Skip to content
Notifications
Clear all

Ultimate USB microphone Addendum

8 Posts
3 Users
1 Reactions
270 Views
marcdraco
(@marcdraco)
Posts: 626
Moderator at Large
Topic starter
 

This is very late breaking news for people interested in building this project.

There is a chance that over-driving the THAT1512 will cause an overload that will destroy the input transistors on the digitiser. While this isn’t expected, it can happen with the THAT being capable of generating 13.5 volts of swing into a device that’s designed to work at 5V (max) with ADC limits much lower, in the order of 3.3V or less. This isn’t specified by the manufacturer but I discovered this by accident (and I should know better) – even my equipment predicted an output in excess of 10v and there’s me wondering why the trace on the USB audio is flatlined.

The solution is to add protection diodes to the output. It’s probably too late to change the original project now and I’m close to creating the “ultimate” pre-amp to go with the ultimate microphone. That part is now completed and the full thing should take no more than one more run to make sure I dotted all the “i”s and didn’t short circuit any of them!

If anyone wants to see how the "fix" works, I'll pop details in but it's really a simple diode clip.


Take everything I say with a pinch of salt, I might be wrong and it's a very *expensive* way to learn!

 
Posted : 26/07/2024 6:34 pm
pshepa reacted
marcdraco
(@marcdraco)
Posts: 626
Moderator at Large
Topic starter
 
image

This is a clip from the V2 designs - the output (pin 6) is fed via the 100R resistance and limits the current should the voltage exceed the switch on point of either diode. The result is clipping which sounds terrible but is already over the limits for an audio input anyway. But it won't destroy the THAT or anything later in the chain.

 


Take everything I say with a pinch of salt, I might be wrong and it's a very *expensive* way to learn!

 
Posted : 11/11/2024 1:45 pm
SalmonWall
(@salmonwall)
Posts: 9
Active Member
 

@marcdraco 
Hello! 

I'm a little confused about the reason for using the 1n4004 diodes in particular. Would it be possible to use other diodes, or does the 1n4004 diodes have some unique properties?

I'm still a beginner in electrical circuits and I think that this fix would make a great addition to the project. 

thank you in advance

 
Posted : 12/11/2024 8:14 am
marcdraco
(@marcdraco)
Posts: 626
Moderator at Large
Topic starter
 

EDIT: Where are my manners this morning, welcome to DiyPerks always good to see another beginner has joined the fray - and what better project to start out with than ones that's actually rather useful and tests, hones many skills!

In this case any standard general purpose diode will work, provided you don’t go over the max Vfr (reverse voltage) you’ll be fine. They’re there to cause clipping when the output goes over about a 1V (max) although the come on about 0.6 to 0.7 volt, it’s not that critical. 

You can use some other tech diodes here (even LEDs) but the forward voltages vary quite a bit from 0.2V for Schottky to several volts for LEDs to 10s of volts for Zeners. 

The peak repetitive reverse voltage is what matters. Pretty much any general purpose diode (I keep  bundles of them) from a spares box will do the trick in this case. The 4004 is just one I had to hand.

This post was modified 3 weeks ago 2 times by marcdraco


Take everything I say with a pinch of salt, I might be wrong and it's a very *expensive* way to learn!

 
Posted : 12/11/2024 10:19 am
SalmonWall
(@salmonwall)
Posts: 9
Active Member
 

@marcdraco 

thank you so much for the help.
I've tried implementing this into my capsule circuit, however I've getting some severe clipping on my signal. 
I was expecting for to getting clipping around the 0.7v range since that is what the diodes I am using are rated for.

image
image

the blue signal is after the diodes and the green the before the diodes. Sorry in advanced for my naivety, I'm still working things out.

Also on side note, what range of values can I expect from the output of the capsule?

 

Thank you in advance

This post was modified 3 weeks ago by SalmonWall

thank you in advance

 
Posted : 13/11/2024 1:23 am
marcdraco
(@marcdraco)
Posts: 626
Moderator at Large
Topic starter
 

That's probably my fault for not explainin this correctly. The anti-parallel diodes are supposed to clip but they don't go from the FET, they are fitted to the output from the main voltage amplifier.

The output from the capsule is going to be feeble anyway and won't be able to switch a diode on in a practical circuit since a diode typically requires about 0.6 -> 0.7V (silicone) to turn on and the microphone capsule can only produce a few 10s of millivolts even if it's aimed directly at a Disaster Area Concert.

Don't worry about getting stuff wrong in the sim., it's a LOT cheaper than doing it in KiCAD and hoping all your math is right! 🙂

This post was modified 3 weeks ago by marcdraco


Take everything I say with a pinch of salt, I might be wrong and it's a very *expensive* way to learn!

 
Posted : 13/11/2024 11:01 am
(@coin_skipper)
Posts: 1
New Member
 

@marcdraco

I am new to this entirely, and have only ever played around with Arduinos with straight forward wiring before, and while browsing this forum I noticed you made a lot of posts mentioning a "v2" of the Microphone, or showing a section of what I think is a pre-amp schematic. Would you be able to provide those? I don't really want to have to rebuild a circuit board if it is flawed, but I totally understand if those files are not easily accessible or sharable.  

Many thanks!

 
Posted : 14/11/2024 4:27 am
marcdraco
(@marcdraco)
Posts: 626
Moderator at Large
Topic starter
 

If you plod back up the thread a few pages (I forget how far...) there's a full KiCAD 8 project for the Michelle preamp which is V2 of the mainboard, complete with the schematics BOM and so on.

That variation covers a number of different options including "traditional" P12 (12v Phantom Power) microphones.

The primary change from Matt's original is in the power supply although I did add a lot of extra filtering that isn't practical on Veroboard. It's not amazingly complicated but the options do require someone with Matt's skills to explain them better than i could in an set of written instructions.

There are also two new capsule adaptors inspired by Matt's original head design. One is a beefed up version of a JFET carrier that can be used in several different configurations ad the other is a P12 or P48 phantom powered head with all the trimmings including a 20dB amplifier. 

You can see the modifications I've made in the circuit diagram but if it's not clear, just hit me up and I'll go through the bits that aren't clear.


Take everything I say with a pinch of salt, I might be wrong and it's a very *expensive* way to learn!

 
Posted : 14/11/2024 11:09 am