just add some verbiage explaining it. In other words, if someone hooks it to a "classic CAN" system, it should work fine, but if they hook it to a CAN-FD 2.0 system, it presumably wouldn't.

#

I'm tinkering with CAN myself, but using different hardware (my goal is to reprogram the ESCs in my electric tractor to not go into "turtle" mode when the battery voltage is higher).

low anchor
#

I have an existing PWM fan rated for 12v at ~0.1A that I want to run at different speeds - is that something I could do using a PWM output from a raspberry pi GPIO pin? What kind of circuitry would be best?

stark arch
#

is better to have crystals directly above a power (3.3v) or gnd plane? or does it not matter too much?

supple pollen
#

It doesn't matter a lot, as both have low impedance.

stark arch
#

would the same go for buck converters?

supple pollen
#

Buck converters are trickier, due to the high current pulses they produce in operation, generally the layout matters in several ways

#

I don't know much about PWM fans, but I'm guessing they have a PWM input to control the speed, so presumably you could connect that to a PWM output. One thing I don't know (and may matter) is the frequency the fan expects.

low anchor
#

Ah sorry, I misspoke. It's just a DC fan

stark arch
supple pollen
#

I'd normally aim for the ground plane as a) it's generally slightly lower impedance, providing better dissipation of noise, and b) things connected to it are often somewhat less sensitive to noise than noise on a power plane.

stark arch
#

makes sense, thanks!

low anchor
#

I took an ec en class a long time ago but I don't remember much

supple pollen
# low anchor I think I'd need a MOSFET to drive a DC fan - would this one work https://www.vi...

You can use a MOSFET or bipolar transistor. The main parameters are the current (you need enough current capability to support the fan's current draw), voltage (whatever the fan supply voltage is, plus a little headroom), and either the current gain (if you're using a bipolar transistor) or the required gate voltage (if you're using a MOSFET: the IRF540 you listed requires more gate voltage than a Pi will provide).

tough matrix
#

is there such a thing as antistatic bubble wrap, for packaging diy electronics such as PCBs with components on them?
There are some pink bubble wraps advertised as "antistatic" on Amazon, but I am not sure if it can be trusted...

distant raven
#

The pink bubble bags I send your logic kits in was anti-static

limpid nest
#

Is rj45 acceptable for a CAN board?

distant raven
#

Probably not without a little planning

stark arch
#

I have a question about PCB Assembly. I am attempting to get a BGA assembled, but on the viewer before I assemble it it shows the white dot in the bottom left, but theoretically it should be in the top left (see datasheet image, A-1 should be in top left). However, in the viewer it shows the dot in the bottom left, even with 0 rotation. Should I trust the viewer and just rotate it 90 in the pick and place file so the dot appears in the right spot?

image.png image.png
tough matrix
#

is it JLCPCB?

distant raven
limber thunder
#

what should i do with shield and gnd?

image.png
#

this is basically testing right now so if they arent mandatory ill leave them be

distant raven
#

Ground connects to the ground of everything else. Shield can be left unconnected, connected to ground, or connected to ground through a ferrite bead.

limber thunder
#

thank you

#

what is this X for

image.png
tough matrix
limber thunder
#

ah ok

#

yeah ima have to do some research on this ddr chip anyways

#

stm doesnt really seem to document the pinmaps well, the only things that lined up were the A pins and these

image.png
tough matrix
#

what is the chip? do you have the datasheet?

limber thunder
#

yes

#

STM32MP157D

#

for example, this pin doesnt match anything and it's not really elaborated on

image.png
#

the DDR chip is IS43TR16640C-125JBL

#

im looking at the ddr chip datasheet and stuff makes a bit more sense now

#

they do a MUCH better job at elaborating on pin purpose but without much help from the stm datasheet i cant do stuff

tough matrix
#

haven't worked with DDR interfaces before, so can't help here - but hopefully @distant raven or others know more

stark arch
distant raven
#

Personally, I'd avoid anything with DDR if you aren't experienced with high speed signaling. I'd work on something with more forgiving timing/length matching/controlled impedance than DDR

distant raven
#

I've done other length matched signalling like MIPI, SGMII, RGMII

stark arch
inland jungle
#

how sensitive is length matching for USB2 D+ D- routing?

distant raven
#

USB2.0 is very forgiving

#

it also depends too what speed you're looking at. FS? you could route it wildly. HS probably need to be more careful.

#

but a lot of things these days still only support FS which is wild to me.

limber thunder
#

I just need to figure out how to wire it is all

distant raven
#

it has a DDR controller, I'm talking about avoiding routing high speed signals until you've got experience routing more forgiving high speed interfaces.

limber thunder
#

i just dont know how it would apply here

distant raven
#

all I'm saying is DDR is not very forgiving for routing. especially high speed ddr like ddr3 and newer.

#

plus it typically requires impedance controlled stackups for PCB, it usually can't be routed on anything less than 6L, 8L is really the minimum you should do.

#

you're talking about a PCB order that could be expensive.

#

not to mention if the ball pitch of the STM part you're using is finer than 0.8mm, you will likely need blind and buried via which are not cheap.

#

you could easily spend $500-$1000 on PCB alone. If you can avoid the DDR, do that.

#

If you're okay with that cost, sure fine. Make sure you take your time, get a professional to review it like through STM FAEs (Field Applications Engineer).

limber thunder
distant raven
#

is there a reason you have to use this particular chip?

limber thunder
#

it has a little bit builtin but not enough to use the chip normally

limber thunder
distant raven
#

what's the application?

limber thunder
#

inbetween a microcontroller and an sbc

#

this chip runs linux

distant raven
#

is there a reason it needs to run linux?

limber thunder
#

if i want to use it for 3d printing stuff

#

it would really only be helpful with klipper

#

also it's just fun to play around with

#

i could just find the next step down that's more akin to a normal microcontroller chip

distant raven
#

I don't think you need a linux capable chip to support klipper. If it's just for processing GCODE, there are tons of fast chips that don't need DDR3 or linux

limber thunder
#

it would be running klipper in it's entirety, by that i mean the actual host thing you'd traditionally put on a laptop or dedicated sbc

distant raven
#

SAMD51 is supported

limber thunder
#

that's for the firmware not the actual server stuff

distant raven
#

doesn't octopi support klipper?

limber thunder
#

i'm using this to connect to a klipper machine not be the klipper machine

#

host the webui and whatnot

distant raven
#

are you translating klipper to marlin or something?

limber thunder
#

klipper is two parts:
the firmware that runs on the printer
and the software that connects to the printer which is traditionally run on a linux server
i want to do the latter in a small footprint

distant raven
#

doesn't octopi operate as a klipper server?

limber thunder
#

octopi runs on linux

distant raven
#

sure it runs on a raspberry pi

limber thunder
#

i dont understand where you're going with this

distant raven
#

I'm not seeing what you're trying to do

limber thunder
#

the firmware's job is to talk to the server with motor and hotend stuff

#

i want to be the server not the firmware

distant raven
#

right, I get that but why do you need something outside of octopi?

#

or any other firmware that can easily run on a pi

limber thunder
#

i dont follow at all

limber thunder
#

im doing this for fun lmao

distant raven
#

that's an expensive way to have fun

limber thunder
#

well that's how most of my hobbies end up

distant raven
#

Well, how about start with something not a raspberry pi like a orange pi or something

limber thunder
#

i'm trying to make a pcb not buy something

#

i have plenty of experience on that front, i am trying to design a pcb that does something similar for fun

distant raven
#

I mean, if you have a few thousand to throw around figuring it out; sure why not

limber thunder
#

don't understand how saying that is in any way helpful but alright

#

maybe i wont end up building it

#

no harm in at least designing the board

distant raven
#

i'm just being realistic here. it is doable, but it will be expensive to do

#

designing is fun, worth the experience

#

maybe look at the IMX RT1170 or RT1180 series

#

they're pretty nice, can run linux

#

pitch is fairly friendly

#

only needs like a 6L PCB which is fairly affordable <$100 for 5

#

They have support for things like LPDDR2, eMMC memory, SD cards, gigabit ethernet

limber thunder
#

i want to at least try and do this

#

ill save that in case i want something more realistic but i'm having fun

distant raven
#

be sure to join the 1bitsquared discord, they can help you or at least give you pointers on the HDI routing for DDR3

limber thunder
#

ty

#

which channel would this best fit in?

distant raven
#

eda-and-pcb I think

inland jungle
#

If I want to interface a 4-20ma sensor/transmitter to the MCU ADC, do I just measure across an inline 150 Ohm resistor and call it a day?

#

maybe if I feel extra motivated, pop a Zener on for voltage protection?

worldly schooner
low anchor
#

for hooking up power to a raspberry pi 4 via the GPIO headers, does it make sense to use some wire to solder to both of the 5V pins? 3A feels like a lot of current

cursive sentinel
#

It is, and that is the predominant design flaw that plagues both the 4 and 5.

hushed ermine
low anchor
#

Theoretically, could one of these be driven directly from a pi pico?

LED Charlieplexed Matrix - 9x16 LEDs - Blue

These are 9x16 Charlieplexed LEDs designed to match with the Adafruit 16x9 Charlieplexed PWM LED Matrix Driver - IS31FL3731. You don't get any resistors or driver chips with ...

supple pollen
# low anchor Theoretically, could [one of these](https://www.adafruit.com/product/2973) be dr...

The GPIOs can switch to high/low/hi-Z, so they can support Charlieplexing in general. The display is divided into two halves, each of which takes 9 GPIO pins to drive. You would need series resistors to limit the current as well. The problems are that the Pico is a 3.3V device and blue LEDs have nearly 3V forward voltage, the GPIO pins aren't capable of much current, and the Pico's power supply itself isn't capable of much current. It seems to me it could just barely work, but it would take some engineering, and wouldn't be very bright.

magic aurora
#

Hi, I'm looking to add the "NeoPixel Nano 2020 RGB LEDs - WS2812B" to a 3.3 volt system. would that work or are these also 5v?

low anchor
supple pollen
#

Many 3.3V systems get their power from some other voltage, which you could use to power your NeoPixels. Many AdaFruit boards do exactly this, powering the NeoPixels from a lithium cell.

knotty tiger
#

main problem with NeoPixels (and other LED devices that are nominally 5V supply) is that 3.3V logic might not be high enough voltage for the data to reliably transfer, if you’re giving it actual 5V power

#

if you want reliable communication, use logic level shifters

bright cape
#

Hello folks - having some fun with MOSFET's here. I'm using a ESP8266 Wemos D1 Mini. It runs from 3V3. I'm using the GPIO in open drain mode to either have HiZ or Pull to GND.
I'm aware there may be some leakage current as the HiZ is to 3.3V, rather than 5V.
VLED is 5V. D0 pushes either HiZ or GND.
I have 6 mosfets on the board. 3 of them are "locking ON" (don't switch off, regardless of gate voltage).
Is this common with MOSFETS?

image.png
#

I'd really like to avoid needing to add nmos drivers. 🙂

bright cape
#

VGS is +/-12V, so 5V is comfortably within range.

#

Current draw from the power LED's is about 200mA - maximum

#

ID limit of the transistor is 3A... so comfortably below the maxmum current

distant raven
bright cape
#

last thing I want is a floating gatE?

#

because the High signal of the ESP8266 only goes to 3.3V, 1.7V below VLED

distant raven
#

Should be fine

#

I suspect since you’re running the d1 and 3.3V and the LED to 5V, you’re doing some weird back powering of the pin which is messing with your ability to toggle the mosfet

#

8266 can handle 5V, but I’m unsure of how well taking 5V in when running of 3.3V is going to bode.

bright cape
#

i just checked the signals from the Wemos with a scope - they are still toggling (pulled up by mosfet pullup, then pulled to ground). (my code just toggles every 1000mS

#

so the wemos is playing just fine.

#

but the output of the PMOS is just "Shorted" ON

#

and only 3 of the 6 on the board.

#

they all start off working fine

distant raven
#

I suspect it’s just always going to see 5V. Though personally I would not put the LED at the drain side. I would put it it and its current limiting resistor on the high side. Then just tie drain to ground.

bright cape
#

the power LED's already have a current limiting resistor on them. so that's not a concern.

distant raven
#

No, it’s the side of the mosfet that your on coupled with pulling up to 5V.

bright cape
#

and to put the LED at the high side would mean I need to use an NMOS

distant raven
#

Remove the 100k pull-up and see, worst case is things don’t change

#

Best case is your LED toggles

bright cape
#

fair

bright cape
#

about to head into the lab (garage) to remove the resistor from one of the broken channels. now also wondering, as I PWM some patterns, I wonder if the lack of resistance between the Wemos and MOSFET Gate would cause excess current to flow in the gate capacitance of the mosfet?

distant raven
#

It could

bright cape
#

removing the resistor did not help.

surreal tangle
#

Yea, Vgs(th) is -0.7v (-0.4 to -1), so 3.3 is always going to be more than 1V less than 5V (ie 5V - 3.3 = 1.7 v). So it is always on...

#

Better to switch to a low side N-Mosfet. Easier to switch from 3.3V...

distant raven
#

So neither would work in the situation

surreal tangle
distant raven
#

Yeah

#

They removed it

#

But I agree, NMOS is easier to switch for LEDs

#

That has been my experience anyway

surreal tangle
bright cape
#

yeah - I'm progressively coming to terms with moving to NMOS.

#

end customers are so used to sharing the negative rail (Ground) around a system thoguh

#

My board is a Wemos D1 shield - tiny - so any extra components make routing even harder!

#

Prototype for reference

IMG_3202.jpg
#

pardon teh bodge wire -- copper fill of GND didn't flow close enough to the edges... I missed 3 ground connections as a consequence. Rushing to get to fab.... tut tut tut

strange ibex
#

Is it common for different individual current sensors to have different offsets? Specifically, I added an ACS71240 current sensor to my prototype board. It's a 3v3 30A version, and it's supposed to output 50% of 3.3v (1.65v) on the current level pin when at no load. But of my five prototype boards, all five have different no current readings, from around 1.45v to 1.55v. Is some amount of software calibration normal per chip? Or did I likely screw up the design somewhere?

#

as I increase current, I'm seeing the expected increase in voltage. It's just the starting "0 current" voltage is different on 5 different identical boards, and all of them lower than the expected 50%

cursive sentinel
#

I wouldn't count the "zero" reference to be particularly stable across chips, and it will likely vary with other external factors. Having dealt with it before, this is an unfortunate reality of the design of those chips. Had they provided some sort of reference output to establish the zero level externally, this wouldn't be as big of an issue. Or, according to the datasheet:

Screenshot_20240606-233811.png
#

Unfortunately, they only provide typical values, no min/max.

Typical values are mean ± 3 sigma values.

#

Translation: "you can't use this part to measure currents accurately". Though if you establish your own reference, you can use that.

supple pollen
regal lodge
#

What would be the best way to make it so that when programming a chip via an ICSP header the voltage supply is kept from powering the entire device? I was thinking a diode but then there's a voltage drop across VCC during normal operation :3

#

it's already on my BOM so think

knotty tiger
#

if you have enough GPIO and board space, make it so all other loads get explicitly switched on by load switch on firmware command

regal lodge
#

there are multiple chips that need programmed :3

jade wedge
#

I'm clearly overlooking something, or these potentiometers are not as they seem.
I have an old joystick that I'm rebuilding to be USB. It's got two 130k pots, that I've commoned together leaving the wipers separate. Metering across the ends, I see 65k, which is expected. Measuring from one end to one wiper I get non-linear values, which upon further reflection makes sense given that they're effectively resistors in parallel, and I'm only changing one of them.
Question though, once I pull one end of the pots high, and the other low, the wipers should work as proper voltage dividers without giving crazy values, right?

regal lodge
#

what joystick?

jade wedge
#

Like, I'm thinking that the only reason I'm seeing values like this is because the circuit is incomplete, and the test is taking unintended paths.

#

It's an ancient CH joystick. No active electronics.

supple pollen
jade wedge
#

Ok, good. I figured it was just bad testing, but always like to make sure.

#

Now I just need to install Arduino again...
Yay the joys of changing operating systems.

#

Side note, while installing Arduino I discovered SimulIDE.
This will definitely help answer some of my dumber circuit questions without bugging people.
Also experiment without risking exploding components.

#

Much to nobody's surprise, it works exactly as you say.

cursive sentinel
jade wedge
#

True true.
Though I'm always thinking I'm being a nuisance. Yes, it's dumb, yes, I need to stop it, no, I probably never will fully stop.

Also while it can't simulate the smoke, it can tell me if I'm using a mosfet vaguely correct. Something I'm still wary about.

cursive sentinel
#

That's... mostly true. Ultimately it depends on how good the simulation model is.

jade wedge
#

Don't worry, I won't be using this to design anything advanced. I'm mostly looking for the basics. Like, do you pull the gate of a PNP above, or below the source in order to turn it on.

cursive sentinel
#

You're mixing terminologies there. PNP applies to BJTs, not MOSFETs. For a p-channel MOSFET you pull the gate below the source voltage.

jade wedge
#

One day I'll get them right.

cursive sentinel
#

Learning!

twilit aurora
supple pollen
#

I don't think so, but you can probably find one pin versions for sale at electronics distributors

#

Then again, maybe so. It's worth a try

hushed ermine
#

I'm sure you can also just rip pins out with pliers, that's a kind of reconfiguration

inland jungle
#

one pin versions might not be mechanically stable

worldly schooner
twilit aurora
#

Nice! That works for me. I am searching for a robust solution to mount some wires temporarily to the breadboard without the need of any tool.

scarlet jay
#

HOLA MONA HEY!
I found FLUX AI and it is better than KiCAD, EasyEDA and many others...AI in HARDWARE!!!! Massive
I have this schematic and I am finishing it with NORDIC chip, DIGIPOTS and INA219 for production if anybody wants to help me for this ADVICE

supple pollen
distant raven
supple pollen
#

It claims it helps with repetitive tasks. I don't have issues with those, I just use copy/paste, submodules, and scripting for that.

distant raven
#

And I personally think that’s the biggest hurdle with making decent assistive tools because people think it’s a replacement for actual mastery of topics

supple pollen
#

I also have issues with web-based design tools. My main beefs are the UI integration via a web browser can't be as smooth as a locally running app; security in that someone else has access to my designs (not a big deal usually, as I open source most of of my projects but some for-pay work could be impacted); and the worry that the tool will go off the air or vanish forever with no warning.

distant raven
#

Altium is trying to switch people from perpetual licenses to subscriptions

#

I’m glad my free monthly subscription to KiCAD won’t ever go up in price 😁

supple pollen
#

I spent $1400 for an Eagle license, then they switched to a subscription model. Grump. Altium was never an option as it's DOS-only.

#

I switched to KiCAD a few months ago.

distant raven
#

But now I feel pretty good about using it. I’ve even donated portions of my board sales to KiCAD. Which I’m actually due to donate some more soon

supple pollen
#

That's awesome of you.

#

I can see myself contributing to the code base at some point.

distant raven
#

This board was derived from one I made in Eagle, ripped out a bunch of it and added a lot to and mostly finished at this point

Screenshot_2024-06-06_at_2.22.02_PM.png
#

The differential routing and meandering is a lot more straightforward in KiCAD than Eagle which is nice

#

A pretty friendly package to add 10base-T Ethernet to a project

pale grove
#

lcsc says discontinued

supple pollen
#

Chips like that get discontinued all the time. AdaFruit generally buys in large quantities so they have some time to find a replacement chip before they run out (otherwise boards go out of stock for a while). NAND flash chips like that are common enough and it should be a straightforward process to find a replacement.

tough matrix
thin falcon
#

I'm working on a project and I'm using the RP2040 Prop Maker Feather for the first time. On the screw terminals, there is a signal labeled "EXTERNAL_BUTTON", GPIO19. Any reason I can't use that terminal to control a relay? Just wondering if the board has any circuitry that makes that connection only suitable as an input from a button.

📎 Adafruit_RP2040_Prop-Maker_Feather_PrettyPins.pdf
supple pollen
#

Right in the product description, it says "Extra Button or Output Pin - One more pin on the terminal screw block can be used for button input or digital output, for activation or a simple LED."

worldly schooner
thin falcon
#

The other thing I need to finalize is power. I

#

have one component (disco ball) that is powered by 120V. Thus my relay. So, after development, the thing will be wall powered. I included a 120V-5VDC power supply. I'd like to connect it to the USB pin on the feather via a diode. Will the feather's on-board voltage reg be OK with ~4.3 (5V less a diode drop)? Does that sound like a sensible approach to working in the 2 scenarios: (1) feather powered from USB connection to PC during development w/ 5V power supply on but no currnet draw due to diode and then (2) feather USB disconnected from PC and feather powered from 5VDC power supply via diode?

supple pollen
#

Yeah, that'll be fine. It uses a low dropout regulator, as it can also be powered by a lithium cell which normally provides 3.6-4.2V.

thin falcon
lyric nymph
#

Does anyone know exactly which USB-C socket Adafruit uses on their QTPY? The schematic just calls out "USB TYPE C", but doesn't show a manufacturer part number...

distant raven
#

GCT, Amphenol, and other brands carry some various of it

lyric nymph
#

Sweeeet. Thank you!!!

lyric nymph
#

@distant raven would you mind if I DM'ed you? I've got a couple of newb-ish questions about that connector..

distant raven
#

Ask them here so everyone can learn

lyric nymph
#

Fair enough. So, I'm used to using USB2.0 headers which have VBUS, GND, D+, D-, and maybe a case ground too. The docs for the item you linked (USB4105-GF-A) clearly shows more pins than even the Adafruit schematic I linked previously has. So on a schematic that only shows D+, D-, how would I go about using the USB-C you recommended, which has DP1, DN1, DP2, DN2?

distant raven
#

Connect DN together, and DP together. It is there because USB C cables usually have 2 sets of DN/DP pairs. USB2.0 generally have one set. Tying them together makes it so your board will work regardless of which orientation your cable plug is

#

I imagine that Adafruit’s part just connects them together in the part library

supple pollen
#

That makes sense, it could just be implemented in the footprint and only expose the final connections to the netlist

distant raven
#

You will also need CC1 and CC2 connected to ground via 5.1k resistors

distant raven
#

Kind of a fun approach

supple pollen
#

Oh yeah, I remember that. An interesting approach, sort of like Monty Python's 3-sided record.

distant raven
#

A classic

supple pollen
#

On the less interesting side, there's the Raspberry Pi which just hooked CC1 and CC2 to the same resistor, which caused various odd problems.

distant raven
#

I do recall that. A lovely mistake to make

lyric nymph
distant raven
#

Probably fine if you use a USB2.0 usb C cable. But bad if you used a complaint usb c cable and power brick

lyric nymph
#

Interesting design choice with two separate micro's for each of the D1/D2's!

#

So with the 5.1k CC1/2 resistors, if I plug the board into a full-blown USB-C power supply, is something bad going to happen?

distant raven
#

USB-PD will try to negotiate over the CC pins if they are left unconnected. If they sense a specific amount of current like which would be set by your CC pull down resistors, it’ll deliver 5V and up to 3A

lyric nymph
#

Awesome! So this is what I have at this point..

image.png
distant raven
#

Looks good

lyric nymph
#

Sweeeeet. Thank you very much for your help, @skerr!

distant raven
#

You’re very welcome

#

Best of luck in your journey, it’s a fun journey to take

lyric nymph
#

I'm building a "prop controller" board. Intending to use the ESP32-C3 similar to the QTPY, and the MAX98357 from the Adafruit I2S 3W Class D Amplifier (https://www.adafruit.com/product/3006). Should be able to handle a bunch of neopixels, and maybe 8 or so PWM'ed LEDs... If I can figure out how to integrate these all onto a single PCB!

tough matrix
lyric nymph
#

Yep, exactly. Though, looking at that product more closely.. I might just give that a whirl instead of "rebuilding earth".

supple pollen
#

If you want to get fancy, you can connect the shield via a 1MΩ resistor instead of directly to ground.

distant raven
pale grove
#

Almost 25 minutes for the 60 files

#

Fine for a prototype but if I have to do it for let's say 100 pcs it will take forever

tough matrix
#

what baud rate do you use for serial connection?

tough matrix
#

even the default 115200 baud translates to roughly 11kB per second, or about 6 minutes for 4MB - slow, but not 25 minutes, even with the overhead
Writing to the NAND chip can be done really fast - SPI on RP2040 runs at up to 62 Mhz.

unique patio
pale grove
#

115200

cursive sentinel
#

Keep in mind that baud refers to total bits per second. That means it includes overhead. Assuming the typical case of one stop bit and one start bit, that means every byte takes at least 10 bit times to transfer.

#

This makes the effective bit rate 115200*8/10 = 92100bps.

tough matrix
pale grove
#

VScode on pc:
https://pastebin.com/JykXsFJb

circuitpython on rp2040:
https://pastebin.com/cTkWXrG7

This is what I use for the file transfer

Pastebin
import serialimport timeimport osimport hashlib# Replace 'COM10' wi...

Pastebin.com is the number one paste tool since 2002. Pastebin is a website where you can store text online for a set period of time.

Pastebin
def is_sd_card_empty(path): try: return len(os.listdir(pa...

Pastebin.com is the number one paste tool since 2002. Pastebin is a website where you can store text online for a set period of time.

worldly schooner
#

Test

distant raven
worldly schooner
#

Oh what

#

I thought I deleted that

#

My internet is messing with me.

tough matrix
#

looks like it is not only you.
Discord was glitching for me as well

obtuse temple
#

Does anyone know if there is a remedy yet for the "align to grid" option in Fritzing yet? or do we still just have to deal with it

stark arch
#

I'm building a PCB soon, and I faintly remember using a software last time which would show where which components go on the board, and allow you to check them off as you populate the board. I can't seem to find what I'm looking for (probably using the wrong search terms), does anyone know what software/website might be able to do that?

cursive sentinel
#

I mean... KiCad gives place and route stats.

obtuse temple
#

sooo close but not quite

#

I am trying to figure out how to get them to align lol

#

if anyone has any ideas

#

it appears the connectors' terminal points are not aligned to the grid

image.png
obtuse temple
#

Just figured it out

#

Fritzing aligns by the first "connector"

#

I removed the battery "connectors" and I2C "connectors" on the Qwiic connector and now it is aligning fine

#

picture with the feather moved to illustrate the connector positions, but I got rid of all connectors except for the pins directly from the feather itself, and now the problem is fixed

image.png
#

This information might be relevant to some big wigs in regards to the Adafruit Fritzing Libraries. <@&327289013561982976>

#

At least, specifically for the ESP32 Feather V2 part as that is the only one I have been messing with so far lol

#

it seems that either the battery or I2C connectors are the main culprit in alignment issues with this part specifically

distant raven
unreal bay
#

I work with custom Fritzing parts. You "may" be able to fix those by changing the numbering, and/or moving the 'off alignment' connectors to the end. It has not been clear how Fritzing chooses which is the 'first' connector to use for setting the origin, which in turn sets the alignment.

distant raven
obtuse temple
distant raven
#

Most don’t really have that ability. There are a few Adafruit folks in here but they don’t generally do those things. They’ll tell you to email support or submit feedback on a learn guide so they have a record of it and can add it to their todo list

#

That’s usually the best way to bring attention to things

supple pollen
ripe dawn
#

Choosing the Correct Capacitor

pale grove
#

Got another question for a "mass production" type project. (Let's say 1000+ pcs as a start)

I need to have components in 2 different locations in my enclosure. It cannot be one flat rigid PCB.
So I guess I have the following options:

  1. Flex PCB

  2. 2 rigid PCBs + a flex cable connector between the two?

  3. 2 rigid PCBs + soldering wires (around 6 wires)

#

I am guessing the flex PCB is the "simplest" solution?

and the flex cable connector option the cheapest?

supple pollen
#

That's my guess too

#

You can get mixed PCBs with rigid and flexible parts, but I'm guessing those are more expensive.

#

The soldering wires might be cost effective as well (depending on who's doing the soldering and how much they're paid)

worldly schooner
#

Other considerations for interconnects include how far, how many signals, how much power, how fast the signals, etc. There’s no singular right answer, but there are plenty of options to select from depending on your answers.

#

If your interconnect only needs six wires, there may be some off the shelf discrete wiring solutions out there as well, depending on distance and volume.

tough matrix
limber thunder
#

can help me understand this battery charger (datasheet, TPB4056B)
i dont know how im supposed to get power from the battery
do i just connect it to the pos and neg terminals regardless of charging status?

limber thunder
#

also, what's the little capacitor here supposed to be

image.png
distant raven
#

To use the battery, you’d use a battery switching circuit like Adafruit uses.

limber thunder
#

this is kinda a general question: if a part im using has multiple pins of the same name, and you're supposed to connect another individual pin to it, how would i go about that? would i link all four of them?

#

LX is the thing with 4 pins

distant raven
#

What’s the part?

limber thunder
#

MAX77787JEWX+

#

this is what it shows in the diagram

image.png
supple pollen
#

Yes, link them together. This is to reduce resistance/inductance and improve cooling and current capacity. The datasheet gives a recommended PCB layout.

Screenshot_2024-06-16_at_10.28.06_AM.png
limber thunder
#

thank you

#

i wish this layout was a bit easier to read

heavy jasper
#

A warning - this is a 0.4mm pitch part, you’re likely going to need some strongly non-standard trace width/clearance/via size rules to make this layout work - and commensurately this may be a pretty expensive board to fab.

supple pollen
#

You can download the design files for the evaluation board(s) and look at the PCB layers individually for more detail

limber thunder
#

just need a chip that can charge a battery at like 750ma over usb, and them somehow get that power to the rest of my circuit

#

i had this chip reccommended to me and uhh

#

i love the recommendation but it's a lot

distant raven
supple pollen
#

BGA is often a bit much for beginners

distant raven
#

That does up to 800mA charge current

supple pollen
#

The usual approach is a 2-chip solution, one to charge the cell and one to condition power to the rest of the circuit (depending on what the circuitry needs: some circuitry can run directly from a LiPo cell, others will need a regulator and/or booster).

distant raven
tough matrix
#

I am probably nto a beginner, but so far, I never attempted BGA....

distant raven
#

This is how you can connect your battery to your power circuit

limber thunder
distant raven
#

But it’s definitely not something you do on your very first, or even your first 10-20 designs

limber thunder
#

can someone explain to me what this does?

image.png
distant raven
limber thunder
#

mk so it isnt shorting to ground with a capcitor

distant raven
#

Helps handle ripple current, and to an extent filters certain frequencies

limber thunder
#

would a multilayer ceramic capacitor work for that?

distant raven
#

That’s what they expect

limber thunder
#

ah cool

distant raven
#

Unless it specifies otherwise, assume MLCCs

supple pollen
#

The datasheets tend to be long, but they're chock-full of useful information, like details on how to choose capacitors and sometimes even specific part recommendations

limber thunder
#

VISHAY Si5481DU-T1-GE3
is there an exact (or functionally equivalent) match for this part
the SI5459DU-T1-GE3 is recommended as a repalcement but im worried because the specs are a bit different

tough matrix
#

P-channel MOSFET is pretty generic part.
You probably don't need very high currents or voltages (nothing above 20 V or 5A, I hope?).
If so, then the only things that are really relevant are resistance (R_DSon) and threshold voltage (V_gs th)

limber thunder
#

the vgs th max on the newer one has a higher voltage at same amperage

cursive sentinel
limber thunder
#

rdson max has more than double the mohm count

cursive sentinel
#

What are you using it for?

limber thunder
#

i ended up picking a more powerful charging chip after i realized the battery would take 5 hours to charge on the old one

#

i need to put the mosfet at mp1

image.png
# 
BATGATE (Pin 17): Battery Charger and Ideal Diode Am-
plifier Control Output. This pin controls the gate of an
external P-channel MOSFET transistor used to charge the
Lithium-Ion cell and to provide power to VOUT when the
system load exceeds available input power. The source of
the P-channel MOSFET should be connected to CHGSNS
and the drain should be connected to BATSNS and the
battery
distant raven
#

Pretty close

#

Weird package but you want that for good physical contact for sinking up to 19W

limber thunder
#

ah sorry i forgot to come back but i already found something close

#

DMP2022LSSQ-13

distant raven
limber thunder
#

that is a good question

#

it's already routed in there, i can probably remove it but i wanna try and finish the charging circuit before i have to leave

distant raven
#

Okie dokie

limber thunder
#

ty tho

hasty solar
#

hey folks, I'm planning a small controller driven device for a planter bed in my yard that I'd like to report metrics out to a server I run. the planned location is well outside the range of my wifi and it's a solid 50 feet from the house, so running cable across the whole yard seems like a poor option. does anyone have other suggestions to get it internet access without just buying a sim card and using the cellular network? bandwidth and latency don't matter, just a stable enough connection to send a small http request out about once per minute

worldly schooner
# hasty solar hey folks, I'm planning a small controller driven device for a planter bed in my...

https://learn.adafruit.com/adafruit-feather/lora-radio-feathers might be a good read. Of course, it's an additional device, but if the goal is wireless connection over 50-100 ft, it's tough to top. A pair of LoRa devices would connect your planter to another internet-connected device like an ESP32 or a Raspberry Pi.

Adafruit Learning System
Introducing Adafruit Feather

Boards of a Feather flock together!

hasty solar
#

that could reduce the footprint of devices that I need to power and waterproof externally actually, that's really great. thank you!

hushed ermine
hasty solar
#

pair this with a solar panel and battery kit and I might just be able to make this controller totally self contained

#

second question: is there a particular type of gasket or other protection that's recommended for sealing cable entry points on device enclosures? I want to keep this device totally sealed against rain if I can
alternatively, am I overthinking this? is it enough to just put a tarp over the top so it doesn't actively get rained on?

digital stream
#

Cool project. Hope you share the end result. I play around with waterproofing a lot of solar enclosures but at lowest cost. Hot glue is surprisingly effective short term. Vaseline can also help against corrosion of stuff if water does get in. I’ve seen but not used potting compounds (they’re expensive but worth it if you have pricey components). Other than that there’s stuff like the wagobox capsules, but they use standard gland sizes.

tribal orchid
#

Hey, I am planning on making a slim virtual dice and using the new-ish displays. I am trying to copy the devboard design (the Qualia RGB-666) but want to ditch the multiplexer and reroute pins to fit it in the esp. Is there a way to reconfigure the firmware? Also, what is sercom? And how is it used to regonfigure the SPI and I2C pins?

  1. Picture is my design; 2. is from Adafruit GitHub
    My schematics arent final at all btw, i am still working on the details but got stuck on the multiplexer.
image.png image.png
supple pollen
noble loom
#

Hey! I am making a board that uses a relay. I plan on switching that relay using a transistor, as my microcontroller can't supply enough current. My relay uses 100 milliamps, and the transistor I picked out can do a continuous current of 75 milliamps. Is that 33% margin enough or should I choose a beefier transistor?

tribal orchid
supple pollen
distant raven
supple pollen
#

Even more when surface mount versions showed up

#

I like bipolar transistors for working with low voltage logic, I have an easier time with bipolars compared to MOSFETs when I only have 3.3V to work with.

distant raven
#

Yeah, they make life easier being current driven devices rather than voltage dominated

#

Add in Darlington pairs and you can control quite a bit of current with just a little microcontroller

supple pollen
#

I've been doing that more lately, as microcontroller drive capability gets reduced. There are a few that I wouldn't feel comfortable pulling 5mA out of a GPIO pin. But there's plenty of voltage to bias two B-E junctions in series, so a 1mA draw into a Darlington pair will generally switch most of the things I want to control.

#

Since I learned on vacuum tubes, I'm still getting used to the idea of 3.3V being "plenty of voltage".

distant raven
#

Hehe yeah, no need for hundreds of volts 😬

noble loom
supple pollen
#

I just threw that out as a common and inexpensive transistor where I live. Most small signal NPN transistors (such as BC547) would be suitable.

noble loom
distant raven
noble loom
#

Oh my goodness I just realised my mistake. The relay pulls 75 milliamps, and the transistor does 100

distant raven
#

Ohhh yeah, that should be fine then

noble loom
#

Thanks!

gloomy lion
#

Hello! Got a question about big ol capacitors and using one to power a microcontroller.

Not really sure how best to describe the scope of the project but I'm planning a circuit that's powered by occasional 12V pulses, from pressing a button, originally meant for plain white LEDs. I want to replace them with neopixel-compatible LEDs animated by a microcontroller. The microcontroller needs to stay running in between pulses so the lights stay somewhat responsive, without the microcontroller having to boot up each press. I can't wire in steady external power, so I'd like to use a Really Big Capacitor™ that gets charged up a bit on each pulse to keep the microcontroller listening for input between presses. The LEDs would just be powered by a 5V regulator while the 12V line is active.

My question is, if the 12V needs to be regulated to 5V for the microcontroller and LEDs, would it be more efficient to store 12V in the cap and regulate that, or place the cap after a 5V regulator, with the regulator fed directly by 12V? Also, is there anything I should consider for the specific capacitor other than having a low ESR value and super high capacitance? I haven't done much with electrolytic caps before so I'm a little out of the loop

distant raven
#

And a super low power micro like a attiny

tough matrix
#

I need some help from experts.
I am reverse engineering so e "smart servos". Initially it looks like they use some version of Dynamixel protocol, which in turn is just some data frames transmitted over half-duplex uart.

But there is something odd that I see on the scope

#

the two periods of low line level are usual start bits.
But that really short leftmost low "spike"?
that is too short to be a start bit

#

here it is at other scale

PXL_20240617_175417337.jpg
#

width is 240 ns (while the width of a bit is 13 us)

gloomy lion
#

Just wanted to get the question out there in the meantime!

distant raven
#

You’ll need to step down the 12V down to ~5V

#

Alternatively you can use a bunch of MLCC in parallel. Like a bunch of 47uF MLCC together

gloomy lion
#

Cool! The supercapacitors are a better price than I expected too. Thanks! I'll look into using MLCCs as well, especially if it saves some vertical space

distant raven
#

It just depends on how much time you need it alive between pulses

inland jungle
#

that's a lot of 47uF to hit .3F 🙂

distant raven
#

Also consider that you will want to have a boost converter if you need to get back to 12V for the LEDs (if you’re using the 12V variety)

distant raven
gloomy lion
inland jungle
#

? The super cap was $3

distant raven
#

Yep

gloomy lion
distant raven
#

If you know your total current usage, you should be able to solve for the capacitance you need with a little calculus

gloomy lion
#

Also, it's just the microcontroller that'll need power between pulses, the lights won't be powered by the cap at all

#

So that makes capacity a bit less of an issue

inland jungle
#

2mA (attiny) * 30 seconds = 0.06 C, let the voltage drop ~1.5V over that 30 seconds 5 -> 3.5V, and you need 40mF

distant raven
#

So 400x 100uF capacitors 😬

#

Or one super capacitor

#

Also which attiny?

#

I think the newer ones can use less current to operate

inland jungle
#

attiny85, but it depends on clock frequency

#

also maybe sleep modes make sense, not sure how long it takes to 'wake up'

gloomy lion
#

If wakeup from sleep is like a few ms tops that would probably be fine. I just need it to be fairly responsive, with very little visual delay between button press > light turning on

inland jungle
#

it should only take a few clock cycles I think, but I've never done it

gloomy lion
#

Sounds acceptable tbh, I might pick up a cheap attiny board just to play around with it

#

Thanks for all the recommendations and info, you guys rock

dusky jetty
#

I have a question: On the Adafruit DC+Stepper Wing there is an AON7401 at the 12V power connector. Is that for reverse voltage protection? I don't know how the TB6612FNGs pass the 12V so can you share why there wouldn't be some form of voltage protection between the TB6612FNG and the motor blocks?

tough oyster
#

Hey all. I have a very simple pcb I need designed. I would attempt to learn how to design myself, but I do not have a computer so it is basically impossible to make it myself.

I come here to ask if it is okay to commission my design. I looked over the rules and couldn't see anything prohibiting solicitation so I decided to ask. Please ping me if you are up for the task or if this is against the rules and I'll take this down asap :)

supple pollen
tough oyster
#

I just need whatever you need to send it to pcbway :P I think that is just Gerber and BOM but I may be wrong

#

I'll pm you if that's okay... Im currently at work but I can send you the requirements when I get home

inland harness
#

I need help figuring out a schottky diode issue

#

Is this the expected voltage drop?

rn_image_picker_lib_temp_cfde8ee0-abf7-468a-9c0c-12bdbfe5c0c4.jpg rn_image_picker_lib_temp_221fded8-df03-4945-b850-db02b6e301ae.jpg
#

I want to drop the voltage in both directions, that's why I did this

supple pollen
#

I see a cell being measured but without the diodes in series either time

inland harness
#

I need the voltage to drop in both directions on the same power rail. Sometimes power will be flowing in, sometimes it will be flowing out.

#

I wanted to limit my battery to charging/discharging between 20% and 80%

#

This is the best idea I could come up with.

#

Am I doing something wrong?

supple pollen
inland harness
#

They don't

inland harness
#

One has diodes, the other has no diodes

supple pollen
#

I thought that was the battery tab. Note that diode voltage drop is roughly a 3/2 power law, or in other words, it depends on the current drawn. A multimeter presents about a 10 megohm load so draws almost no current. You'll see a higher voltage drop with some current draw

inland harness
#

Will these work as I'm intending them to though?

#

Also I didn't know that current draw was a factor

tough matrix
#

how much voltage drop do you need?

inland harness
#

When charging the battery I need it to stop before the battery is full because of the diodes

#

And i have a chip that has a voltage cutoff of 2.75v and I want it to cut off sooner for battery health

#

.4v drop is ideal and that's what I was led to believe by internet explanations of how schottky diodes work

supple pollen
#

You'll see from .2 to .5 volt drop depending on the diode, for a reasonable range of currents

inland harness
#

Most of the wear when charging and discharging the battery happens between zero and twenty percent and eighty and a hundred percent.

inland harness
supple pollen
#

Almost no drop. I don't know why you'd want to drop voltage coming out of the battery, that's just wasted power.

inland harness
#

I already said

#

I need to trigger a cutoff circuit sooner

#

What do you mean wasted power?

supple pollen
#

The voltage drop in the diodes represents wasted power that's dissipated as heat

inland harness
#

I specifically did this for battery health...

#

Well guess there goes that plan

supple pollen
#

I get that, just trying to wrap my head around the use case

inland harness
#

Meshtastic node on the top of a random mountain

supple pollen
#

For battery health, I normally opt for lower voltage chargers (harder to find) and protection circuitry with a reasonable cutoff (or an adjustable one). Or just reach for LiFePO4 and enjoy much better lifetime and safety

inland harness
#

Also it gets cold so heat is not an issue and may be a good thing

supple pollen
#

Ah, you don't want to charge the below freezing

inland harness
#

I'm not using lifepo4 because temperature

#

Notice the diodes are right on the battery contact

#

This is lithium IMR so it should still work for a bit below freezing

#

I have insulated the box as well.

#

How much power do these diodes actually use?

#

I'm trying to figure out if the solar panel can offset their usage over time

#

It's a 2w panel and that board sips power

supple pollen
#

The power is simple, the voltage drop multiplied by the current through them

#

So if they're dropping 0.4 volts at 1 amp, they'd be dissipating 0.4 watts (which is a fair amount for a small package like that, especially in an insulated box)

inland harness
#

Huh so it'll barely keep up

#

Thank you for your help!

#

By the way, because it's only going to barely keep up, I'm just going to go without the diodes, because the battery should never get to cutoff voltage, even if it's in a basement facing away from a dirty window

#

The downside will be that the battery will almost always be at 100% which I know is not good for them.

#

The battery should last for weeks if the solar panel suddenly dies somehow.

supple pollen
#

You could try the diodes and see, I suppose

tough matrix
# inland harness Huh so it'll barely keep up

well, what madbodger said was an example.
1A is a lot - typical microcontroller project uses a lot less (unless you have beefy motors).
So if your project uses 100mA, then the diodes would use 0.1A*0.4V= 0.04W

inland harness
#

Micro-amps are very small.

#

Not milliamps

#

When that rak board is in power saver mode, it uses 10 microamps.

#

I think my multimeter uses more power than the rak board.

#

I can power the board on a 3.7v 2032 lithium battery for a few hours in normal mode

inland harness
supple pollen
#

Presumably the current draw is to small to achieve much voltage drop

dusky jetty
# dusky jetty I have a question: On the Adafruit DC+Stepper Wing there is an AON7401 at the 1...

Replying to my own question with an answer from the AF forum:

The TB6612FNG has internal diodes to protect the output MOSFETs from reverse voltage. See the output topology diagram on page 3 of the datasheet: https://cdn-shop.adafruit.com/datasheet ... 121101.pdf

It is also possible to short-circuit the generated back-EMF current so that it acts as a breaking force on the motor. See page 4 of the datasheet for a diagram of that.

pale umbra
#

is there a better way to mount this than soldering directly to wire as I dont have the gear to do surface mounting ? This would be soldered to a 14 awg

#

Like I see fuse holder and look up fuse holder and that fuse and had no result but maybe there are generic fuse holder for that mounting ??

worldly schooner
cursive sentinel
worldly schooner
#

I don’t think that’s necessarily any better for this particular case, but perhaps…?

#

I’d probably solder this to some sort of tab connector intended for quick disconnects, then crimp the wire in. Don’t think it’s any easier, but it could present some mechanical advantages.

pale umbra
#

Ill look that out seems like everyone I ask as a different opinion about the rating of the fuses I need but eh I guess I'll check if 60A would be throught hole

copper lintel
#

https://www.digikey.com/en/products/detail/espressif-systems/ESP32-S3R8/15822451
does Digikey only sell engineering samples? I am running into some weird issues, and I think it might be because the chip is an early revision (v0.2), so for my next board I want to use the same model number, but a full release version.
I looked at mouser, and there was no disclaimer about the parts being a sample, but I am not sure

DigiKey Electronics
ESP32-S3R8 | DigiKey Electronics

Order today, ships today. ESP32-S3R8 – IC RF TxRx + MCU Bluetooth, WiFi 802.11b/g/n, Bluetooth v5.0 2.412GHz ~ 2.484GHz 56-VFQFN Exposed Pad from Espressif Systems. Pricing and Availability on millions of electronic components from Digi-Key Electronics.

supple pollen
#

If you wanted to get more fancy (and expensive), you could use ANL style fuses.

#

If you prefer traditional cartridge style glass tubular fuses, you can get AGU types in 60A ratings

#

Fuseholders designed for wires are available for all of these. Note that 14AWG seems insufficient for currents in the 60A range.

pale umbra
# supple pollen Fuseholders designed for wires are available for all of these. Note that 14AWG ...

I think 20-30A will do since my current uses would be in the 11-12A range max. I thought Id be able to protect the battery from internal short with an external fuse but from what I can gather it isn't possible and I also don't know what the lipo batteries amps capacity is, I merely have it's C rating (ie: discharge rate) and its capacity in mAH. So it seems what I have to do is to protect what is connected to the battery instead

supple pollen
#

The ampacity is basically the capacity times the C rate.

pale umbra
#

And I suppose Ill have to be careful to not connect the battery to a inverted diode by accident

supple pollen
#

A fuse will protect from external shorts, but you're right, a cell with an internal short is going to self-destruct

pale umbra
supple pollen
#

Connecting it to a series inverted diode would just block the current flow. No big deal. Connecting it to a parallel inverted diode is basically a short circuit (a fuse will help here)

pale umbra
#

sorry for the edits just woke up 🤣 my math was a bit off

supple pollen
#

While your math is right, 640A is unrealistic.

pale umbra
#

yeah but I dont have any other information except that it is suitable for 125A r/c ESCs

supple pollen
#

I'd take all such marketing claims with a large grain of salt

#

But it's wise to avoid short circuiting such a pack in any case.

pale umbra
#

I dont actually expect it to ever provide anything close to 640A either I just dont know how much it could provide by accident like if I connect something wrong like a diode

#

and hence Im not sure what diode to get for such a pack to protect me and the circuit in case I connect something wrong and dont realize it. y understanding for a diode plugged the wrong side is that it would pull the max the battery can provide

supple pollen
#

Hence the fuse.

pale umbra
#

yeah and the max fuse I could find were 100A or 200A but they are surface mount incompatible with any holder or circuit breakers assemblies

#

obviously it cant do 640A, but I dont know how much it could do and my understand the fuse current capacity have to be higher than the battery

#

so I dont know what fuse to get for it

supple pollen
#

Surface mount is silly at such currents, copper traces can't handle that kind of current anyway

#

No, you set the fuse capacity to the maximum current you want to allow. I'd probably use an ordinary 10A fuse.

#

The fuse voltage has to be higher than the battery, but in low voltage circuits like this, it isn't much of a limitation, there are plenty of high current fuses with a 32V rating for automotive use that would be totally fine.

#

The idea is to damage the fuse instead of damaging your battery or circuitry. And the battery would certainly be damaged by a 640A current flow

#

You make the fuse the "weak link" intentionally so it blows first

pale umbra
supple pollen
#

Yeah, 1A would be fine in that case, and provide nice protection.

pale umbra
#

what about the breaking capacity ? I think we are talking about current capacity so far ?

supple pollen
#

Since you said you were connecting with 14AWG wires, I'm unsure why you were looking at surface mount anyway.

pale umbra
#

I only had 4 choices at 100A/200A, 2 of the them were over 10$ the 1$ 2$ ones were surface mounts after filtering on digikey

#

so it wasn't by choice but having to pick 1 chocie between 2

supple pollen
#

An ordinary cheap common ATOF style automotive blade fuse is rated for interrupting 1000A, so that would be plenty

#

An ordinary 3AG glass fuse is rated for 10kA interrupting capacity

pale umbra
#

yeah that seems good and I suppose there are a lot of fuse holders for it with wire in and out on the holder where I can just solder my own wire directly on it

pale umbra
#

Ok Im checking littelfuse website to get 2 fuses and 2 fuse holders

pale grove
#

Got another question. Let's assume I finish my PCB design with the RP2040. I get PCBs made, put my program on it (circuitpython) and distribute the devices to potential users.

Later on I need to update the code to fix bugs and add features.
How can I let the user update the device - but not give "access" to the python code and so on?

Can I somehow add a USB-C connector and have them updated it that way (but not give access to the file system?)

#

Or what would be the best approach?

limber thunder
#

is there an official footprint for the rp2040?

pale umbra
limber thunder
#

found it

oblique vale
#

you can also get little resettable circuit breakers in an automotive blade fuse package, which are cool

limber thunder
#

answered should i use aluminum electrolytic or aluminum polymer capacitors here

image.png
hasty solar
#

i'm exploring the idea of building a fairly simple plant watering device entirely with circuitry without needing to handle running a controller or writing control software, but am hitting the obstacle that every capacitive moisture sensor i can find seems to be an I2C sensor, which doesn't lend itself well to non-software controls. i found this one, but its datasheet doesn't explain how the analog pin works, and it seems to do a bunch of extra stuff i don't want and thus don't want to pay for. is there a simpler product that just outputs capacitive moisture readings to an analog pin and documents how that functions?

hasty solar
#

new question as i look at circuitry: i'm interested in setting up a switchable 555 timer to control watering, with both "on" duration (how long the valve is open) and minimum "off" duration (how long before the valve can reopen once closed) being controllable with separate potentiometers. could i do that by modifying this monostable diagram to add a potentiometer between C1 and the 6/7 pins, but not in series with the switchable resistors they have in their diagram (which i'll replace with a second potentiometer)?

waveforms-tim40.gif
hasty solar
#

seems like my idea won't work since it adds too much resistance to the discharge path for it to stay discharged in its off state, but this diagram looks like it'll work (with pin 2 separated so i can control the on cycle) by keeping the resistors in series but shorting out the discharge resistor during charging https://www.electronics-tutorials.ws/wp-content/uploads/2019/07/tim72.gif

hasty solar
#

circuit diagram i ended up speccing out - will this work? resistance/capacitance values are semi-arbitrary based on components i have and the range of cycle lengths i wanted

circuit.png
supple pollen
#

I don't think that will work, but you can pull the setpoint around with an external voltage into pin 5

#

After reading your description, it might work depending on what the signal from the moisture sensor looks like. You'll get a lot of emitter degeneration.

limber thunder
#

if i wanted to do a 4*3 grid of buttons using as few pins as possible, how would i go about it?
there's a way to do it with resistors and an analog pin but i havent found any good concrete info on how exactly it should be done

drowsy drift
limber thunder
#

looks like i can use an i/o expander

drowsy drift
# limber thunder problem is that's like 80% of the pins i have remaining and i still need to wire...

The multiple-buttons-with-analog pin can work but can be finnicky and usually analog pins are precious. Here’s one technique: https://forum.arduino.cc/t/multiple-buttons-on-1-analog-pin/8622

Arduino Forum
Multiple Buttons on 1 analog pin

Thought i'd share a little project i've been working on. As many may know there is a method for using a single analog pin to control multiple buttons. There is even a library for it. Problem is the way they connect the circuit doesn't allow for the use of button combinations. So you couldn't press 2 buttons at the same time to perform a diffe...

drowsy drift
# limber thunder looks like i can use an i/o expander

Yes! An I/O expander is perfect. For buttons I really like using something like the 74HC165 shift register. Here’s an Arduino example. You can use keypad.ShiftRegisterKeys if in CircuitPython https://dronebotworkshop.com/shift-registers/

DroneBot Workshop
DroneBot Workshop
Shift Registers - 74HC595 & 74HC165 with Arduino

Learn to use the 74HC595 and 74HC156 shift registers to add extra input and output ports to your Arduino, then use them to build a fancy LED light display!

supple pollen
#

You could also use a chip specifically designed to operate keypads

hasty solar
#

my hope is to have the 555 tripped when the analog voltage from the moisture sensor goes above some threshold

limber thunder
#

is this a correct key matrix?

image.png
supple pollen
#

Yes, but without series diodes you can run into an effect known as "ghosting" if multiple keys are pressed at once.

limber thunder
supple pollen
#

For true ghosting prevention (3 or more keys at once), you need a diode per switch. But a diode per row is sufficient for up to 2 keys at once.

limber thunder
#

where would the diode go

#

if i wanted 2 key

supple pollen
limber thunder
supple pollen
#

Between the matrix and the GPIO

Screenshot_2024-06-23_at_5.24.05_PM.png
limber thunder
supple pollen
#

I think that's the side connected to GPIOs configured (one at a time) as outputs?

limber thunder
#

(dumb question) why would there need to be a diode on an output?

#

if something goes back into it would it matter?

supple pollen
#

Good point. I misread your question about pin 3 and the row wire, and mentally jumped to that diagram (which is for a combination LED and keypad driver, which needs the diodes to keep the LEDs from confusing the keypad circuitry). You can omit the diodes if you don't need to prevent ghosting

limber thunder
supple pollen
#

Then you don't need any diodes

limber thunder
#

i was trying to pick a ~2.7-3.2" mono backlit lcd but all of them use like 10v

#

not like one of the super wide rectangular ones but something that's more square but not entirely square

supple pollen
#

There should be 5V ones that aren't too hard to find

limber thunder
#

all im seeing is 8-12v ones

#

e-ink might be an option tbh

supple pollen
limber thunder
supple pollen
#

You just gave one dimension and specified mono, backlit, and LCD

limber thunder
#

not like one of the super wide rectangular ones but something that's more square but not entirely square

supple pollen
#

How much more square? Like 4x16?

limber thunder
#

similar to this shape

#

but not 3.52"

supple pollen
#

Character or graphic?

limber thunder
#

either or

#

im looking at eink displays but none of the ones im seeing come with frontlights

supple pollen
#

https://newhavendisplay.com/sv/4x16-character-lcd-stn-gray-display-with-yellow-green-backlight/

Newhaven Display
4X16 Character LCD | STN+ Gray Display with Yellow/Green Backlight

Newhaven 4x16 character Liquid Crystal Display shows characters with dark pixels on a bright yellow/green background when powered on. This transflective LCD Display is visible with ambient light or a backlight while offering a wide operating temperature range from -20 to 70 degrees Celsius. This NHD-0416BZ-FL-GBW display has an optimal view of 6...

#

Winstar, Crystalfontz, New Haven Display, 4D systems, lots of vendors offer them

limber thunder
#

they are not cheap lmfa

#

it's like $130

supple pollen
#

Yow. I have a bunch of their smaller ones, they were like $20. Not sure why that one is so expensive.

limber thunder
#

do you know where you got them?

supple pollen
#

Looks like prices have gone way up. The CU16025 I used to buy for $20 is over $80 now

limber thunder
#

i'm like really mad about that

#

those look so cool

supple pollen
limber thunder
#

i wish they werent all that aspect ratio but tbh that's a price im willing to pay

#

this was originally going to be a bit more phone shaped

supple pollen
#

I have some that aspect ratio and size, but they're giant 1 and 2 character displays, really low rez

limber thunder
#

most of the cost is the controller it seems

supple pollen
limber thunder
#

every single ounce of my being wants me to stick to vfds

supple pollen
#

I'm now really glad I bought a boatload of them when they were cheap

limber thunder
#

istg i was born in the wrong year

#

:(

#

if i can find an oled with that blue-turquoise-ish tint im seeing on the vfds it would be perfect

supple pollen
#

Might be able to manage it with a color filter

limber thunder
#

this is your fault madbodger

#

lmao

limber thunder
#

im not opposed to the idea

supple pollen
#

Yes. It totally is. Wait 'til you get hooked by cathode ray tubes...

crt-vector.jpg
limber thunder
supple pollen
#

I normally get my filters from Lee Filters

limber thunder
supple pollen
#

That's the tricky bit: it depends on what color OLED you're starting with

limber thunder
#

wouldnt white be the easiest?

supple pollen
#

White is probably your best bet

#

What I'd probably do is get one of their Cinegel swatch books ($10 or so) and try various ones (and combinations - stacking gels gives you a lot more color choices). The swatches may be big enough to use directly with your display.

limber thunder
#

that isnt lee. still the correct one?

supple pollen
#

Oh right, that's Rosco. Another big color filter vendor, they're good too.

limber thunder
#

wouldnt the filter make the background look blue as well as the light?

supple pollen
#

Yeah, but the background is dark (darker with the filter) so it's not terribly obvious. And using VFDs with filters was very common.

limber thunder
supple pollen
#

Yup. That's totally valid.

limber thunder
#

how do they adhere to the display?

supple pollen
#

Attaching them is left as an exercise for the user

#

You can use double sided tape, clamps, covers, etc.

limber thunder
#

hmm

#

looks like there are some cheaper chinesium vfds available

supple pollen
#

Yup. Saw a $2 one I like the look of

limber thunder
#

do i want to change the shape of the entire device just for a vfd though

supple pollen
#

That depends on your design priorities

limber thunder
#

it was originally gonna be a phone lmao

supple pollen
#

It's easier to put a VFD in than try to find room for a CRT but still...

limber thunder
#

do i want it to be the coolest stuff ever or do i want it to be phone shaped

supple pollen
#

I remember the old Motorola brick phones with an 8-digit LED dot matrix display

limber thunder
#

this might be the first rp2040-based phone

#

putting a vfd in it would just make it even funnier

#

(ignore the swag rat)

supple pollen
#

I don't recognize that phone module offhand. I do rather like the rat.

limber thunder
#

that's the ec25

#

it's an expensive chip but it does sms and calling

#

as well as 4g lte

supple pollen
#

Lots of them do SMS and calling, but yeah, 4G is more expensive than 3G

limber thunder
#

most of them are only data, no sms or calling

supple pollen
#

I was looking at something like the SIM-7600

limber thunder
#

yeah the ec25s are around $70

#

so far the features i have are: a lion battery setup (which i might need to change later), an rp2040, buttons (which i will need to change if i use the vfd), proper hifi audio (lmao) and that's pretty much it

#

i might need a microsd reader (the ec25 has one built in but idk if i can access it using the uart commands)

#

havent bothered to check

#

if i went with this display what shape would the rest of the device be

image.png
supple pollen
#

There are several possibilities.

limber thunder
#

im kinda at a loss

supple pollen
#

I'd be tempted to use an ordinary bar format, with the display in the middle, vertically, with the speaker above it and the mic below

limber thunder
#

what abt buttons

supple pollen
#

Several possibilities. You could have a keypad between the display and mic, or columns of buttons flanking the display, or get creative and have the buttons on the side or back (I'd have a few on the side anyway for volume and the like)

#

I'm thinking old school phone style

motorola-ultra-classic.jpg
limber thunder
#

that's a really cool idea tbh

supple pollen
#

You could also get fancy and hide the mic between the buttons (it's just a little opening)

limber thunder
#

i wanna see how big 7cm is holdon lemme find my ruler/caliper (whichever turns up first)

#

can the characters on these displays display vertically?

supple pollen
#

It would be tempting to grab one of those old school phones and build an updated one into it and reuse the buttons and display, but those old school phones seem to have become expensive collectors' items (even though you can't use them these days)

limber thunder
#

i was gonna 3d print a shell

supple pollen
#

Although the one pictured is only $60 on eBay (unlike one I saw for a whopping £33k)

limber thunder
#

wow that display is really small

supple pollen
#

Big displays weren't really a thing in those days

limber thunder
#

7cm is tiiny

supple pollen
#

... maybe

limber thunder
#

wait why are these so expensive

#

what step in the production makes them so pricey

#

i guess supply & demand

supple pollen
#

$2 is too rich for your blood?

limber thunder
#

im talking about the larger ones

supple pollen
#

Could be the power converter: it needs to create a low AC voltage for the filament and a moderate (like 30V) DC voltage for the grids and anodes. Also the control electronics also has to deal with 30V outputs.

limber thunder
#

its a shame tbh

supple pollen
#

The VFD module itself requires phosphors, cathode activation, vacuum processing, gettering, etc.

#

Some creative folks repurposed VFD technology to make a flat glowing triode amplifier tube http://www.nutube.us/

limber thunder
#

it almost looks similar to a vfd

supple pollen
#

I suspect it is VFD

limber thunder
#

on a receiver that modern?

supple pollen
#

Some folks like the look of it, and audio equipment design doesn't change very fast

limber thunder
#

i think we have one spare - could probably yoink it

supple pollen
#

Used equipment is a great source of VFDs. You may have to come up with your own drive electronics, but there are chips and power supplies available

vfds0010.JPG
limber thunder
#

ima have to pick up on this tomorrow

#

ill sleep on it now that i know my options

limber thunder
supple pollen
#

Yow, that is a nice find!

limber thunder
#

no controller

#

but i can probably buy one

supple pollen
#

One of the related items seems to have the controller but it's $78

limber thunder
limber thunder
#

you can solder vfds straight onto boards right?

supple pollen
#

Yup.

#

The far left one in the pic above is taken out of a microwave oven and soldered to perf board for an art project I'm building (working title: "Blinking 12:00")

limber thunder
#

wow i need some very steppes up voltages

supple pollen
#

Yeah, that's a good chip

limber thunder
#

so i gotta step lion voltages up to what like 60v?

#

that sounds like a current eater

supple pollen
#

There are a few ways to get the voltage. The LM9022 chip is designed to create both the filament and drive voltages, or you can use a voltage multiplier, step-up transformer, boost converter, or pre-built power supply module

#

30V is plenty for most of these

limber thunder
#

the vfd wants like 45

supple pollen
#

These days, a boost converter is probably the easiest approach

#

The board with the microwave display in the picture has a handy little VFD power supply module under it, that I got from a surplus electronics vendor. It provides both the center tapped AC voltage for the filament and the higher voltage for the grids and anodes.

limber thunder
supple pollen
limber thunder
#

idk what ic to use

#

the boost converters im seeing probably dont do 4.5v to 45v

supple pollen
#

It's a big ratio, the popular choices are the ancient 555 timer, MAX1771, and MC34063

limber thunder
supple pollen
limber thunder
#

the mc34063 looks like itll work

#

so any input voltage can result in any output voltage on there or are there limits to the stepping up

supple pollen
#

There are limits, but getting 45 or even 60 volts from a lithium cell is fairly straightforward

limber thunder
#

can i do 40v on the vfd?

supple pollen
#

Personally I like the LT107x and LT117x chips, but there isn't as much info out there on using them, as they work differently and are more expensive.

Screenshot_2024-06-24_at_8.42.54_AM.png
#

40V should work. For testing, I'll often just use three cheap 9V batteries in series for 27V. My preferred source for cheap 9V batteries is a dollar store.

limber thunder
#

how much current would lion voltage to 45v draw

supple pollen
#

It depends on how much current is drawn at 45V

limber thunder
#

30ma it looks like

supple pollen
#

Basically it'll be about 11 times the draw, so if you're drawing 5mA at 45V, you'd probably be drawing 60mA or so at 3.7V

limber thunder
supple pollen
#

30mA out would be 350 or so mA in

limber thunder
#

i might need a different battery

supple pollen
#

It's the law of conservation of power

limber thunder
#

mine can do like max 3a out safely

#

the cellular chip peaks like 2a but ill probably only be doing up to 1a max because no high speed data transfers needed

supple pollen
#

30mA at 45 volts is 1.35 watts: no matter what you do, you'll need to start with at least that much power (generally a little more due to circuit losses)

limber thunder
supple pollen
#

No, that's a reasonable load for battery operation

limber thunder
#

and then with everything else on top of it?

supple pollen
#

I'm not sure which lithium cell you have in mind, but lots of them are capable of that kind of power delivery

limber thunder
#

it does 3a maximum current draw

supple pollen
#

Should be sufficient

limber thunder
#

will the vfd even look unique if it isnt a character display

#

probably right?

supple pollen
#

Yup.

limber thunder
#

@supple pollen how do i set the voltage of the lt1070

#

they have examples on certain voltages but none are what im looking for

supple pollen
#

In the diagram above, the resistors to the feedback pin form a voltage divider. The idea is to adjust the ratio so when the output voltage is the voltage you want, the feedback pin voltage will be 1.24V

#

The equation you want (and the explanation of how to choose resistor values) is on page AN19-18, equation 21.

limber thunder
#

what unit is this? ohms or kohms

image.png
supple pollen
#

R2 is 1.24kΩ, and R1 is calculated from it. Since it's a ratio, it will work either way.

#

That looks right, that would make R1 43.6kΩ

#

I normally do all my calculations in basic units to avoid such confusion

limber thunder
#

i can do 43.7 but not .6

#

that's close enough

#

does it matter what power the resistor iss?

supple pollen
#

Only about 1mA is flowing through it, so power dissipation shouldn't be much of an issue. 45V times .001A is only 0.045 watts.

#

The VFD isn't going to care if it's 44V or 46V, so the tiny error in the calculated resistor value isn't going to matter.

limber thunder
#

it looks like the inductor should be approx 150uh, same as what the doc says

#

i find that weird because im using very different values but whatevs

supple pollen
#

It looks like a lot of the terms more or less cancel out in your use case, so it makes sense that the value doesn't vary much.

limber thunder
supple pollen
#

That seems suitable to me

limber thunder
#

muahahaah all your high voltage are belong to me

image.png
#

(at least in theory)

#

not sure abt this tho

image.png