Why white noise is always present?

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi everyone,

I've made a simple mix table as a university project, the problem is the white noise, when the output potentiometer is at a higher settings it gets worse. I can't find the source of that noise.
I will post the PCB and the circuit layout later.

PS:
Why the headphones output of the PC is absolutely noise free? :confused:
All my circuits have always white noise. That's weird. :p

Best regards,
Daniel
 
It depends largely on the gain structure and fundamental design of your analogue circuitry. Usually the headphone output of a sound card wont have much, if any gain, applied to the signal. Heck lots of sound cards wont even output a solid standard 2vRMS signal either.

If there's a high gain stage somewhere in the signal chain, or if you've got a high gain amplifier at the end of the chain, then this will amplify the noise of the stages before it. If one doesn't need to increase the end signal magnitude in whatever the end design is, then applying more gain than is necessary will only worsen the noise figures.
 
Thank you for your help,

The only high gain stage is the MIC amp maybe that's the problem.

Best regards,
Daniel
 

Attachments

  • Preamp.png
    Preamp.png
    41.1 KB · Views: 255
AFTER the mic amp!
Generally you have the mic preamp (Sometimes with variable gain, but the resistors used are small) then the faders then a post fader summing amplifier.

Also you have lots of very high value resistors which never ends well from a noise perspective, I would suggest a visit to the library for Doug selfs book on small signal audio design which has much on this subject.

10K might just about be ok at line level, but those 100K resistors in the mic amp input stage are a disater.

Also your mic amp appears to be an attempt at a balanced input, which is totally broken by the presnece of the volume pot (Huge impedance inbalance at most settings).

Mostly for low noise audio I see resistors in the hundreds of ohms to a few K region, not hundreds of K.

Regards, Dan.
 
Thank you for your advices Dan,

Variable gain is possible?

Changing the value of the feedback resistor is safe?

I thought that could cause oscillations and bandwidth variations. It could affect gain/phase margins I think.

@jplesset

I've used TL084, and yes I know that I should've used LM49740 or something similar, but that project was intended to use common and low cost parts.

PS: Do you know why unity gain buffers in non inverting configuration get unstable when a large feedback resistor is used?

Best regards,
Daniel
 
Instead of a pot on the input with a fixed gain amplifier following it, you could use a variable gain amplifier with no attenuation on its input. Those 1K resistors in series with the microphone are far noisier than the mike itself, so you can ditch them with a different circuit.

The basic mike amp circuit you want is some form of the three op amp instrumentation amp. It uses one variable resistor to provide gain control and can be made, with suitable choice of amplifiers and resistor values, to add little noise to that inherent to the mike.

Here's a commercial chip that will show you the basics of the instrumentation amp circuit: http://www.ti.com.cn/cn/lit/ds/symlink/ina217.pdf

Regards,
 
@ danny92

Here's what i would do.

After you've decided which MIC PreAmp circuit to go with, insert it in place of U1 in my screenie. Yes i've drawn the MIC's + & - connected reversed to the OpAmps - & + see below.

In your diagram, the mixing circuits were not ideal & interaction etc could occur. It's much better to mix signals to a Virtual Earth circuit, which isolates each one from the others outputs. This requires the OpAmps being used in the Inverting Mode.

To ensure the final output is Non inverting, the AuX & Mic inputs get reinverted in U3.
 

Attachments

  • Zero D 4 danny92.png
    Zero D 4 danny92.png
    6.4 KB · Views: 139
Thank you all for your great help,

@Zero D

The circuit seems interesting but why there is an opamp without feedback? I've never used anything like that, what's the purpose of it?
It's safe to change the feedback resistor value? It doesn't influence gain/phase margins?

Best regards,
Daniel
 
Last edited:
Originally Posted by danny92

The circuit seems interesting but why there is an opamp without feedback? I've never used anything like that, what's the purpose of it?

As i alluded to earlier, i wasn't meaning to use it like that. It was just a general placeholder image of an OpAmp. What you would Actually do is, replace that OpAmp with a Mic circuit based around an OpAmp/s of your choosing. Others made some suggestions for you. The output of that would connect as shown in my circuit via a 3k9 resistor to U3.

Take Note the way i've shown the + & - Mic connections to U1. U1 = your choice of Mic circuit. I've done this to save you needing another OpAmp between the output of U1 & U3. U2 inverts Aux 1 & 2 & get reinverted in U3 to maintain correct phase. If you don't connect the Mic inputs as i've described & shown, the Mic will be out of phase relative to it's input. Therefore it would require an extra OpAmp inverter between U1 & U3 to correct it's phase.

It's safe to change the feedback resistor value? It doesn't influence gain/phase margins?

Which one/s ?
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.