help-with-hw-design

[limpid nest]

oh there's the signal

ok, it's measuring at 1.93 volts

between pins 1 and 2 here

regardless of the 595 state

[worldly schooner]

What about your return and 3V3

[limpid nest]

I don't have 3v3 picked out anywhere that I could attach a signal probe to

the return isn't technically ground right? Since it sinks into the 7407?

[worldly schooner]

It should sink to wherever the 7407 connects to GND.

Assuming the appropriate signal is high, anyhow.

[limpid nest]

low in this case, it's an active low chip

[median wind]

ok, now that im implementing the actual circuit I can see I made a dumb mistake, I converted A'B'C' to A NOR B NOR C, but I was treating that on paper as if it were A NOR (B NOR C) which is not the case

[worldly schooner]

A NOR (B NOR C)' should work. The number of gates matched up, so I thought you had it figured out already haha

[median wind]

indeed that does, but I'm short one gate, hmm, I'll just check all my expressions again

oh wait,t his should be the same as (B nand C)' right? I already have that term

wait, more like B' nand C'

[worldly schooner]

FW = A’B’C’ = A NOR [(B NOR C) NAND (B NOR C)]
FX = B’C + BC’ = (B NOR C) NOR (B NAND C)
FY = C
FZ = D’ = D NAND D
Ferror = AC + AB = (A NAND A) NOR (B NOR C)

Total should be 4 NORs and 4 NANDs, double check it please.

[median wind]

I think there should be an extra NAND in FX to invert (B NAND C)

F(B’C + BC’ ) = 1,2. Minterms
F((B NOR C) NOR (B NAND C) ) = 3. Minterms

unless my calculation is off

[worldly schooner]

Oh, you might be right

[limpid nest]

why might my 5V signal be sagging to ~1.93V by the time it gets to my terminal block?

[unreal flax]

Best guess is that the current-limiting resistor on your LED is too small, so the output signal isn't powerful enough to drive the LED, and the voltage is sagging until the LED just barely turns on at a smaller current.

[limpid nest]

hmm

It worked on a breadboard though..

[unreal flax]

That would seem to indicate that there's some wiring difference between the original breadboard and your current setup.

[limpid nest]

hmm

shoot

[unreal flax]

It's a classic debugging situation. If something really should work, but it doesn't, then that's a sign that something you believe to be true about the system is not actually the case.

[limpid nest]

darn

mega darn

let me get the design up, if you feel like checking it out

pt 1

[unreal flax]

I mean, one simple thing to do is to disconnect the LED from the terminal block and see if it's outputting 5V in that case.

[limpid nest]

pt 2

[limpid nest]

I tried that, got 1.93V

[unreal flax]

Cool. My next step would be to put a weak pullup / pulldown resistor on the output, and see if it's being actively driven at all, or is just floating at 1.93V.

[limpid nest]

Let me see if I can figure out what you mean.

Is 1k OK for this? Does that count as a weak pull up?

And would I pull it up to the desired signal, 5V?

[unreal flax]

Yes and yes. And/or pull it down to ground, too.

[limpid nest]

interesting

I got some advice from a friend that I need to add pullups on the outputs of my 7407s, like so:

That doesn't seem to jive with what you are saying though

[unreal flax]

What I'm advising is not a modification of your real schematic, just a troubleshooting test to gather evidence about what's going on.

[limpid nest]

ah ok

any thoughts on the advice I was given?

[unreal flax]

No opinion offhand... I'd need to do a little more investigation of your design and the datasheets, but I don't think it would be the source of the problem you're seeing.

[limpid nest]

ok yeah makes sense. Shoot this is quite troublesome. Appreciate the input.

what could cause the voltage to be floating at 1.93V when it's literally connected by a trace to a pin providing 5V?

[unreal flax]

How confident are you that the pin is providing 5V?

[limpid nest]

very, I measured it directly. It's the 5V pin on pi

let me check again though

[unreal flax]

How confident are you that the trace is making electrical contact?

[limpid nest]

hmm, less now! How could I check that? continuity?

got the expected beeps when I touched the 5V pin to the terminal block output

reads as 0 Ohms

[unreal flax]

Cool. Just to double-check, when you're measuring voltages, are you using the same ground reference each time?

[limpid nest]

yeah

I used the return path as a ground but I can try using a true known ground

[unreal flax]

Please do. As I understand your circuit, the return path is another chip output, not really ground.

[limpid nest]

ah ok, let me do that real fast

iiiiinteresting

that gives me 5.2V

measured between the output and the board's ground

[unreal flax]

Great, that starts to make sense. So the return path is probably at 3.3V.

[limpid nest]

why would that be?

hm

[unreal flax]

5.2 - 1.9 = 3.3

[limpid nest]

ah no haha I got that, I meant what design mistake would lead to that?

I also tested all the connections between my parts on a bare board, they are all fine

fine = how I designed it, not necessarily correct

[unreal flax]

I'm not sure what would cause that offhand. Do you measure 3.3V there?

[limpid nest]

where's there in this case?

[unreal flax]

The return-path wire versus system ground.

[limpid nest]

let me see

yup almost exactly 3.3

[unreal flax]

Cool. Well, investigating why that is would probably be the next step. It seems to be the smoking gun to explain your earlier weird readings.

[limpid nest]

yeah, I just can't see an electrical reason why that would be? Let me see if I can dig any deeper than I already have

[unreal flax]

The first thing I'd check is whether the 7407 is accidentally being powered from 3.3V instead of 5V, and you're just seeing a logic high on the output. Closely related, I'd check the input voltage for that buffer, to see whether it's being asked to output high or low.

[limpid nest]

hmm would that be in my schematic or are those physical tests I can do?

let me solder some jumpers onto the power pin of 7407

[unreal flax]

Physical tests, just measuring voltages.

You should be able to just poke the pin on the chip with your multimeter.

[limpid nest]

I'm too underfed and they're too close together, hands are shaky

[unreal flax]

Heh heh, understood. You might also be able to follow the traces from the pins to an easier-to-access spot on the board.

[limpid nest]

I'm not as skilled as madbodger, but I am madly bodging, so I'm there in spirit

[limpid nest]

OK! It is at 3.3V, but not on my schematic

weird, so are the transistors inside the chip just not fully "turned on" because it's at the wrong voltage level?

another weird thing: This worked for about a second when I went to re-seat the pi hat after it sagged a bit

at the beginning of my setup it did.

[unreal flax]

Any chance the hat is mounted off by one pin row so the 3.3V on the Pi is connected to the 5V pin on the hat or something like that?

[limpid nest]

I don't think so

But I have to go home (OT not allowed) so I can't answer questions about the physical setup

Thanks so much for your help. How would you proceed next?

[supple pollen]

It sounds like a bad connection plus a short circuit to me

[limpid nest]

Hmm bad connection between the pi and the hat?

Would that explain why I'm getting 3.3V at the power pin of the 7407?

[distant raven]

No, that likely means you’re using the 3.3V power pin

[limpid nest]

Huh

I'm away from the system right now but I'll double check

[distant raven]

Though something odd I encountered with an Arduino Nano

If I powered via USB, the 5V pin never was 5V

It was 3.68 or 3.75V

But it should have been 5V

If I powered by 5V externally, that wasn’t the issue.

So so strange

[limpid nest]

Huh

[limpid nest]

I am 99% sure that I seated the 2x20 pin header on the pi right, I might be missing something obvious but from my schematic I am not powering the 7407 from 3v3 in design at least

[distant raven]

I’ve not designed a pi hat before

But I do know that the pins are pretty close for 5V and 3.3V

[limpid nest]

They are yeah, idk I'll go over the schematic vs pinout.xyz again

[tough matrix]

what do you use to dispense solder paste when using stencils?
of course there are proper squeegees specifically designed for that, but I guess most people use old credit card or something like that .. what works best for you?

(had to look up proper spelling of "squeegee")

[supple pollen]

An ordinary putty knife can work, even the cheap plastic ones.

[limpid nest]

I checked and double checked my pinout. I am really sure it's right. I got some advice to add pullups to the outputs of my 7407. Does that seem reasonable?

[round rivet]

how do I get a differential impedance of 100 ohm for MIPI DSI?

[unreal flax]

That'll come from the trace width and spacing you use on your board. The second scenario in this calculator would cover the differential case, I think. https://www.pcbway.com/pcb_prototype/impedance_calculator.html

[limpid nest]

is it possible voltage is sagging due to the limits of the pi?

[worldly schooner]

It would more likely be the limits of the power supply, not the pi itself. 1.9V doesn't sound like a sagging voltage, though.

[limpid nest]

what does it sound like?

[worldly schooner]

Something connected to something it shouldn't be?

It feels like the ground path connected to 3v3 somehow...

[limpid nest]

yeah I'm tearing my hair out over here trying to understand

[round rivet]

Something like this? Ka-ro recommends 85-115 ohm, JLCPCB can do 0.127mm spacing. Also looking for he ZDiff measurement in the case of differential signalling?

[limpid nest]

so there's a small chance that I either got a bad chip from digikey, or somehow fried a good chip. Hopefully that's the issue, because there don't seem to be any differences between my prototype and the PCB

[twilit mango]

Oh bleh. I need Eagle help. I've done this before and can't remember how to do it. Nevermind! Of course I figure it out as soon as I'm about to ask.

Took notes this time.

[limpid nest]

what could cause the output of my 7407 to not work as expected? A short somewhere? I was careful to check over my work

[supple pollen]

Doesn't that chip have open collector outputs?

[limpid nest]

yeah

[supple pollen]

Then it won't drive its outputs high

[limpid nest]

hm

but one of them was high, which is what's weird

let me re-do the test and I can be confident of the values I see

[supple pollen]

Huh,yeah. It normally can only pull them low

[limpid nest]

yeah it would be "high Z" normally right?

[supple pollen]

Kinda

[limpid nest]

what would a hi z output state read when measured between the output pin and system ground?

I got 3.3V there

but only on one of the outputs, which is weird

[supple pollen]

It's, not really driven to any particular voltage so it's undefined

[limpid nest]

gm

hm*

[supple pollen]

You can use pull-up resistors to set a default level but in most cases, you'd want a different chip

[limpid nest]

really? I'm just trying to sink current into it

It seemed like a great choice

[twilit mango]

(up arrow in Discord will pop open your last message for quick editing, and then enter saves it. no clicking needed.)

[limpid nest]

ah I knew that, I was just being lazy

[supple pollen]

Yes it will sink but not

[twilit mango]

Fair enough!

[supple pollen]

Give a voltage

[limpid nest]

ah ok

I just want it to sink current, that's kinda what's it for from my reading at least. Am I wrong?

[supple pollen]

You can't really use a voltmeter ti check it unless it has a load connected

[limpid nest]

ok interesting

[twilit mango]

Gnarly mess.

(Have to flip parts to create Fritzing objects. Fritzing doesn't care that you destroyed the routing, it only cares about the pads you tell it to.)

[distant raven]

Lol

Fritzing:

https://tenor.com/view/april-fools-joke-dog-its-fine-this-is-not-gif-16757454

[twilit mango]

Entirely.

I had to then flip the part in the SVG to get it right, and as long as assigned the pads properly to the right parts of the image, Fritzing didn't even care that the original pads were in the opposite order. 😄

[brisk hearth]

Never use Fritzing for PCBs

LOL

I did sometime

[twilit mango]

I'm not using it for PCBs, I'm creating the Fritzing objects you see in all of our Learn guide wiring diagrams.

[brisk hearth]

Oh, cool

[twilit mango]

That's a word 😆

It takes the Eagle file and makes it into an SVG, basically, and Fritzing objects are packages of SVG files.

[brisk hearth]

I just thought, since it was in #help-with-hw-design...

[twilit mango]

Well, it was because I initially needed help with getting Eagle to do what I needed it to

so I posted the thing I figured out that I didn't end up needing help with.

[brisk hearth]

Aha

I used it for one DRV8833 board once, and never again

[twilit mango]

I do have an Eagle question that I can't find an answer to. Can you flip a group horizontally? Not the whole board, but a specific component or group of components.... I can rotate which (obvs) rotates, and I can mirror which flips it to the opposite side of the board, but I couldn't find anything about flipping on a horizontal or vertical plane. Is that not a thing?

[distant raven]

Uhhh, I don’t think so?

I know you can flip it for editing

But it doesn’t actually change the absolute orientation

[twilit mango]

Right.

Ok, so I didn't miss something, and my google fu didn't fail me.

I was considering flipping the part properly in the Eagle file, but did it in Illustrator instead. For Fritzing, it worked out in the end, but for something else, it might not have.

[limpid nest]

I need to decouple an IC input with a 1 uF cap in ambient normal temperatures (max probably around 95F, min in the 40s), is X7R acceptable?

[unreal flax]

Yeah, that should be fairly standard.

[limpid nest]

cool thanks

[daring night]

Finally making the jump from KiCad to Eagle after a day of crashes while just trying to get a prototype sketched out 😂 I only learned KiCad and general board design about six months ago so it shouldn't be too painful, hopefully.

The thing I'm failing to Google properly is just basic library acquisition. I've found the Adafruit guides for creating library items, and I found the page for the Adafruit-Eagle-Library but most of these things look pretty out of date.

Specifically I'm hoping to find a symbol/footprint for the Trinket M0. The Adafruit library is pretty dang cool and has a Pro Trinket symbol/footprint but not the 3v, 5v, or M0 Trinkets. I've also found the PCB files for the Trinket M0 (hooray for true open source hardware! ⚙️) but it looks like a lot of work to just dumb that down to a symbol/footprint. 😦

Does anyone have recommendations for a good source for things like this? Or should I just stop looking and make my own symbol/footprint?

[twilit mango]

Can you get it from the board files for the Trinkets? (Noob here, not sure what the difference between the files and the symbol/footprint really is.)

[daring night]

Yeah, I've grabbed the board files and they're awesome to have. But in this case I'm using a Trinket in a project as a drop-in, socketed module.

Its not crazy time intensive to just make a schematic symbol and board footprint for it from the original project files but I was hoping to save time if someone's already done the work.

[twilit mango]

Ah fair enough.

Wait... thinks

[daring night]

I might not be using the right words too, still new to this space 😛

[twilit mango]

What about the PyRuler?

I still think I'm grasping at the wrong thing here.

But it has a "fake" Trinket on it.

[daring night]

I totally forgot about the PyRuler! I'll grab the schematics and see if its laid out as a module or as discrete components

[twilit mango]

Looks like components unfortunately... but I could be wrong.

Anyway, I searched for "trinket" in "add part" (if that's even right) and didn't find anything in my Eagle. And I have the more up to date version of Adafruit's stuff.

[daring night]

I could just skip ahead in my learning and wrestle with SMD. My little miniware hotplate is probably getting lonely anyway

I've enjoyed prototyping with the Trinket form factor, really handy for smaller projects. If I end up cooking up a library item for myself I'll throw it in a github repo or something

[twilit mango]

Nice.

[median wind]

so, I normally don't like asking before I have a good idea of what I need to do, but in this case I think I need a little kick to get going. we have 4 inputs,one representing someone from a company where they have 1,2,3,4 shares respectively, the inputs represents their vote which is worth as much as their shares. we use a 7 segment display to represents the vote (0 for 10 votes, nothing for 0 votes). We have to use four 74151 mutiplexer (8 to 1), but I honestly don't have an idea how to work with this

im having trouble figuring out why I need multiplexers here 🤔

[unreal flax]

It sounds like all of the vote addition logic would need to be done via the multiplexers, unless you have some other circuit components available.

[median wind]

so I guess I'd need to use the selectors hmmm

honestly I feel like I'd have an easier tie with normal gates and no multiplexers 😅

[unreal flax]

Yep, you probably would. I assume it's a homework exercise rather than a real design problem, so the limitations are no doubt imposed just for the sake of learning.

[median wind]

yeah, that's exactly right

[frigid seal]

Hey, How do you all handle DNPs on your boards from the schematic perspective?
I normally remove the comment and write DNP but I'd like to preserve the design intent..

[worldly schooner]

Usually I just leave a text note next to the component in question, and exclude it from the BOM.

[median wind]

ok, so I arranged a table with the inputs and what I believe should be my output., the problem is getting there. I need a decoder (what will received the so caleed output) a 7 segments display, and four 8 to 1 multiplexers.

so each one of the inputs represents an actionist where each has a different value attached to their vote, 4, 3, 2 and 1 respectively, and their vote has to be represented with the display, the "oddities" is that 10 votes (if they all vote) would be represented by a "0", and if nobody votes there's no outut (why 0000 is 1x1x and 1111 ends as 0000)

im tryign to figure this out manually but there's got to be a better way

[median wind]

for example, I think this should serve to give my least significant bit for the output

if D (worth 1) and B(worth 3) are both 0, or both 1, then the output is 1, but then it gets more complicated with the following bits. But ill keep working on them one at a time

actually thinking about this, I think I can just use karnaugh with this to find out what I need out of each bit

yeah, I think this might work

ugh, the least significant bit needed just 2 variables, but the next one needs 4 variables and i only have 3 selectors

[unreal flax]

Why 4 variables? I think the "4" vote wouldn't matter one way or another for the second bit, right?

[median wind]

im actually changing things arround a bit, rather than working with don't cares for the output of 0000, I've decided to just set them as 1111, now the multiplexer for the least significant is like this:

ill revisit the second bit now

I0,I1,I2,I5,I6 = 1, rest = 0

so uhm, A should only affect it sometimes when it's 0

[unreal flax]

Ah, right, sorry, I didn't catch the "10 = 0" part of the spec.

[median wind]

gaffe, A+B is supposed to be A+C

the problem is I'm getting this through uhh intuition you could call it, and not from the expression I get from the karnaugh map simplification

(the karnaugh map for that image above is f(a, b, c, d) = b'd + bd' + a'b'c' )

[hollow steeple]

I've designed a custom board using an ESP32-WROVER-E module. I'm using a CP2104 breakout from adafruit to upload to it. The first program uploaded perfectly, but I can't upload any new programs to it now, it always fails to connect. It runs the original program (just cycling colors on an onboard neopixel) and outputs start information to serial, but fails to connect. I have seen a lot of mention of adding a capacitor, but I believe my design incorporates that (I based the autoreset portion off of the Adafruit esp32 feather). Any pointers on what area I should be exploring more to get to the bottom?

(I've also tried holding down the reset button which doesn't help, but that also warms the board and turns off the power LED which doesn't seem right)

[unreal flax]

Would it be possible for you to post a schematic showing the reset / serial interface portion of your board?

[distant raven]

Are you crossing the UART lines by chance?

[hollow steeple]

No, I get serial output from the ESP32 on boot

[distant raven]

Okay, just curious

There a running joke about no one getting UART right the first time

[hollow steeple]

Yeah it was one if the first things I double checked 🙂

[distant raven]

Okay good

[unreal flax]

I'd be investigating the soldering of R1. It sounds like the reset button is shorting out the power supply instead of just having a weak pullup.

[hollow steeple]

Let's see!

R1 and C17 both seem ok to me

[unreal flax]

If you have a multimeter, checking the 3.3V rail when the button is pressed would be my next move.

[hollow steeple]

Would just checking across ground and the output pin of the regulator work?

[unreal flax]

Yep, that should be fine.

[hollow steeple]

It drops to nothing

.. and my computer just gave me a warning about a usb device using too much power (I'm on a different computer than normal so I didn't see it before.)

So I guess something is shorting when its closed?

Continuity says yes, 3.3 and gnd close when pressed

So now I just have to figure out if it's soldering or on the pcb

The schematic looks ok on that front I'm guessing?

[unreal flax]

The schematic looks good, and normally your CAD software would have flagged a layout error causing a short in the ERC/DRC.

[hollow steeple]

I don't know that EasyEDA has something fancy like that 🙂

[unreal flax]

Ah, I'm unfamiliar with EasyEDA, so I couldn't say.

[hollow steeple]

I did connect RST to some other chips so that seems like the next place to look

Aha, I checked the connections for R1, I did not check R1 itself.. its shorted

Thanks so much for your help

Frankly things are going much better than expected given the complexity of this board combined with it only being my third

[unreal flax]

Yep, you're ahead of the game... so many possible things to go wrong. My last major board wouldn't even power up at first, heh heh. 😅

[hollow steeple]

The one I did before this I did the pinout to a boost converter that wasn't in the library backwards

I actually did it right the first time and then I moved it to the back of the board to save space and later got confused about it.. ended up "fixing" it

I'd consider this my first major board, prior ones had part counts around 10, this one is around 100

[median wind]

so im like 99% sure my inputs and data selectors for that multiplexer above are right (well, besides writing A+D instead of A+C), the problem is I don't know how to get there from anything other than intuition, and intuition hasn't helped me for the next multiplexer 😫 . My attempt on seeing the relation between the simplified expression (from karnaugh) and what I ended up with hasn't been fruitful

[hollow steeple]

Huh, the pads on the board are shorted somehow. Not on an empty board though at least.

[round rivet]

someone know of a similar PCB footprint that could fit something like this?

[unreal flax]

I don't expect that pinout is standardized, so it would generally need to be a PCB designed to fit that module.

[round rivet]

That is what I am going to do, I need a similar stamp module footprint that I could use as a basis

[worldly schooner]

It’s a castellated pad, so the pads shouldn’t be too strict. As long as you have pitch and width for the pads, you should have an easy time with it.

If you do need something just for reference, you could look at the pads for soldering something like the pi pico or the rp2040 stamp?

[supple pollen]

I wrote a footprint generator a while back for stuff like that.

[median wind]

Well i think i got a way to find out how to do the mapping, the issue is that it's illegal as per the instructions

Basically the instructions say to not use 5v for logic ones, instead to use inverted ground (though i imagine this applies to the multiplexer input only and not the selectors)

This is what I did to find how to wire it, with this thing that resembles but isn't actually a karnaugh map

You reckon this might fly 👀 ?

In a way I'm passing the inverted ground ...

[median wind]

Ok, it seems i can feed it the input, success! Though i still would like to know if there's a way to find out to wire up the multiplexer the same way i did the first one

I'm assuming there had to be some kind of formula

[hollow steeple]

@unreal flax thanks again! the short was underneath the esp32 module itself. reflowed it and now everything is working perfectly

[unreal flax]

Nice troubleshooting! 👍

[limpid nest]

If I want to run shift registers at 5V but talk to them over 3.3V do I only need unidirectional level shifting?

[distant raven]

Unidirectional should be fine

How fast do you plan on clocking?

If not terribly fast a BSS138 mosfet would probably work fine

[limpid nest]

Not terribly but I'd rather not have any extra sources of error. And I'd need 3 mosfets which take up more space than a small ic

Thx

[distant raven]

I mean, they are sot23-3 so pretty small

[limpid nest]

Yeah

I'd be changing the state of the shift register at less than 1 Hz

[distant raven]

Get a resistor array and save a little more space

There are small level shifters but they usually have max low side voltages of like 2.65V

(I went looking last week)

Which is sad

[limpid nest]

Transistor array?

[sinful depot]

For level shifting you could use something like the TXB0104. You can use 3.3v on one side and 5v on the other

It has 4 level shifters in the package

[limpid nest]

That's a nice one, thanks

[heavy jasper]

Another low part count solution if you’re clocking slowly: just pull-ups to 5V on each of the outputs and configure your microcontroller output buffers as open-drain.

But this relies on you using an MCU with five-volt tolerant pins (despite 3.3V digital supply)

(Some are, some aren’t)

[limpid nest]

I'm using a raspi, if that helps

I have a good bit of room on the board so idk why I'm stressing, I haven't even had to place parts on the bottom

[heavy jasper]

I don’t believe the raspberry pie has 5V tolerant IO; more common on the rp2040, stm32, and those types

[limpid nest]

Ah thanks thay was my understanding as well. I'll just use either a 4 channel level shifter or a few mosfets

[distant raven]

Pico doesn’t have any 5V tolerant pins

[median wind]

this is a decoder, does anyone know what "test" "RBI" and "RBO" are for?

[supple pollen]

For a reliable answer, look up the data sheet. From memory, "test" is "lamp test" (all outputs on), RBI is "ripple blanking in", and RBO is "ripple blanking out". Ripple blanking is used to elide leading zeroes.

[median wind]

thank you!

[limpid nest]

"Elide leading zeroes" what does that mean in this context?

[supple pollen]

If you had a row of these chips displaying the number "0034", you could tell them to blank the leading zeros, so the first two digits didn't show, and it would just display " 34"

Basically you use the RBI and RBO pins to daisy chain the information "all the digits to the left of me are zeroes", so if that's true and the chip has a zero on its inputs, it will disable all its outputs and pass that "true" on via its RBO pin to the next chip.

[limpid nest]

Ahhh ok thx

[supple pollen]

Again, this is from memory and I could have some details wrong. Refer to datasheet for reliable information, please drive gently, do not litter.

[limpid nest]

A guy who I consult with is big on using decoders over shift registers. Are decoders an old school way of expanding GPIO?

[supple pollen]

Just a different way. A decoder lets you select one output at a time. A shift register can enable any pattern of outputs.

[limpid nest]

Ah ok

[supple pollen]

Decoders are also fast, straight through chips, so they're useful for selecting things like memory banks, output drivers, etc.

[limpid nest]

So am I misunderstanding this "a decoder let's you select one output at a time"? It seems like you could set the output pins to whatever pattern you wanted?

[supple pollen]

No, each input pattern only enables one output. So if you put "0010" in the input, it would enable output B, if you put "0111", it would enable output G.

[limpid nest]

Ahhhh ok

So decoders are good when you 1)need speed 2) don't need to place an arbitrary pattern?

[supple pollen]

Right. Also, when you want to avoid selecting two outputs at once by mistake (like the memory bank example above)

[limpid nest]

Ahhh ok

Cool! Thanks

[supple pollen]

Here's one typical use, with one as a display multiplexor (for a KIM-1 computer), that lights one digit at a time, but sequences between them quickly so they all appear to be lit.

[limpid nest]

Ah very cool so almost like pwm?

[supple pollen]

I'm not sure about that, I suppose it depends on how you think about it.

[limpid nest]

Like it's turning on and off so fast that to the human eye it seems like it's on? Sortof PWMing your eyeball

[supple pollen]

Yeah, in that sense, yes.

[limpid nest]

Just reminded me of how at least one model of hoverboard was open loop controlled. The human provided the PID system heh

[supple pollen]

I heard about a powered wheelchair design that also took the "put a human in the loop" approach, to good effect.

[limpid nest]

Interesting

A dream of mine is to work on powered assistive devices but I'd have to move from an area I love

[limpid nest]

deleted and moved here as it's closer to PCB design:

Can one of these (https://www.digikey.com/en/products/detail/te-connectivity-amp-connectors/1-1776296-2/2038206) be used like one of these in the attached picture?

I'm asking because the TE-connectivity connector on digikey seems to have a nice small physical size

and space is at a premium in my enclosure

it seems like they can be? Anyone have a differing opinion?

[supple pollen]

Yes, those "European style" barrier strips work nicely for that sort of duty, and with less exposed conductors as a bonus.

[limpid nest]

nice ok I'll grab some. Also have some Qs about choosing shrink tubing parameters, let me get sorted on that and I'll BRB

so I think I need 3/16ths inch but I'm going to get an equal amount of 1/4" just in case, but how do I choose between parameters like Inner Diameter - Supplied vs Inner Diameter - Recovered?

I am assuming that recovered is after shrinking?

also I need to bodge some caps onto my design. I bought some thru hole ones for this purpose but a friend suggested soldering one terminal of a SMT cap directly to the power pin and then using wire to connect to ground. Which sounds smarter?

[limpid nest]

it seems like for ferrules, DIN, vs W, vs T comes down to color of the insulation only?

[supple pollen]

I like leaded caps for that sort of thing (I used to just wire wrap the leads around my regulator pins for really local capacitance). And yes, I think "inner diameter - recovered" is after shrinking.

[limpid nest]

leaded like with leads (pronounced leeds)?

[supple pollen]

Yes, that's what I was trying to say, apologies for the ambiguity

[limpid nest]

no worries, I just thought there was a kind of capacitor with Pb involved

that I hadn't heard of

[wild glade]

What's good free pcb/schematic design software?

[unreal flax]

Kicad is generally considered the best free alternative.

[wild glade]

I'm just sad Eagle got so expensive

So, now I want to get back into the hobby, time to go free

Thanks for the advice, Ed ❤️

[unreal flax]

If you have the need for mechanical CAD software too, Eagle is included in the Fusion 360 license, BTW.

[wild glade]

I know, but I'm a Solidworks user

Well, Idk for how long, since they stopped the 40$ per year EAA license thing..

Fusion360 does too much in the cloud, I like to have all my stuff locally

And not have to spend tokens to do CFD

[unreal flax]

Solidworks does have a maker 3D Experience license now, but I'm not sure how limited that is.

[wild glade]

Solidworks shoveled their cloud version into the maker license as well, sadly

KiCAD looks promising however

Looks like a lot of EAGLE folks migrated since the last time I used it

[unreal flax]

Yeah, it's improved a lot in recent years. The v6 release is coming soon, too, so a lot of people are using the 5.99 daily builds instead of the older stable release.

[tough matrix]

6.0rc1 is expected to be released today

[velvet patio]

Kicad's workflow is quite different from Eagle. So much that I might pay for Eagle just to avoid having to learn it.

[frigid seal]

I promise you, kicad ends up being better.

And they fixed a lot of the terrible workflow in 4.0

[limpid nest]

what's the difference (on digikey) between m2.5x.45 and plain m2.5?

[worldly schooner]

For what? One specifies an extra dimension, but without context, can't say for sure.

[limpid nest]

for mounting a PCB. It's moot actually they don't have any "plain" m2.5s in stock

[worldly schooner]

Length of hole?

[limpid nest]

I would assume it has to do with pitch? but I'd just never seen it that way before

[worldly schooner]

I still have no idea what product you're trying to filter. What are you searching for, a panel or screws?

[wild glade]

Welp, time to make a ionizer driver circuit, a microphone preamp, and some other doohickeys

[limpid nest]

sorry, I was doing the thing where I have a conversation in my head and assume you're in on it.

huh, why did my link get censored?

[wild glade]

No links allowed I think?

[worldly schooner]

Certain links are filtered out, though I don't know the criteria.

[limpid nest]

huh, it's just digikey. I know if you have three specific letters in a row the bot will catch it

let me check the link

[worldly schooner]

From what I did see, you posted a super long digikey link that probably looked suspicious.

[limpid nest]

I've def posted links that long before

[worldly schooner]

A description should suffice.

[limpid nest]

I'm here filtered by machine screw m2.5x.45

[wild glade]

whut

[limpid nest]

did you get dinged too?

[wild glade]

Not even a link

What, you're not allowed to say " d a m n "

[limpid nest]

ah yeah

[worldly schooner]

Yeah, it's just thread spacing

[limpid nest]

G rated here

[wild glade]

Thank god for my buddy Clive that depotted those ionizers

[wild glade]

Christian server

[limpid nest]

actually non-religious as far as I can tell. It's just supposed to be a place kids can come and not see adult only content

[wild glade]

I know, it was a joke

[limpid nest]

ah sorry, I can't always catch those over text

[wild glade]

Either way, back on topic 😬

Ionizer circuits are quite interesting

Context: I'm DIY'ing an air cleaner

Which should turn out to be much much cheaper than buying one

[limpid nest]

not rated by UL tho heh

[wild glade]

It's all just a fan with a filter lol

All I'm doing is 3D printing some parts and making a proper electronic circuit for it all

Gonna custom design a PCB that takes a 'duino IC

[tough matrix]

https://twitter.com/KiProEDA/status/1460311298273546246?s=20

KiCad Services Corporation (@KiProEDA)

True fact.*

*tagged today. Will take a day or so for releases on various platforms https://t.co/uKnad8OhLc

Twitter · Nov 15, 2021, 6:18 PM

This is about KiCad 6.0 rc1 (release candidate)

[sinful depot]

@limpid nest the .45 on a 2.5mm thread is course pitch, .35 would be fine pitch

[limpid nest]

I figured it was the pitch, was just confused as to why it's called out in one option but not another. Who knows

Thanks

[elder peak]

Oh boy.

This is going to be exciting, trying to move to 6.0.

[tough matrix]

latest nightly, posted today, is in fact KiCad 6.0 RC1

[elder peak]

I don't have the patience necessarily to play with nightlies of most anything unless someone's paying me.

So, yeah, this'll be interesting just because it's probably got a bunch of things that I really really want to mess with except that also I haven't been playing with the nightlies so I'm going to be all WHO MOVED MY CHEESE.

[bright thistle]

laughs in -git

[limpid nest]

simple circuits question. When I have a pin configured as an input with internal pullups enabled, what happens when I short the pin to ground? Does current flow thru the pullup to ground? That's how I've always understood it

[worldly schooner]

Yes.

[tough matrix]

Yes.
Internal pullups are usually rather high resistance (something like 50kOhm, iirc), so the current will be very small

[compact portal]

yep

[limpid nest]

Thanks friends

[velvet patio]

8 days order to delivery on my most recent JLCPCB order, using the slowest, cheapest option possible... They must not want to sell any express service.

[limpid nest]

interesting. I used the fastest shipping because I needed it at the time, but good to know that cheaper options are OK. They must be flying stuff in

[velvet patio]

$2.18 total for the order

Granted this was a reorder, but that can't cut more than 1 day off the amount of time necessary.

[limpid nest]

China's PCB subsidies are anti competitive but dang if they don't benefit me immensely

[velvet patio]

Well, the entire process seems exquisitely automated. I'd wager the only time the boards actually get touched by human hands is when they're depanelled and packed.

[limpid nest]

probs yeah

[tough matrix]

what is the optimal layer arrangement for 4-layer boards that need to dissipate heat from chips with thermal pads?

usual layer order is signal, GND, power, signal2

but wouldn't it be more efficient to place GND plane on the bottom layer so it can dissipate heat better?

obviously, the pad will be connected by thermal vias to GND plane

[worldly schooner]

Strictly for heat dissipation, sure. 4-layer boards are usually used for routing complex signals with stricter requirements, such as length-matched high-speed signals or impedance-matched RF traces. Putting the GND layer on the bottom means your largest stable surface no longer acts as a shielding layer for protecting the sensitive signals on your two signal layers from interfering with each other.

That being said, if you have no such signals crossing, I'd be interested in its thermal dissipation efficacy.

[tough matrix]

in this case, there will be no RF or other sensitive signals

the only reason to choose 4 layer is to have uninterrupted ground plane for heat dissipation

does it make sense?

[worldly schooner]

It does, to some extent. How complex a board is it? Is it feasible to consider an aluminum-backed board instead?

[tough matrix]

interesting idea
let me check

they are 1-layer only... might be difficult

[worldly schooner]

I mean, multi-layer aluminum boards do exist, though I doubt JLCPCB offers that capability.

I see several manufacturers that offer 2-layer aluminum boards, but I can't see the price offhand.

Shipping alone probably makes the metal alternative pricier than you'd like...

[hexed wren]

hello, I would like to design in KiCad a project that uses a Nordic nRF51822 SoC, but after googling some and digging in forums I'm not sure about the availability of nRF51 / nRF52 reference designs for KiCad. Can anybody here point me to updated resources on this? Thank you !

[unreal flax]

You'll generally only get manufacturer reference designs as PDF schematics rather than full CAD files, so you'd need to incorporate them into your own board design in most cases.

[hexed wren]

got it, thanks. In the meantime I've found that the Micro::Bit reference design, which is nRF51-based, is also available in KiCad format (https://github.com/microbit-foundation/microbit-reference-design) so that might be a great starting point.

[limpid nest]

For using ferrules, is this a decent length of wire to have stripped out?

Or do I want more?

[tough matrix]

you would probably want it about as long as the metal part of the ferrule or slightly shorter - it shouldn't stick out of ferrule

[limpid nest]

Ok thanks!

[tough matrix]

more importantly, you do need to use proper crimping tools - do not attempt to do it with pliers!

[limpid nest]

I bought a cheapo ferrule crimper kit. Seems to work fine, my buddy has the same one and it works fine

In case anyone is looking for one, the brand is "Sopoby", found on amazon

[limpid nest]

On a solid state relay, when I power the relay on, should a multimeter read a short between the AC load pins?

[gusty iron]

It should be 0.2ohm or less

[limpid nest]

thanks

[gusty iron]

well, it may depend on the type of SSR, you should check the data sheet

i just know ones at work that I use (very expensive ones by Opto) are 0.2ohm typically

[limpid nest]

let me see what the datasheet has to say, I should have gone there first.

[ember laurel]

How does one go about sourcing pretty much any buck converter or mosfet these days?

TI has pretty much 0 stock on everything, with 60w lead time

[distant raven]

What voltage and current max?

[ember laurel]

40 in, 5 out, 3A, 1MHz

[distant raven]

Ah… yeah..

Good luck

I found this unknown part in Digi-Key for 3.3V 700mA with a built in inductor

I found this part with the Digi-Key mobile app. https://www.digikey.com/product-detail/en/XCL224A333D2-G/893-1419-1-ND/8256121

[ember laurel]

Yeah that won’t fly

https://www.ti.com/product/TPS54340B

That would be ideal

I can also do with Silergy SY8303 or SY21153

[distant raven]

I found this part with the Digi-Key mobile app. https://www.digikey.com/product-detail/en/MPQ4313GRE-5-AEC1-Z/1589-MPQ4313GRE-5-AEC1-ZTR-ND/13684276

That’s 45V 3A

But not your switching frequency

470kHz

[ember laurel]

Yeah I’d prefer it a bit higher, otherwise the L gets so large.

I’d also want it synchronous.

I’ve used both the SY8303 and 21153 in prototyping, and they are very good.

That MPS part is very pricey too.

[distant raven]

Yeah, 5.6uH inductor

Not sure you could find a 3mm inductor that value. Cheaply anyway

[ember laurel]

World is just insane.

[distant raven]

Yeah

[ember laurel]

It’s like Starbucks has introduced a buck converter Frappuccino or something.

[distant raven]

I have 8 TPS62237 left and there are none in stock anywhere

[ember laurel]

And now they all get ground down for that purpose

[distant raven]

It’s a 3.3V 500mA buck

[ember laurel]

Yeah that’s a good one

[distant raven]

Super low Iq for a 3.3V buck

I have a was selling a 5V 500mA variant breakout if it

TPS62173 or something like that

But I can’t find that either

If you need a 5V 3.9A boost, there’s this: I found this part with the Digi-Key mobile app. https://www.digikey.com/product-detail/en/TPS61253YFFT/296-30503-1-ND/3451639

Lol

[ember laurel]

I wonder when this will end. They must be ramping up production?

[distant raven]

They just bought the micron fab down the road from me, but said late 2022 before they start making TI parts

[ember laurel]

I’m just about to run my first batch of 500 units here

And currently Several of power related components are missing

And mosfets

I have already managed to stock STM32’s and motor driver chips required

I expected it to become hard to get those

But I didn’t expect that Some regular mosfets or bucks would be an issue at all

70w lead time doesn’t sound great.

[distant raven]

Power mosfets?

Or regular mosfets

[ember laurel]

Power

[distant raven]

Oh yeah.. good luck

[tough matrix]

it is insane, but sometimes you can get lucky
I was recently able to score an Infineon motor driver chip

[distant raven]

LCSC sometimes has stuff I can’t find anywhere

Like a 3.3V ldo I regularly use

Which funny enough is a TI part

Digi-Key doesn’t have them, mouser doesn’t have them

But LCSC does

[limpid nest]

having a hard time finding a part that I assumed existed: a female to male audio jack/plug that is panel mount. Does that sound far fetched?

[distant raven]

Hey @ember laurel is this one you can use? https://lcsc.com/mobile/product-detail/DC-DC-Converters_Texas-Instruments-TPS54340DDAR_C45886.html

I’m not sure the difference between the B variant and D variant

Ah, 0.8V adjustable

Guess the B makes a big difference

Ranges from 0.8-42V out

Would require extra bits to make it 5V out

Hot dang! The load switch I’ve been after for months is on LCSC

$0.20 more than DigiKey

[limpid nest]

nice!

[tough matrix]

this is to allow connecting audio cable through enclosure wall?

[limpid nest]

yup

I found a female to female that isn't IP rated, the product engineers are gonna get back to me as to whether there is a hole all the way through or not. It doesn't seem there is. So I can source some O-rings and make it IP-ish

cutting an aux cable and soldering the wires to solder cups has turned out to be prohibitively fiddly and diffuclt

[tough matrix]

would XLR work? I expect panel mount XLR jacks are easier to find, and there are cheap audio jack to XLR adapters

[limpid nest]

XLR? Let me google that

I did just place an order but I'm not unused to wasting my money

[tough matrix]

XLR is larger size connector, typically used for microphone cables

[limpid nest]

ah ok let me see

you'd think there would be an IP rated way to pass audio through an enclosure but what do I know

[limpid nest]

This seems like a great option for my next revision. Thanks!

[sick geyser]

Pair of RCA might be another option

[limpid nest]

let me check those out

looks like the XLR ones are more likely to have some kind of ingress protection rating

[tough matrix]

there are some waterproof XLR

[limpid nest]

yeah I think next revision will use XLR female-female and then an XLR-Male adapter in the enclosure and an XLR-female adapter on the outside

[limpid nest]

this panel mount connector has 4 pins on the back, but three male pins on the front. What is the 4th pin for?

ah it seems it's for grounding to your panel

[tough matrix]

I have to share this marketing gem with you:

I need this microscope

[limpid nest]

It's not as bad as the ones where the model is holding the iron by the literal iron

[tough matrix]

true.
there is also a recurring theme where a scope is showing some waveforms while no probes are plugged in

[limpid nest]

Lol

[distant raven]

It’s picking up cosmic background microwave radiation 🙂

[ember laurel]

@tough matrix I have that microscope, but without the virtual lens teleportation feature.

@distant raven the TPS54340 is a great buck, but good luck finding it anywhere.

[distant raven]

It was on LCSC, the adjustable version

[ember laurel]

shows 0 in stock

https://octopart.com/search?q=TPS54340&currency=USD&specs=0

[distant raven]

And it’s gone

Lol

[ember laurel]

shows more on octo

yeah - that's how it tends to be

I've resorted to checking TI.COM every hour

[distant raven]

There was 1400 yesterday 😅

[ember laurel]

I managed to get 160 of my mosfets thee yesterday

https://www.ti.com/product/CSD88539ND

[distant raven]

Nice!

How much current can those sink?

[ember laurel]

did the same, every day 10x checking on TI.COM, to get 525 of the DRV8323S

[distant raven]

Geez, 15A

[ember laurel]

the TI mosfet can do 6.3A continuous

[distant raven]

15A peak is a lot

[ember laurel]

yes

[distant raven]

For instantaneous current, that’s pretty nice

[ember laurel]

other options there would be the DMTH6016-LSD13

[tough matrix]

nice
are you making a motor controller?

[ember laurel]

yes

another option is the https://lcsc.com/product-detail/MOSFETs_Wuxi-span-style-background-color-ff0-NCE-span-Power-Semiconductor-NCE6005AS_C126151.html

would be ok according to specs

but I had one of these fail inexplicably in a prototype unit.

[tough matrix]

I was looking for driver chips to get 4A continuous and was able to get some. But I need smaller numbers

[ember laurel]

I have about 10 units in the field with this NCE6005AS in it

and 10 units with the TI mosfet

[tough matrix]

are they pin compatible?

[ember laurel]

they both do the trick, the TI runs slightly cooler

yes - pin compatible, very similar specs

[tough matrix]

and you need 500 units?

[ember laurel]

I need 1500 plus spares

say 1600 ish

doing a production run of 500 units

[tough matrix]

part sourcing for that is nightmare nowadays

[ember laurel]

yeah ... not just that

I'm really glad I got the STM32H750 that I use

lcsc shows them in stock now though

but 2x more expensive than a year ago

I just read that Tesla is delivering cars with the wireless charging pad disabled

and without front usb ports

due to missing chips

[tough matrix]

BTW
somewhat unrelated question for you.
In dir/pwm control drivers, motor is freewheeling during off part of the cycle

it is claimed that this gives less precise control than with pwm*2 which has motor braking

is it really a significant difference?

[ember laurel]

what kind of motor are we talking about?

[tough matrix]

brushed dc

say 12v, 2A or so

e. g. this is from pololu site:
"This DRV8256E uses a phase/enable interface that allows bidirectional control with only one PWM signal, but it is limited to drive/coast operation. The very similar DRV8256P has an IN/IN interface instead, requiring two PWM signals for full bidirectional control, but it offers drive/brake operation (which usually provides a more linear relationship between PWM duty cycle and motor speed)."

[ember laurel]

I'm not really working with DC motors

[tough matrix]

oh, ok

[ember laurel]

but hmmmm

so what happens during the low part?

[tough matrix]

for dual pwm, motor is in regenerating braking

[ember laurel]

I know a good deal about BLDC/PMSM, but unfortunately not much about DC motors

I'd say, try it out 🙂

[limpid nest]

how can I tell, for this power supply (https://www.digikey.com/en/products/detail/cui-inc/PSK-S25D-5-T/9997570?s=N4IgTCBcDaIA4GcDWBaBYCsARFGAuIAugL5A) whether V_o(-) is truly at 0V with V_o(+) at 5V, or if they are just floating somewhere above 0V with a 5V difference between the pins?

[worldly schooner]

Well, voltage is inherently relative, so if you don't ground the Vo(-) terminal, you'll have a 5V difference of some floating voltage?

[limpid nest]

hmm that's what I was thinking but wasn't sure. Many supplies I looked at don't have an AC ground does this thing having an AC ground imply that V_o(-) is grounded?

[worldly schooner]

It is isolated, so there isn't any connection to the AC GND

[limpid nest]

ah

ok

Hmm, so would I need to physically connect AC ground to V_o(-)?

or could that be bad?

[worldly schooner]

I don't see a need to connect them, unless you're doing something else on the AC side of everything...?

[limpid nest]

I was mostly concerned with whether my sticker with arrows warning of hi voltage on both sides was accurate

[worldly schooner]

Isolated means there should not be any high voltage on the DC side.

[limpid nest]

interesting OK, I just wasn't sure. I think I'm the only person who's ever going to service this but I don't want a coworker to be injured

thanks!

Sanity check, short lead with the oval lines is negative on this aluminum cap?

hmm, the leads are only barely long enough on the DC side to reach between V_o(+) and V_o(-)?

oops that's not a question

Can I use the NC terminals and some wire to make the connection?

Another confusing thing I haven't been able to determine from the datasheet. They have a 0V ground symbol here, and it seems it's specifically not labelled as V_o(-). Does the trim resistor go to ac ground?

[limpid nest]

from left to right, the pins are V+, NC, TRIM, NC, V-. Can I use the left most NC pin as a bridge for the aluminum cap that doesn't reach all the way between + and -? I have a little insulated wire bridging the pins.

Also how would you insulate the wires of the caps? This should really be on a board and will in the future but I don't have time for one right now

Shrink tubing?

[gusty iron]

some 1mm heatshrink would work fine

[limpid nest]

yeah the 1mm I have is old or something and isn't shrinking. Short term solution before I can get this on a board is just electrical tape

[sick geyser]

The tube would still be fine even unshrunk

[limpid nest]

it didn't stay on

[sick geyser]

Sandwiched between the part and the screw terminal though?

[limpid nest]

maybe, I've already bodged it with electrical tape, so it's a sort of moot point. These need to be on a board, and will be in the future

what do you think about my bridge idea?

[tough matrix]

tbh, I'd just solder extra length of solid core wire to cap rather than do these bridges.

[limpid nest]

interesting ok, I'll see if I can get my tape off and try that

[limpid nest]

I am trying to find an analog to the 7407 chip, but with more outputs. I'm on digikey. What does "number of elements" refer to?

[heavy jasper]

Generally the "number of elements" is in a case where there is a repeated logic element inside the chip, see for instance this 75453:

For 1:1 buffers, the product of bits/element and element should be what you want, but there are some special cases like this one

FYI: if what you're looking for is just an open-drain load drive buffer, you may have better luck in the: "PMIC - Power Distribution Switches, Load Drivers" category

[limpid nest]

I think I want open collector?

I'm using it to sink current

[heavy jasper]

open-collector or open-drain

[limpid nest]

ah right, they are semi synonymous right?

[heavy jasper]

open-collector is bipolar transistor, open-drain is FET

[limpid nest]

ahh ok

why is "PMIC - Power Distribution Switches, Load Drivers" a better place?

[heavy jasper]

It's a more expanded set of parts that are less likely to fall into the traditional 7400 series form factors (which you likely don't really care about) but are specifically intended for use as load drivers

[limpid nest]

ahh neat ok

wow they have a 32 output one, that's perfect. I need to drive 25-26 loads

[heavy jasper]

I think that one may be obsolete

[limpid nest]

ah yeah they only have 2 available

and not even in stock

[heavy jasper]

If you need to drive 25-26 loads - I'd probably stick with 8-output chips

they're widely available

and just get 4 of them (and have a few spare channels!)

[limpid nest]

ok, would you stick with them because of their wide availabilty?

also how would i know whether to pick a P-channel output or an N-channel? I'm thinking low side switching.

[heavy jasper]

N-channel is generally used for low-side switching

[limpid nest]

ok cool thanks

Does "On/Off" mean just drive the line hi or low and it turns on or off?

[heavy jasper]

that would be my guess

[limpid nest]

thanks, I can datasheet that to be sure

[heavy jasper]

probably mostly for 1-output devices

[limpid nest]

hmmm

[heavy jasper]

(or just different manufacturers having different names for it)

[limpid nest]

what does 1-output devices mean?

[heavy jasper]

a load switch with one input, one output

(one enable signal, one load)

[limpid nest]

oh I see

ok, I was trying to picture the utility of an 8 input, 1 output device lol

[unreal flax]

Majority vote wins! 😁

[limpid nest]

haha

these chips look perfect for my needs

[ember laurel]

Yay, TI.COM just got 2000 of my mosfet in stock! Snapped up 6x250 reels, so 1500 pieces.

[tough matrix]

you didn't leave much for others :)

[frigid seal]

cries in chip shortage

I bought an alternate part because our part went out of stock. Then I did the redesign around the chip. Then our vendor canceled the order for the replacement chip.

[ember laurel]

ouch ouch

@frigid seal I get the feeling that you kind of have to score your chips first, then design for them

first when you have them in stock - you can be sure to be able to start production now, and not in 72 weeks or whatever the lead time happens to be

[frigid seal]

That's what I thought I did! Guess you have to have the chips on your doorstep before you design for them.

I'm surprised that people aren't stealing containers in transit.

[distant raven]

😳

I’m hoping to grab the TPS62237 but who knows when it will be in stock

[ember laurel]

check ti.com every day

[distant raven]

I have email in stock notifications

Hoping to catch it one day

I just want to order 250, that should get me through a few product iterations

[tough matrix]

Yeah, maybe we should play Robin Hood of electronics and collect a band of archers near the road to TI warehouse...

[frigid seal]

If we are playing robin hood we'd be better served hanging out in scalper's forest.

shakes fist at the person who quoted $250/ea for a 1500 roll of $5 chips.

[tough matrix]

what??? $250 each chip??

[distant raven]

😳

[distant raven]

What chip?

[frigid seal]

An isolated Op-amp.

[distant raven]

TI or Analog?

Or another brand?

[frigid seal]

TI, why you got a fix?

[distant raven]

Nah, just curious

Analog has been better at keeping amps stocked than TI

Is there a certain frequency you need?

There are some Broadcom ones that have -3dB frequency at 200kHz

I found this part with the Digi-Key mobile app. https://www.digikey.com/product-detail/en/ACNT-H790-000E/516-3926-5-ND/5413956

2x the TI price but there are 600 in stock

100kHz variant has 2500

I found this part with the Digi-Key mobile app. https://www.digikey.com/product-detail/en/ACNT-H870-000E/516-3884-ND/7606533

[ember laurel]

@distant raven I also always sign up for the TI mail notifications - but they come way to late

I found checking the website 10x per day is more effective.

[distant raven]

Lol

I had a chance to buy a ton of the TPS22917 10W load switch a while back. I thought: “I’ll buy them in the morning” as I got the notification and they were in stock. Bad decision lol

[ember laurel]

yeah you gotta instabuy

[tender cedar]

If I am trying to make a surface-mountable 'breakout', are there significant drawbacks to putting bottom pads onto the pcb like in the left image, rather than using castellated pads? I'm trying to avoid the cost of castellated pad manufacture. I would be soldering the PCBAs onto larger boards using hotplate + hot air. I don't see many designs that do this beyond exposing thermal ground pads like the ESP32 series, so I feel like this isn't as feasible of an idea as I am hoping. Thoughts?

[distant raven]

I’ve done castellated pads with JLCPCB no issue, others have as well

But under pads seem to make a more “seamless” look

[supple pollen]

I had some special ceramic boards made with bottom pads like that, but they were annoying to solder and I ended up flipping the board and mounting it upside down (so they were top pads) but it took some blobby soldering to make it happen.

[tender cedar]

Hmm I see. Maybe I should just order a run of sample boards to experiment with then; the castellated pads may be the best option afterall. I could also experiment with putting DFN-like pads near the edge of the board and deleting the solder mask between the edge and the pad. I do plan to use castellated pads on one of the boards I am ordering, but the price creeps up pretty quickly if I put them on all the different designs i am hoping to make.

[marble scaffold]

after many hours of grueling work

i present to you: my first pcb footprint

based off of this datasheet: https://datasheet.lcsc.com/lcsc/2005251034_XTX-XTSD04GLGEAG_C558839.pdf

how does it look? am i missing anything?

i have the pads (obv), silkscreen, pin 1 indicator, and courtyard

and just realized that the height is 1mm too short

[limpid nest]

Nice work!

[marble scaffold]

the datasheet didn't specify the length between the outside of the package and the first pin so i decided to ignore for now and try to figure out out later

[marble scaffold]

thanks :)

[limpid nest]

I know the care that goes into a good footprint

[marble scaffold]

after more pixel-perfect (or i guess, micrometer-perfect) nudging and lots of measuring, here it is

[limpid nest]

Be aware of what some call "CAD goggles", sometimes the part looks different from what you want only because of the CAD system

[heavy jasper]

One note - are you drawing the footrpint based directly on the drawing on page 7?

Because by my read, that's just a physical drawing of the package itself, not necessarily a suggested footprint (e.g. if you make the soldering pads pads exactly the same size as the component pads, it may not have a decent space for a solder fillet). The official guidelines there would be from IPC (paid spec) but you may be able to find a footprint generator that takes the package dimensions and turns them into footprint dimensions. Or if not, this guide from onsemi seems like a decent start and calls out how the pads should be sized relative to nominal footprint: https://www.onsemi.com/pub/Collateral/AND8211-D.PDF

[limpid nest]

Harry's always got the goods

[heavy jasper]

Correction: this is an LGA8, not a DFN8 so slightly different (the lead doesn't extend out to the edge of the package to be a fillet-able pad)

Annoyingly - probably your best bet is to check against the easyEDA model (since I'm guessing you're buying this part from LCSC)

since that's at least been cleared by the vendor and hopefully used before

But failing that - I think you're still probably best to treat it like a DFN with the guidelines shown above (and the LCSC footprint seems to show something similar)

[marble scaffold]

laying out/routing high density PCBs feels like playing tetris

[marble scaffold]

tysm, never thought about this

gonna have to spend more hours working on this footprint

but at least it'll be usable

and i won't need to order more pcbs to fix it (hopefully)

[limpid nest]

It's time well spent. You learn as much redoing it

[marble scaffold]

yknow now that i think about it

i could take the footprint from https://github.com/adafruit/Adafruit-SPI-Flash-SD-Card-PCB

import the pcb into kicad, strip away everything else, and save the footprint

it really was that easy

i just assumed that i wouldn't be able to import it into kicad

(tysm kattni for making this footprint)

well at least i learned a lot

[steep fog]

Hi! I'm working on burning a bootloader onto a custom pcb I made, but I'm having some problems.

I'm using a JLINK edu mini with microchip studio and I'm trying to program an atsamd51j19

when I try to read the device signature, this pops up:

After clicking yes, the program freezes for awhile and then just says this:

anyone have an idea whats going on?

[marble scaffold]

where was this samd51j19 from?

[steep fog]

microchipdirect.com

[marble scaffold]

huh

[marble scaffold]

sam microcontrollers have a security feature that lets you secure the chip

preventing you from debugging it or reading the flash memory

companies that integrate sam microcontrollers into their products use this to prevent you from debugging their stuff or dumping their firmware

so i'm not sure why this is showing up on your chip

[steep fog]

I've tried removing the chip and soldering on a different one with the same result

I tried using the jlink on an itsybitsy m4 and it didn't have any problems

so i think the problem probably has something to do with my pcb?

[heavy jasper]

Do you have the packaging as sent from Microchip, and can you check the MPN there to see if there's any weird custom numbering that you wouldn't expect based on the normal datasheet?

It's unlikely, but not impossible, that they accidentally sent you a pre-programmed part intended for some other customer

[steep fog]

I do not have the original packaging

Shouldn't erasing the internal flash reset the security bit even if it was already programmed?

[heavy jasper]

I haven't dug through the SAMD51 security information specifically - some chips don't even let you do the erase in such a mode. But it seems like that is always an option based on a cursory read of the datasheet. Can you try not reading the device signature at all, but proceeding directly to an erase?

When it's in this "longer than expected" mode, do you see any activity LEDs on the jlink tool?

[steep fog]

It lets me try to, but after about a minute this shows up

[steep fog]

Yes, the led rapidly blinks

[heavy jasper]

If you're already doing chip swaps... can you swap to the ItsyBitsy?

That would be pretty clear whether it's a board problem or a chip problem

[steep fog]

My board uses the 64 pin qfp package while the itsybitsy uses the 48 qfn package

[marble scaffold]

wish that the usblc6-2lc6's pinout didn't suck

usb traces need to cross twice

rotating it doesn't work cause now the traces are on the wrong side

[heavy jasper]

Can't you just assign traces opposite in the schematic?

[marble scaffold]

flip the usb connector to the other side?

not backside of the pcb*

right side of the usb stick

nvm

didn't read the last word

[heavy jasper]

The USBlc6-2lc6 is flow-through routing, by my read there isn't an "in" or "out" side of that particular chip

[marble scaffold]

yeah, i could do that

feel stupid now

[heavy jasper]

so you can just arrange the schematic so it lines up and flows through

[marble scaffold]

from this

to this

there we go

tysm

[heavy jasper]

Seems like it'd work. Or you could just swap D+ and D- sides in schematic too

if you want to have VBUS/GND oriented a particular way

[median wind]

so I need to implement a full adder with a ic 7486 and 7400, but as far as I can see, I an user 2 XORs for the sum, but would need 5 NANDs for the carry over. Is there actually a way to implement this with two CIs alone? the 7486 has 4 xor gates, 7400 has 4 nand gates

[distant raven]

9 nand gates to implement a full adder

But a nand and xor chip should be suitable

[median wind]

NAND AND XOR, or NAND & XOR 😛 ?

the problem I saw is that I think I need 4 XORs for anand and pressumed I couldn't do better with the sum than use just 2 XORs, but i should convert that to nand two and see what I get

[marble scaffold]

final pcb design

schematic for it

back side is the ground plane

[median wind]

ok, so I was having way too many nand gates in my design, I cheated and found a website that has the full adder https://de-iitr.vlabs.ac.in/exp/half-full-adder/theory.html and it indeed has 9 nand gates but I don't know how the author arrives to that conclusion

[marble scaffold]

this is second design, and i feel like i did way better than the first

[median wind]

like, why does this have 3 terms instead of four?

[heavy jasper]

The muxing to the samd21 pins is confusing to the uninitiated and I’m only on a phone right now, but you should check: I’d expect CMD pin on the flash chip to be MOSI, and SDO to be MISO, since the SAMD is the Manager and the flash chip is the Subordinate on that bus.

Your actual pin mapping may yet be fine with only reversed net names, but it’s worth a double-check.

[limpid nest]

Hoping for some help building a case for my next installation to use IP rated bayonet connectors. I was told no because the crimp tool for them was $700 and each connector was $10+. I'm using cable glands instead. I find them really lacking. My main case is going to be that 1) bayonet connectors are easier to remove when the installation needs to be taken apart and 2) they are more water/dust tight than cable glands. Does that sound right and can anyone think of more reasons to use bayonet connectors? I can provide more info if needed

[heavy jasper]

Can you find connectors that terminate to solder cups instead of crimps? I think Switchcraft (previously linked) has a bunch of options that could work well.

[limpid nest]

hmmm

I'm not sure what you mean

Oh I think I see

hmm maybe?

I did some research early on and only found crimp connectors.

Like these? https://www.digikey.com/catalog/en/partgroup/bnc-connectors/2241

[limpid nest]

How would you pare these down? I'm in a situation where I don't know enough so I don't know what I'm looking for
https://www.digikey.com/en/products/result?s=N4IgTCBcDaIEIDkDCIC6BfIA

[limpid nest]

how do I find a mating cable for something like this? https://www.digikey.com/en/products/detail/amphenol-sine-systems-corp/RTHP0161PN-H1/5227720

Not sold on this particular one, just curious

[sick geyser]

This is NZD so conversion required but I can get a tool for 84 and the crimps for 5 retail even cheaper bulk/wholesale price

[limpid nest]

this?

Oh link incoming I assume

I need single and double position bayonets

[sick geyser]

https://cdlnz.com/BNC-CTC75 https://cdlnz.com/CT-BNC

I'm suspecting now that you're not using coax

[limpid nest]

I'm not intending too, no

I need 2 single pole connector to carry 110V AC and a 2 pole connector to carry 5V and GND

[sick geyser]

That's the only place I ever come across BNCs

[limpid nest]

You see them in scopes too. My buddy did a design with them that I really like but they required an expensive crimp too.

Tool

[limpid nest]

Here's a good example of what I'm after: https://www.digikey.com/en/products/detail/conxall-switchcraft/7382-2PG-3ES/3515391 But I don't know how to source/create a cable that mates with this

like would I need a crimp tool for the connectors that go to the cable?

[sinful depot]

Here is the mating connector, https://www.digikey.ca/en/products/detail/conxall-switchcraft/6282-2SG-311/3515079

This particular connector uses solder cups so you solder the wires. Use the data sheet to create the part number for the features you need.

Lots of different specs for connectors even if they are from the same family.

[limpid nest]

ahh ok, I thought it was something like that but didn't have time tonight to dig thru. If I don't need an expensive crimp I think I can sell this

thank you

[marble scaffold]

good catch; didn't mean to flip those pins

fixed version

my guess is that while messing with pins trying to make routing easier, i accidentally flipped those 2

also does it matter which sercom pins i use on the samd21?

like am i allowed to assign clk, mosi, and miso to any pin within the sercom

[mystic mantle]

Not just any, they have groups

Just check the SAMD21 SERCOM mux table

[marble scaffold]

using sercom0

i know that i'm only allowed to use these 4 pins for sercom0

but within these 4 pins, are specific ones designated to be mosi/miso/sck/cs?

or can i pick which pin i want to become mosi/miso/sck/cs

[mystic mantle]

lemme check

[marble scaffold]

https://ww1.microchip.com/downloads/aemDocuments/documents/MCU32/ProductDocuments/DataSheets/SAM_L21_Family_DataSheet_DS60001477C.pdf

33: SERCOM SPI - SERCOM Serial Peripheral Interface

(this is the SAML21, on purpose)

i can't find anything here that gives fixed pad numbers for the spi pins

meanwhile i2c has specific lines that are to be used for scl/sda

[mystic mantle]

I can't seem to find any specific info either

[marble scaffold]

so i think it should be fine

[distant raven]

Checkout adafruits schematics for the pins they used for SPI and I2C

PA08/PA09 is often used for I2C

I know I used those two pins for I2C

[tough matrix]

Public Service Announcement: right now, ATSAMD21G18A-AUT is in stock at Mouser: https://www.mouser.com/ProductDetail/Microchip/ATSAMD21G18A-AUT?qs=sGAEpiMZZMtXHE36kCvv36ajhsTJC%252B%2FBgmvnKz0leZ0%3D

I expect it won't last more than 2 days - so if you need it, grab now.
(I've already ordered mine before posting this)

And Digikey has ATSAMD21G18A-AU: https://www.digikey.com/en/products/detail/microchip-technology/ATSAMD21G18A-AU/5056473

[fervent shore]

Where is the best place to look for breakout boards , or is that something people normally make themselves? (need a way to connect mipi-dsi 4 lane, power and an i2c on a ribbon cable). I'm going from a flat panel connector to jumpers (is that a terrible idea?)

[tough matrix]

breakout boards are normally made specifically for a given chip; there are some "generic" breakout boards, usually to convert a surface mounted part to a breadboard-friendly one, but they are very few.
In your case, I'd suggest designing your own - sounds like a realtively simple one. You can have them produced at Oshpark for cheap

[limpid nest]

Oshpark is nice

Plus puuuuurple

[limpid nest]

Should I tin the end of my stranded cable before putting it in here?

[tough matrix]

no!!

https://electronics.stackexchange.com/questions/29861/tinning-wires-that-will-be-screwed-in-to-a-chocolate-block-terminal-strip

if inserting stranded wire is difficult, use ferrules

[limpid nest]

Ferrules will be used on one side of the strip but can't be used on the other for reasons

What do they mean by "fully tin"?

[tough matrix]

the tightening screw should rest against untinned strands

so it is OK to tin very tip of the wire that goes past the screw

[limpid nest]

Ahh ok

I'm too sleep deprived and antsy to just get the tips so I'll just twist them and leave them un tinned

Thanks @tough matrix

[tough matrix]

you can strip longer segment, tin half of it, the cut off almost all of the tinned part, leaving just a very short length tinned

[limpid nest]

That's a good idea

[fervent lance]

Hah. In all these years I thought I was being lazy. Turns out I was being correct (so does the laziness just instantly vanish from the universe, due to a change of understanding? I think not. ;)

I have never really decided what the form should look like when stranded copper is tightened under a screw head (designed for that purpose) as on a barrier strip.

A Google search on 'proper forming of stranded wire under screw head' produces a rich vocabulary of related ideas and concepts.

[limpid nest]

I have a bunch of these (https://www.digikey.com/en/products/detail/molex/0350689702/3184883). On my test setup they are easy to remove, but once I add shrink tubing they don't seem to want to come off. What gives?

[fervent lance]

https://forums.mikeholt.com/threads/landing-stranded-wire-on-screw-terminals.126214/

[limpid nest]

when crimping ferrules, is it important to have the wire length be as close to the length of the housing as possible?

[fervent lance]

Experience will tell. I would say allow 1/16" exposed conductor. All you want to do there is reduce the surface area of a possible short circuit contact-point.

In some circumstances I'd suppose crimping partially onto the insulating jacket would add strength. Just a guess on that.

[worldly schooner]

Length matching is important to ensure maximum crimping grip on the strands inside, as well as minimizing contact resistance. It's only beneficial to a certain point, so no need to go crazy with tolerances.

[limpid nest]

thx

I'm making them so that I can just see the stranded wire near the end before crimping

[limpid nest]

any tips on soldering neopixels or dotstars? The board I'm planning is bigger than my hotplate

[floral spear]

Hi. I'm working on my first real PCB design to send to Oshpark.
What I need to know is, who do I have to pray/sacrifice to, to ensure that I get back a working board with no screwups? Is there a god of gerbers?

[limpid nest]

Unfortunately that god's a fickle master

Posting it here will usually get you constructive feedback

I was considering mounting my own SAMD21/RP2040 but the qtpys of each are so smol it makes sense just to mount the board itself!

[twilit mango]

What OatsNHoney said... if you post it here, you'll find a bunch of folks who can help you through the process and find things you missed. But be prepared for it to take a couple of iterations... even pros end up having to bodge boards to make them work on the first round.

[limpid nest]

let me get a good pic of a bodge. I'm getting paid to do this so I'm a "pro"

[distant raven]

Only downside of rp2040 qt py is parts underneath

Samd21 would probably need SPI flash to be more useful

[limpid nest]

The little 3d printed feet are because I didn't remember to make sure my hat was the right size

[limpid nest]

I can use female headers

[distant raven]

I made a smöl-ish rp2040 design

Single sided

[limpid nest]

Neat

[distant raven]

Just waiting on my JLCPCB panels

[limpid nest]

Neat.

For my Christmas tree design I'm gonna do a panel once I have the prototype done

If I get the green light I forsee needing many of them

[distant raven]

this took me so long to do

[limpid nest]

Wowza well I know who to ask about panelizing

[distant raven]

haha not me xD

i spent hours trying to get this to look right xD

i kid, i can offer help for what i've learned along the way

[limpid nest]

I don't care about aesthetics much, is that what you meant by "look right"?

[distant raven]

not aesthetics

just getting the board areas for the mouse bits and rails to come out

it was one big PCB blob

so much fine tuning

[limpid nest]

Have you ever used "panels by JLC"?

And did you have to lay out your PCB 15 times?

[distant raven]

I've used panel by JLCPCB which is what i usually would use when ordering panels through JLC

but the castellated pads on three sides got my original order cancelled

[limpid nest]

Ahh

[distant raven]

so i had to make my own panel

with luck, these will be in and ready to sell before christmas

I ordered 150 total PCB but who knows how many I’ll sell 🙂

Starting with ~40

[limpid nest]

How would folks go about making a Christmas tree shaped board in EDA?

Really I need 2 or 4 half Christmas tree shapes

[sick geyser]

What price?

[distant raven]

They’ll be $5.50

[tough matrix]

There are quite a few such projects sold on Tindie. Some of them can give you inspiration: https://www.tindie.com/search/?q=christmas+tree

[limpid nest]

Oh I meant more like mechanically, how would you do it. I'll check that out.

[tough matrix]

you can do it same as any PCB design - draw board outline composed of lines and arcs.

[limpid nest]

Ah just painstakingly?

[tough matrix]

But there are also plugins that allow you to draw the board in Inkscape and then convert it to board outline in Kicad: https://github.com/badgeek/svg2shenzhen

[limpid nest]

I know about SVGtoEAGLE and have used it well in the past

Just thought I'd read the room and see if anyone had any killer tips

How do you reckon they made these connect together? https://www.tindie.com/products/modulefun/modulefun-electronic-christmas-tree-kit/

Because that's a better version of what I've been envisioning

[tough matrix]

made slots

[limpid nest]

Yeah

Hm

I see now

[tough matrix]

check other photos on teh product page, it shows it

[limpid nest]

Hmmm, I have to figure out how I want to electrically connect them together

I saw those I was just confused by them. Haven't gotten much sleep

[tough matrix]

electrically, they used right angle male header on one board and right angle female, on the other

looks like 2x2

2x2 female 90 deg headers are unusual, but one can find them on Aliexpress

[limpid nest]

Interesting. I want to use neopixels so I have to figure that out. That's some precise work.

I may crib pretty heavily from this. But I'll make changes to make it my own

I want an rp2040 on there doing neopixel animations

So I'll need power, gnd, and data

[tough matrix]

TBH, idea of connecting 2 boards at 90 deg using slots is hardly new, so they certianly didn't invent it

[limpid nest]

That's fair

[tough matrix]

actually, what they had is 2x2 90 deg male headers on each of half-trees

and 2x2 female headers on the base

[limpid nest]

My idea was 4 1/2 trees connected by PCB rings in the middle.

This looks more like a tree though

Looks like there's a fusion360 to svg plugin too.

I wonder if I can define a shape in f360 normal CAD and then create a board outline from that

[tough matrix]

probably - Fusion 360 now is integrated with EDA

[limpid nest]

Yeah I should have included that I'm thinking there must be a way for that reason

At least I could do is export face to dxf and then import the dxf

[supple pollen]

My usual approach is to get the path somehow, then write a simple Python script to spit it out in the format the CAD package uses (I think it's a fairly simple XML format for KiCAD) or alternatively the scripting language (Eagle has such, I don't know about KiCAD).

[tough matrix]

AFAIK, KiCad uses Python as scripting language

[limpid nest]

I'll have to figure this out, I'll be back. Thanks folks

[raven hill]

does anyone know why it is showing it as if it needs to be connected despite being connected to a via

I assume it doesn't matter because it is not flagged in drc, but I am an absolute noob so I may be wrong

[worldly schooner]

My guess would be the lack of a ground plane, as usually the ground pour is one of the last elements to be added in a PCB design?

[raven hill]

that was it, thanks

[limpid nest]

Can I expect continuity testing with a multimeter to work when my circuit is assembled? That's kinda vague but I have a bunch of connectors going from my board to off board components and I want to sanity check that they are all right.

E.g. there are circuits on board that connect to these, but I want to test the connection between the terminal blocks on board and the off board terminal blocks and then from the off board terminal blocks to the off board components

[worldly schooner]

If it doesn't, you assembled it wrong?

[limpid nest]

That's my assumption

[wild glade]

I was thinking, would it be possible to use a mono 4K LCD and UV light from an SLA printer to form photoresist on a photoresist type PCB for etching?

Like, normally you'd put a transparent sheet with the positive design on it, but why not use the LCD instead

[heavy jasper]

hm - it's an interesting idea. I'd be a bit more worried about reflections bouncing around and causing problems when you're beaming onto a metal surface, instead of a bunch of opaque resin

[limpid nest]

Trying to figure out why the hat I made for my pi is suddenly preventing boot from occurring. Has anyone encountered that before?

I've tried 2 different pis

Both boot when no hat, but fail to boot with the hat on

This is a new issue, previously there were no problems. I can't find a failure on the board

[heavy jasper]

"preventing boot from occurring" what does this mean? Do you get no output at all? Do you get through to loading OS but the OS fails out? Have you checked to make sure that your power supplies remain stable and the extra load from your hat isn't causing a problem?

[limpid nest]

No output and no action on the ACT pin

No indication that anything is running

Nothing seems to be hot either

[heavy jasper]

When you have the hat connected, do you still separately supply power to the pi? or are you expecting the hat to supply power?

[limpid nest]

The hat is powered by the 5v and 3v3 pins on the pi

It's just shift registers and 7407s

I'm flummoxed

It was working before!!

I suspect it isn't booting because the main program is run with a service on boot

[wild glade]

I reckon it wouldn't cause a lot of issues, a mask is a mask, right?

[heavy jasper]

Have you tried hooking up a display to confirm it's not booting (if you have a display-capable unit?)

or tried to get some console out on USB?

[limpid nest]

Yeah hdmi cable is plugged in.

I don't have a way to get console out over USB I don't think