|
taqueso posted:reminded me of solidworks pcb, which I've seen in the installer but never used. I assume it is terrible, anyone actually tried it? Yes. It's a reskinned version of altium with some features cut and a little bit of version control added. You can move stuff around quickly between solidworks and the pcb software, but it's really not worth it. Also discontinued this year!
|
#
?
Sep 1, 2023 04:42
|
|
|
|
| # ? May 28, 2024 16:51 |
|
|
I'm looking to upgrade my home use Saleae Logic 4. I want at least a couple of analog channels but it's hard to find something cheap. Low speed is fine; 1 MS/s is probably plenty for my purposes. I do have an older 100 MS/s scope that I can use for faster problems. The Saleae Logic 8 has 8 analog/digital channels for $500 list or $250 'enthusiast' pricing I'm aware of Saleae clones but I haven't seen any with analog channels Bitscope Micro, 2 A/D + 6 digital channels for $145 Hantek 1008B/C, targeted for automotive use. Specs look okay but I feel this would be more of a project to use. 2.4MS/s and 8 analog channels for $100-120 What else should I look at?
|
|
#
?
Sep 1, 2023 05:12
|
|
|
Bad Munki posted:IS31FL3197 What is the PWM frequency for this part? Can it be adjusted? I skimmed the data sheet and didn't see anything. I don't know what the max PWM frequency is for your power supply, but the Meanwell one wants 100 Hz - 3 kHz, which might be slower than what you'd get from an RGB LED controller. A more generic PWM chip (i.e. not specifically for controlling RGB LEDs) would let you pick a PWM divisor to set the frequency.
|
|
#
?
Sep 1, 2023 06:31
|
|
|
Captain Cool posted:I'm looking to upgrade my home use Saleae Logic 4. I want at least a couple of analog channels but it's hard to find something cheap. Low speed is fine; 1 MS/s is probably plenty for my purposes. I do have an older 100 MS/s scope that I can use for faster problems. I've used one of these before: https://digilent.com/reference/test-and-measurement/analog-discovery-2/start It wasn't bad, software for it is pretty good. I think there's an analog discovery 3 now which is presumably similar. Still pretty expensive though.
|
|
#
?
Sep 1, 2023 09:13
|
|
|
ryanrs posted:What is the PWM frequency for this part? Can it be adjusted? I skimmed the data sheet and didn't see anything. Well, two thoughts on that, I guess: 1) I am absolutely struggling to find something like what you’re describing on digikey , I’m just out of my element here. Do you have an example? 2) Well geez, software PWM on the raspi runs in the 10hz-8khz range, I may be able to JUST use the optoisolator directly on a handful of GPIO pins. That’d simplify things nicely. Bad Munki fucked around with this message at 23:26 on Sep 1, 2023 |
|
#
?
Sep 1, 2023 15:51
|
|
|
Just ensure that if it locks up or crashes and sets the output to "constantly on" the power supply doesn't react by outputting everything it's got, or that if it does that's not catastrophic.
|
|
#
?
Sep 1, 2023 16:17
|
|
|
Yeah, that's just full brightness, which is annoying but harmless.
|
|
#
?
Sep 1, 2023 16:31
|
|
|
Speaking of, during the hurricane our power didn't go out but it blipped at like 4 AM and I got to find out that apparently the LED controllers I'm using default to "cold white on at 100%" so that was fun
|
|
#
?
Sep 1, 2023 16:43
|
|
|
That's just your emergency egress lighting.
|
|
#
?
Sep 1, 2023 17:56
|
|
|
Fragrag posted:One thing that's not clear to me yet is where the framebuffer is stored? Is it in the SRAM, PSRAM or somewhere else? The OV2640 camera module max resolution is 1600x1200, which is still too large for the SRAM even at 1BPP - treat the address in the framebuffer pointer as an integer and print it out - go to the processor documentation and find its memory map - see what region the framebuffer address is in
|
|
#
?
Sep 1, 2023 21:17
|
|
|
Bad Munki posted:Well, two thoughts on that, I guess: Yeah, I kinda expected PWM I2C chips to be more plentiful, but they are not! How bad is the Raspberry Pi PWM, really? It's not calling into user code for every transition...is it? (wow that would be gross)
|
|
#
?
Sep 1, 2023 21:37
|
|
|
I think the pi has two hardware PWM channels. I would assume they are like MCU PWM channels where they run on their own.
|
|
#
?
Sep 1, 2023 21:46
|
|
|
Cojawfee posted:I think the pi has two hardware PWM channels. I would assume they are like MCU PWM channels where they run on their own. Yeah, but I need 6 total. But if we’re right, software PWM on the GPIO pins should do fine. I’m not sure how that functions internally, I believe it’s interrupt driven, but beyond that, I haven’t looked into the details yet.
|
|
#
?
Sep 1, 2023 23:27
|
|
|
If you wanted to have dedicated PWM, you could spin up a couple MCUs that you can talk to with UART or I2C or something and use their hardware PWM channels. It's a couple bucks to get the MCUs, but then potentially several tens of dollars for a programmer though.
|
|
#
?
Sep 1, 2023 23:41
|
|
|
Scope creep to the max. This is absolutely in the domain of things I would try entirely on the Raspberry Pi first, on a breadboard
|
|
#
?
Sep 2, 2023 00:25
|
|
|
Meanwhile, I've poked Lumissil's US office re. IS31FL3197 PWM frequency. I'll post their answer.
|
|
#
?
Sep 2, 2023 01:08
|
|
|
I'm going to guess 244 Hz.
|
|
#
?
Sep 2, 2023 02:23
|
|
|
I thought the data sheet said 1Mhz, or at least under features it says "1MHz constant frequency" and elsewhere it lists "charge pump operating frequency" as 1Mhz typical. Is that not it?
|
|
#
?
Sep 2, 2023 02:51
|
|
|
Bad Munki posted:I thought the data sheet said 1Mhz, or at least under features it says "1MHz constant frequency" and elsewhere it lists "charge pump operating frequency" as 1Mhz typical. Is that not it? That's the clock speed (probably), which is then divided by each bit of the PWM cycle, 2^12. 244Hz is also what I guessed, but I've tested some Lumissil parts, as well as both TI's TLC line and their LP line, and even within the same family, I've seen anywhere from 500Hz-30kHz, with literally no way you could tell that without chip in hand. It's ridiculous, but now also kinda funny that Ryanrs is independently discovering that.
|
|
#
?
Sep 2, 2023 03:43
|
|
|
Well I'm gonna order some 16-DIP 4-channel optoisolators (which is what I expect to use on the final product) so I can breadboard it up and see if I can actually make the thing work. If it does, I'll be off to the races. Bonus, these changes have resulted in a full redesign on the low-voltage side of the board, and it ended up making the layout massively simpler.
|
|
#
?
Sep 2, 2023 04:18
|
|
|
Am I trying to be too clever? I've realized that the the connector I want to use for socketing my raspi onto this control board provides enough clearance to tuck the larger capacitor comfortably underneath it, with about 1mm of head space to spare. There are zero components on the underside of the raspi. I realize that if the cap catastrophically blows, it could ding the pi there, so I guess that's a risk, but is that actually likely, and beyond that, is this just a bad idea in general? Doing this allows me to shrink the board and everything fits so much more nicely in general.  Raspi outline is the dotted line.  Same as above.  I crudely overlaid a pi to see how things line up, if that's relevant.
|
|
#
?
Sep 2, 2023 04:57
|
|
|
You're putting a lot of trust that the capacitor you end up choosing isn't slightly taller than you expect. Two smaller ones in parallel or maybe a tiny tantalum one could give you some more headroom
|
|
#
?
Sep 2, 2023 05:01
|
|
|
It absolutely physically fits: I have it on hand and put the calipers on it and held it to the board. But I'll look at the tantalum option, more space is good. Gonna be ordering some parts anyhow, so might as well throw some extra bits in. Bad Munki fucked around with this message at 05:11 on Sep 2, 2023 |
|
#
?
Sep 2, 2023 05:07
|
|
|
Tantalums are tricky wrt voltage derating and other factors. I wouldn't use them without a real reason.ante posted:That's the clock speed (probably), which is then divided by each bit of the PWM cycle, 2^12. Ha ha, this is all based on noticing a conspicuously missing item in their data sheet. You'd think they would list typical values in the electrical spec (bonus points if they included a freq vs temp graph). But having zero info stuck out to me. e: My suggestion with the optoisolator assumed you had easy hardware PWM. If it turns it software PWM isn't good enough, maybe a DAC + op-amp is easiest, since it won't need constant updating. ryanrs fucked around with this message at 06:08 on Sep 2, 2023 |
|
#
?
Sep 2, 2023 05:55
|
|
|
Bad Munki posted:It absolutely physically fits: I have it on hand and put the calipers on it and held it to the board. If it fits it fits - go for it. Tantalums are awful avoid them like the plague. if you can't do it with ceramics and you know that electrolytic cap fits, you're in business.
|
|
#
?
Sep 2, 2023 06:01
|
|
|
I'm sure it happens that parts get made out of spec occasionally, but you should be able to trust the specified maximum dimensions given by the datasheet. Imo, avoid tantalum if you can, it's got terrible failure modes and as a bonus it's a conflict mineral.
|
|
#
?
Sep 2, 2023 06:07
|
|
|
Okay, RIP tantalum. Just finished hitting that undo button about a zillion times but I'm back to the aluminum. Thanks again to the whole thread, you've all been more than patient and beyond helpful. 
|
|
#
?
Sep 2, 2023 06:09
|
|
|
taqueso posted:Imo, avoid tantalum if you can, it's got terrible failure modes Anytime tantalum is mentioned now, it should be assumed that tantalum polymer is actually what's being referred to, and it doesn't have bad failure modes The rest though yeah probably
|
|
#
?
Sep 2, 2023 06:10
|
|
|
quote:you should be able to trust the specified maximum voltage given by the datasheet. e: I didn't know about tantalum polymer being better. All my tantalum jokes are NOS from the 1990s. ryanrs fucked around with this message at 06:12 on Sep 2, 2023 |
|
#
?
Sep 2, 2023 06:10
|
|
|
Does soldering iron sponge have any special properties or can I just cut out a piece from a regular kitchen sponge?
I would blow Dane Cook fucked around with this message at 14:19 on Sep 2, 2023 |
|
#
?
Sep 2, 2023 14:15
|
|
|
You need a sponge made from cellulose. Kitchen sponges are often synthetic and can melt to your tip. I think many car sponges are cellulose. Try it out - get a piece of dry sponge and hold a lighter near/against it. If the material melts it's no good. If it burns/chars without melting first, it's fine.
|
|
#
?
Sep 2, 2023 14:30
|
|
|
yeah it charred with the lighter and i'm not seeing anything stick to the iron tip when i use it so i think it's fine.
|
|
#
?
Sep 2, 2023 14:37
|
|
|
I would blow Dane Cook posted:yeah it charred with the lighter and i'm not seeing anything stick to the iron tip when i use it so i think it's fine. If it's not melting into oblivion then it's working, right?
|
|
#
?
Sep 2, 2023 15:01
|
|
|
seems to be working
|
|
#
?
Sep 2, 2023 16:42
|
|
|
You know your electronics fixes are good when they involve a trip to the local auto parts store.
|
|
#
?
Sep 3, 2023 05:15
|
|
|
What's the standard 4-pin 5V PWM fan connector? JST something I'm sure, but there are too many options and I'd like to throw a couple standard fan connectors on my board for simplicity. Maybe it's not as standardized as I'd like to imagine, beyond pitch. Through-hole is fine, right angle is strongly preferred. I guess I could just throw a 4-pin header on.
|
|
#
?
Sep 4, 2023 00:29
|
|
|
Noctua has the connector they use written out in their datasheet. Not sure if the other manufacturers use it as well. There is no right angle version
|
|
#
?
Sep 4, 2023 00:32
|
|
|
ante posted:Noctua has the connector they use written out in their datasheet. Not sure if the other manufacturers use it as well. quote:There is no right angle version
|
|
#
?
Sep 4, 2023 00:36
|
|
|
I spent so long trying to figure out what the gently caress those weird-rear end 4 pin with the friction lock offset headers are called when I needed some and as far as I can tell Molex just considers them part of the KK line of connectors and doesn't make right angle ones (the straight ones are molex part number 47053-1000 for the record). I eventually found a bag of just unbranded right-angle ones on I think amazon but I cannot for the life of me find the order in my order history. e: No clue what the hell this site is so trust it at your own risk but they appear to have em': https://www.moddiy.com/products/4-Pin-3-Pin-2.54mm-Pitch-2510-Fan-Male-Angled-Connector-Black.html Shame Boy fucked around with this message at 00:48 on Sep 4, 2023 |
|
#
?
Sep 4, 2023 00:45
|
|
|
|
| # ? May 28, 2024 16:51 |
|
|
How do you control the PWM for the fan? I've read they need 25kHz signal. I'd like to not have to change timers on an arduino to get it to work, but maybe that's the only practical option.
|
|
#
?
Sep 4, 2023 01:04
|
|
