Question on noise.

Hardware questions and issues with the FV-1

Moderator: frank

Post Reply
ice-nine
Posts: 192
Joined: Thu May 24, 2012 9:03 am

Question on noise.

Post by ice-nine »

Hi Frank , I've just been having some trouble with white noise/hiss with the SKRM modules (and other FV-1 builds). I just wonder if it can be eliminated in any way.
I will try and give a good example of what is going on first.
I use the SKRM module for these tests and on the input of the SKRM I have a simple unity gain op amp buffer based around an OP275 but have also used a bog standard TL072, the input impedance is around 470k and the output is taken directly to the SKRM and I know that to have the input filtering components already in place. The output from the SKRM is taken to the other half of the op amp which is also a unity gain buffer with both input and output impedance set to around 100k.
Signal is good as I can just get the clipping led to light if I play loud enough but have it backed of just enough not to cause clipping.

The issue I am having is a large amount of hiss on the output, I have also programmed the FV-1 to just pass the input straight to the output of the SKRM so that what goes in should be what comes out.

If I remove the SKRM and pass the signal only through the analogue side of the circuit then there is zero hiss so it would seem the white noise/ hiss is coming from the FV-1 for some reason. Is there something I need to check out or is a lot of filtering stages needed on the input/output to combat this issue?
I look forward to opinions on this hiss problem.
frank
Posts: 1244
Joined: Wed Oct 19, 2005 12:26 pm
Contact:

Post by frank »

Without seeing the PCB it is hard to say but I would check that the power is clean and if it is maybe a little extra filtering on the output and the input, could be out of band noise that is being aliased at the ADC and needs a little more aggressive filtering on the input. If you have a digital scope that has math functions then it probably also has FFT built in and you could look at the spectrum of the noise and follow it through the circuit.
Frank Thomson
Experimental Noize
ice-nine
Posts: 192
Joined: Thu May 24, 2012 9:03 am

Post by ice-nine »

Thanks for the info Frank, I dodn't have a quality scope at my disposal anymore to use but have managed to get the result in the end.

The power lines are clean and no noise on the inputs to the FV-1, so I tried a program with just a freq generator to output to the DAC's which still produced the white noise.

So the simple answer in this case was to alter the LP filter at the output to 1.6Khz which has reduced the hiss to almost nil while still keeping the audio response correct.
Sweetalk
Posts: 141
Joined: Thu Oct 15, 2009 5:13 am

Post by Sweetalk »

I have the same hiss on the output, more hiss as the output volume of the FV-1 increases.

Ice-nine can you send me the values of the modified filter to check out if it works for my board?
seancostello
Posts: 74
Joined: Mon Sep 11, 2006 10:04 pm

Post by seancostello »

ice-nine wrote:Thanks for the info Frank, I dodn't have a quality scope at my disposal anymore to use but have managed to get the result in the end.

The power lines are clean and no noise on the inputs to the FV-1, so I tried a program with just a freq generator to output to the DAC's which still produced the white noise.

So the simple answer in this case was to alter the LP filter at the output to 1.6Khz which has reduced the hiss to almost nil while still keeping the audio response correct.
If you put a complementary highpass filter on the input, you can have a flat frequency response through the system. Pre/de-emphasis is a trick that was found in the analog sections of most early digital audio processors. A bonus to this trick is that it gives you more headroom, as the low frequencies are attenuated before being processed (in my experience, the low frequencies tend to clip in my FV-1 algorithms before the highs).

Sean Costello
ice-nine
Posts: 192
Joined: Thu May 24, 2012 9:03 am

Post by ice-nine »

Thanks Sean, the pre/de-emphasis is how it has been done now and the results are a big improvement.

@Sweetalk in the version I have that just has unity gain buffers which I was asking about originally, on the output of the FV-1 using a 1k - 100n R/C combination works well but experiment from 33n to 100.
Sweetalk
Posts: 141
Joined: Thu Oct 15, 2009 5:13 am

Post by Sweetalk »

ice-nine wrote:Thanks Sean, the pre/de-emphasis is how it has been done now and the results are a big improvement.

@Sweetalk in the version I have that just has unity gain buffers which I was asking about originally, on the output of the FV-1 using a 1k - 100n R/C combination works well but experiment from 33n to 100.
Thanks, I'll try that!
blackbee045
Posts: 1
Joined: Fri Oct 24, 2014 4:58 am

Post by blackbee045 »

I play loud enough but have it backed of just enough not to cause clipping.

The issue I am having is a large amount of hiss on the output, I have also programmed the FV-1 to just pass the input straight to the output of the SKRM so that what goes in should be what comes out.

If I remove the SKRM and pass the signal only through the analogue side of the circuit then there is zero hiss so it would seem the white noise/ hiss is coming from the FV-1 for some reason. Is there something I need to check out or is a lot of filtering stages needed on the input/output to combat this issue?
GuL
frank
Posts: 1244
Joined: Wed Oct 19, 2005 12:26 pm
Contact:

Post by frank »

Try the filters recommended above.
Frank Thomson
Experimental Noize
Post Reply