Hello, I am trying to make a USB microphone that's entirely powered by the computer, but I am running into a few problems.


Project background:

For the actual USB interface I have bought a PS2 Singstar input, which apparently works on PCs too (and it does; very well infact). Although it sounds very good, it wasn't designed for electret microphones, and is effectively a line-level input.
So, what I went on to do next is source out a mic preamp that also had plug in power (for the electrets), and bought this low noise preamp kit.
After buying and building it, I have run into a few problems that are stopping me from progressing any further.

Problem 1:

At first, I couldn't power the preamp by either the USB power, or AC/DC 15v adapter. Powering by USB results in lots of interference (not a ground loop; tested with an isolator and also by plugging the output into a separate battery powered device with the same result). Powering with an adapter (even one intended for audio) results in a massive overdrive buzz. It's the correct voltage (preamp supports 3-25v), so I don't know what's going on with that.

Powering the preamp with a battery works, but I want the unit to be on whenever the computer is, so batteries are not ideal.
Thankfully, I have a "DC To DC Buck Step-down Voltage Module", which very efficiently downconverts 9-25v to 5v. Powering the preamp with this works very nicely, so I do have an alternative to batteries, but again it's still not ideal as it will have to be powered by a separate adapter and won't power on/off with the computer.

Problem 2:

Now, the second problem I've run into and haven't been able to solve is to do with electret power. According to Maplin's tech support, this low noise preamp should provide power to electret microphones... but as far as I can tell, it doesn't! I'm not sure whether it can't by design, or whether it's because mine is faulty. If I posted a picture of the front and back of the board (it's a very simple circuit), would someone who knows more about powering microphones be able to tell me whether it should or shouldn't provide power to the electrets?
I have tried powering the microphones with an additional circuit, but it only works properly with batteries, which I'm trying to avoid. USB power results in lots of noise again, and powering it with the "buck step down voltage module" (the same one that now powers the preamp) results in excessive hiss! Argh!! Not tried using the AC/DC 15v adapter...

Problem 3:

The third and final problem is that the preamp is a tad too loud for the USB mic input, and the variable resistors that are on the circuit to serve as gain control result in the bass disappearing if they're set to anything other than half a turn from max. volume. What could be causing this? Is there any way I can reduce the volume without affecting the bass response?

Any help greatly appreciated.

 

A picture is worth a thousand words, post a schematic what you got.

 

If I had one, I'd have posted it. Despite this being an "educational kit", there's no circuit schematic! These pictures are the best I've got I'm afraid.

The IC used is an NE5532e.

R9 & R10 are meant to be VRs, but because the bass gets reduced when they're turned I removed them.

 

Here's a schematic I found on the web site.

This looks like a dual (ie stereo) single-stage op amp amplifier. Pretty simple.

I don't understand yet why you are having problems.

 

Okay, so now that we have a schematic, we see that there is no power supply regulation on this board. So any noise on the supply is going to come right through the amplifiers. Also, it does not have a supply for the electret [condenser] microphones.

I'd recommend putting a huge filter on the supply to the op amps (a big electrolytic cap and a small resistor (10-100 Ohms). Running from a USB supply is going to have a lot of digital noise on the 5V supply.

Also, is any of this shielded from 60Hz fields (or 50Hz in England)? You are amplifying whatever comes in the input by a factor of 100. Try putting a 'scope on the output to see if there's a lot of distortion. Try it with a shorted input (to ground) to see if it gets quiet.

 

Thanks for the help!

It's good to know that it doesn't supply mic power (though still disappointing), as now I know it's not faulty in that regard.

Regarding using a capacitor to filter the power: will I have to have it in series with the 5v (USB) wire, or will it go from positive to ground?

+5v >----|----->to amp
         |
        C1
         |
        R1
         |
ground>--|----->ground

Or:

+5v >-------C1---R1---->to amp
ground>----------------->to amp

I have a large 2200uF cap, will it do the job?

Also, I'm quite sure it's not being affected by any 50hz fields, as I have it in a properly shielded box. I will, however, try shorting the input when powered by the adapter, to see if that's the problem.

 

Here's how to connect the RC filter.

This has a time constant of .022 seconds (nominal), which gives a cut off frequency of about 7.2Hz, which should be adequate.

IF you have a higher load current (>10mA) then you need to be careful about increasing R1. If the load is very small (like uAmps), you can increase R1 to perhaps 100 Ohms.

Remember the load current causes a voltage drop across R1.

A better approach is to add a voltage regulator after the filter - but do one step at a time. See if this helps.

 

mcasale, that works brilliantly, thanks so much!

Subsequently, problem 2 has been solved as well after filtering the power, as I'm able to power the electrets with USB power as well.