Hissing and whistling in the STe audio output

[Paulo P]

Hi.
I would like to ask about the "STe DAC Audio Noise Fix". I've noticed a little "noise", "high hissing" or "whistling" in the audio output from my STe when the volume on my amp is a little bit higher. I saw a post about it at the forum here: "https://www.exxosforum.co.uk/forum/viewt ... ?f=17&t=38" and also found a homepage to this "problem" here: "http://www.exxosforum.co.uk/atari/last/S ... /index.htm". I could send some short sample of it if it would be necessary.
So I would like to ask for your opinions to this "problem". Could it be the PSU? I think that my PSU looks like original Atari, but it is recapped I think. It looks very well.
I've also noticed a little screen flickering in the hi res mono mode. Don't know if it has any connection with the audio or with the PSU.
Any advice or help will be very appreciated. Thanks very much.

[BlankVector]

I have been suffering from the same issue since the beginning on my STe. I would really be interested to know if there could be a fix for this annoying noise.
NB: STe cinch output suffers of that noise (differently beween L and R), mono output in the monitor port is better.

[troed]

Yeah this is something I remember from my original STE way back then, and I also have it on my main STE now (not the same ). My memory tells me it's the left channel, but ...

I have exxos' dac-cleanup but haven't had time to test it ...

edit: So, original PSU when new as well as on my modern picoPSU:d machine. I'm quite certain the cause has nothing to do with PSU caps ..

[exxos]

I would first try with headphones to rule out other issues firstly. Amplifiers can "hiss" "by design" and its nothing to do with the STE in that respect. Also bad cables between the STE and amplifier can cause all sorts of noise. Its why I suggest headphones to hear exactly what the STE is outputting. You only need a few mV of voltage drop between gnd points between the STE and audio amp to generate noise. So "chunky" quality cables should always be used, not those cheap cables where bend them twice and they break!

The PSU can still have 0.2V noise even after recapping. Noise gets onto 5V rail, causes noise in the audio circuit. My own designed PSU has very low noise and rules that part out. Secondly, my DAC fix beefs up the "regulation" on the DAC chips which rules out more noise. Things like Various RAM access cycles can cause various audio noise also. Likely parity simms can also contribute to this noise also, though I haven't looked into it.

All the STE's I've tested I can barely hear any noise at all anyway. I have had some reports back that my DAC boards helped a lot, but didn't solve the issues fully. Others say it cured the problem fully. Though there hasn't been much testing by people to make anymore progress than that.

[troed]

How high frequencies can you hear? That's one of the key differences between those who hear this easily and those who don't, I think.

It's a constant high pitched whine in one channel, in my experience. I'd wager around 15.5 or slightly higher (I know ...).

Actually I have a wav I could analyze, since I captured the bonus track on Amplitude Problem's "Synchron Assembly" from my STE and can readily hear it there.

/Troed

(Even at my advance age I can hear up towards 18kHz)

[Paulo P]

I'm almost 100% sure it comes from STE's audio output and it isn't the amplifier's hiss. I've tested it on two amplifiers with audiophilic cables, 0,5m long. I can record it and put it here as an attachment. It's about -40dB and it's some kind of whislitng, crackling or buzzing in both channels, not in very high frequencies. It is nice to hear. The sound is changing according to what is actually happening (for example floppy drive access, black/white screen change... in black screen the hissing is much quieter).
It would probably be better to record it to talk about definite thing. I would like to cooperate fully on the solution if there are more people with this problem and if there is an interest. I would like to deal with you all. I think that Exxo's experiences will be very valuable. Of course I can buy an Exxo's PSU, but... as he said... this doesn't have to solve the problem.
So, what do you think? Should I record a short example?

Thanks very much!

[troed]

Interesting.

.aiff with a completely silent (except for the whining noise) attached, it's from the 7an song I mentioned above, captured from my STE.

Looking at the spectrum analysis (done in Audacity) there's definitely something interesting happening here. The two obvious peaks are harmonics, the second is twice the frequency of the one I put the cursor on.

spectrum.png (111.27 KiB) Viewed 10716 times

(I almost feel like wanting to bluebox .. for those who get the reference )

regards,
Troed

edit: Yeah, my theory about how high the frequency was seems way off. Yet the noise "feels" like being at very high frequency. Loop the attached sample and let me know your opinions

edit2: I haven't saved a stereo sample, but I should be able to find my capture-thingie and make one

[exxos]

The sound seems like digital noise. Try running RAM test and the "tone" will likely change.

All the noise is generated and coupled all over the motherboard. When the floppy drive pulls more power from the 5V rail, it affects the way the noise is generated and routed across the board. Floppy drive has a motor which will couple noise back to the 5V for example also.

So better 5V regulation and smoothing caps across the board will help clamp down on that noise. Though I fear its never going to be solved fully. It would need the whole audio circuit "lifting" off the motherboard, placing in a metal can to reduce RF pickup. So all we can really do is beef up the regulation and DAC biasing (like I do on my kit) to make the DAC less prone to noise.

Oddly though, like I said, the boards I tested the noises were so slight I could barely hear them, after my new PSU and DAC boards, it was inaudible. As to "why" other people have much more problematic issues I don't know yet.

My DAC board does improve things, but again, its not known the setup those people had. They could have bad amps and/or cables etc contributing to the problem. So the board could have solved the issues fully on the STE itself, but the user could still have issues elsewhere and thus problem isn't fully solved. There isn't enough data & experiments to prove it eiterway.

If someone has a really bad "hissy board" then send it to me.. I can look into it. Other than the PSU and DAC board fix, I can't suggest much else as I haven't had hardly any issues with sound yet.

[BlankVector]

When the floppy drive pulls more power from the 5V rail, it affects the way the noise is generated and routed across the board. Floppy drive has a motor which will couple noise back to the 5V for example also.

On my STe, Floppy activity can definitely be heard in the speakers. However, I don't know if it comes from motor, DMA, or something else. My ears would say disk-DMA, but I can't tell for sure.

Note that my STe originally had another problem. When STe sound was enabled (Protracker, STe Melody Maker...), there was some random glitches. Something like random spikes, every few seconds, not regularly... I sent the STe to reparation, it has been fixed. But the "hissing and whistling" issue is still here.

[exxos]

On my STe, Floppy activity can definitely be heard in the speakers. However, I don't know if it comes from motor, DMA, or something else. My ears would say disk-DMA, but I can't tell for sure.

DMA access will create various noise issues, same as RAM access, that I have heard before at some point. Though the spin & stepper motor likely causes issues as well. A ferrite core around the floppy power cable may help clamp down on some HF noise.