[Sticky] USB-C Microphone (official topic)

Hello, I can't for the love of god figure out whats wrong with the board, I have changed a few stuff from the tutorial for example:

1. I'm not using a potenciometer (just a 1k resistor)
2. Im also not using a usb c board instead i directly soldered the the power from the board

The board normally connects to my pc and gets recognized but i dont get any sound.

Some other problems: the black box and the THAT chip get pretty warm and also when i measured resistances i got bad readings on:

1. Pin 6 (THAT chip) to the white cable on the green board 175k omhs instead of 0 apparently
2. Pin 4 and 5 (THAT) have a resistance of only 500 ohms should apparently be around 100k

So does anyone know what could be wrong? I also noticed that the black capacitors are oriented correctly in the video tutorial but not in the schematics

I have marked - and + for the green capacitors red is negative and blue positive btw

Circuit impedance (measured in circuit) can be very deceiving unless you designed the whole thing yourself and know what to expect. As the THAT is a black box, we don't.

So - it sounds like you hit the H/F noise problem - the THAT is just blasting supersonic noise and there's so much of that there's no power to amplify the real signal.

If you ponder these pages you'll find you're not alone in this, the base design was plagued (and why I spent years, literally, developing one that was fire and forget). I also did a PCB which some people found useful however, let's go with what you got.

A few pages back there's a chap with similar issues so rather than repeat all that I want you to do all the voltage tests there (with black to GND/0V) and red to the other point with an without the THAT in place. That helps me figure out what might be the issue.

@marcdraco thanks. Just to be on the safe side with my understanding: I'll need to solder the Js by myself later, and test with different R409s (or try with a pot.) to adjust the gain, am I right?

EDIT: This is the final quote, isnt it too expensive?

i dont want to say these very well designed amplifiers are cheap, but i was expecting a price close to the original entire project (30-40$). If not, in your experience with users in this forum, is it worth it to pay the extra cost in the PCB relative to assembling original electronics? Does the original assembly usually come to a bad result?

Sorry @wecan I'm not being deliberately rude, just I'm on a dead-lion trying to make improvements to my apartment (before the landlady shows up at the end of the week. So far I'm about 2 weeks behind and the place looks like a building site. Complete with paint on the carpets (and worse) ... Soooo I'm kinda slow 0- by all means kick me a PM if desperate.

The price you have there is for TWO completed boards - that's a bind but it's what we have to deal with. If we can get this sufficiently tested (a few other people chime in with changes) I'm sure we could find a distro and get a much better deal but hey ho.

So for practical purposes you're getting two Michelles for the price of... one (sorta) but that's still a damn drag. The delivery pushes it a bit more too.

The gain you get with a board from JLC is fixed - *by design*.

This allows you to jiggle about with other bits and not have to worry about fooling with the gain. The value there is sufficient to knock the VUs of the charts from a typical "cereal packet" condenser.

This allows a little bit of experimentation with the input resistors for the "left" channel - remember this thing is MONO (but with a stereo front end which is for future expansion).  

R409 is where your "6-pack" or your pot would (will/can) go. But don't forget that it will be in parallel with the existing 100R value.

NOW, if you're asking should you do it yourself or have this done in the factory: factory every time. You can make this at home if you have a serious setup including a hotplate or an oven. This is not for the feint of heart (well I wouldn't)!

The factory made ones I've had (over the life of the project) I've never had a single one fail due to an error at JLCs end and that's 100s of boards over the last few years.

If you cost out the V1 vs. this they compare very favourably when we compare a veroboard job to a proper factory produced one. Be sure you have the songbird logo right (it'll look awful if your font is wrong). I need to make a SVG up at some point.

Anything I missed? Just holla!

@marcdraco Thank you for your hard work ❤️ 🌺 

Tomorror I am going to order pcb and parts. So in 2-3 weeks I should post how it wents.

I have two questions:

1. Is this board good for faint sounds, like fan or whisper, capsule probably plays big role in it, but for sure board also plays a role?

2. Are you planning on any other changes that would further improve a quality, or only polishing?

Good morning! 
@marcdraco, sorry for tagging. I would like to confirm about 2 version of microphone. We have discussed several months ago and wondering about where to start! Thank you for your patience and support!

Posted by: @spofforth

↑

@marcdraco Thank you for your hard work ❤️ 🌺 

Tomorror I am going to order pcb and parts. So in 2-3 weeks I should post how it wents.

I have two questions:

1. Is this board good for faint sounds, like fan or whisper, capsule probably plays big role in it, but for sure board also plays a role?

2. Are you planning on any other changes that would further improve a quality, or only polishing?

1. Exquisite - it depends more on your microphone capsule's self-noise (which is caused mostly by air molecules hitting the diaphragm).

2. Anything beyond this point is basically polish. The next "jump" is the sort of thing I can't do on my own, I simply don't have the resources. I've done a P48 compatible power supply but that's not quite ready I'm backed up. However, there's nothing saying someone else can't take the digital (I2S) outputs and hook them to anything from a headphone adaptor to a mixer. There's space on the board for a full sterero setup but it's only designed for Mono (particularly over USB).

@ba-ichigo 

Hi there, no worries, better to tag me if it's my design(s).

There are some design files further up the thread that can get the final (beta) version. Anything else would be polish and fixing any gremlins that crept in afterwards.

Ideally get a few of these out there. What's crucial is the impedance matching. If you don't get an impedance matched board to the USB spec (it's available from JLC at no extra cost) it likely won't work.

Beyond that and needing some sort of mic head (I use mine with a cheap(ish) capsule because I needed a particularly flat response. The amp itself is steady as a rock (within reason and at a fraction of the usual price) of stuff that would cost far more. 

But it's mostly plug and play - you just plug your electret in at the input and audio comes out of the USB-C connector at the other end. Want to listen in? The I2S is available - including the 12MHz audio sync so you're golden there too.

So I accidentally fried mine messing with it. I had to get a new DC-DC convertor. I made a few changes. The 220uf caps are 'cus I happened to have a bag of them and I wanted to save space. Also...the datasheet for the DC-DC convertor says max output capacitance is 330uf (which is more than the 10uf of the original model. Which might be why it fried itself having to charge those caps every time I plugged it in.) I also added those .1uf caps. Again - no math, it's just I had a bag full of them. I moved all the smoothing caps as close as possible to the DC-DC convertor and moved the 100 ohm resistors right next to them, so the long runs are after the caps and resistors now. And I added a fuse. but Imma change that a bit. I borrowed some real estate on the ground strip, right where I'd ideally want the ground wire from the USB to come in. So I'll borrow a bit of the strip to the left instead. (OR just have the circuit board be one row longer. DUH. It's width-ways where space is at a premium in my enclosure) Other than that, it's the same circuit, just with the dirty laundry kept farther away from the clean.
 Oh, and I did away with the USB adapter, because I've fried three of them! Audio is gonna go through a 100 ohm resistor and split through two 1k resistors to a 3.5mm jack socket. Then Imma use a 3.5mm to RCA lead to just plug into an External USB adapter. It was just too cramped in my enclosure and the risk of shorting against the sides was too much. Plus all the fiddly little wires...
 So I have power via the USB-C socket, and a line-out via the jack.
I just finished the drawing and it's 2:37am, so there WILL be mistakes. Feel free to point them out. I included a version with all the unused strips removed. I'll try to scooch all the stuff below the THAT upwards a bit. I have room to expand lengthways, so I can get a bit more separation between the power section and the rest that way, too. Probably quite considerably. And I can move the socket that connects to the mic capsule back down where it is on the original, so I can shuffle everything after the 100 ohm resistors upwards by an extra 3 holes. I wanna see how much space I need for the jack socket first. it will probably come in from the bottom left.

@marcdraco 

Thanks for your answer! I'm new to this, just checking if I got it right.

I can take the latest board design, order it, then just connect an electret mic capsule — and it's basically ready to go?

Thanks!

Posted by: @ba-ichigo

↑

@marcdraco 

Thanks for your answer! I'm new to this, just checking if I got it right.

I can take the latest board design, order it, then just connect an electret mic capsule — and it's basically ready to go?

Thanks!

pretty much, yeah. A typical electret might need some gain (you can set the gain on the board with a resistor or a pot)! 

Be aware it’s still beta grade so I’m sat on my hands until a few real people are brave enough to try it. 

Performance of the original was a hard act to follow and keep the cost reasonable.