Hey guys,
So I was working on a generic schematic for a clean blend that I'd use with an effect that reverses the phase (overdrive, compressor, distortion).
I picked up some blocks in a few different schematics and came up with that (see below) and it works well, but there is some hissing going on, even with the clean pot at maximum (clean sound) compared to bypass. And I have no idea where the noise comes from!
So what am I doing wrong? I looked at the Florist schematic and it is pretty similar (but with a Jfet buffer) and I'm sure it isn't noisy!
If anyone had any idea, it'd be greatly appreciated! :)
Cheers!
If pot CLEAN lets Clean signal out, and FX loop is at minimum, does it hiss? Or if FX loop goes out, and clean is at minimum? I ask, to find the direction/source of noise coming. And what opamps do you use?
Quote from: Zerro on July 02, 2021, 03:23:21 PM
If pot CLEAN lets Clean signal out, and FX loop is at minimum, does it hiss? Or if FX loop goes out, and clean is at minimum? I ask, to find the direction/source of noise coming. And what opamps do you use?
Thanks for the answer!
Full clean : hiss (even with volume of fx loop at zero)
No clean : hiss (even with volume of fx loop at zero)
Basically hiss is everywhere, in a little yet noticeable amount.
Im using 2n5088 for the input buffer and tried tl072 and ne5532 for the opamp. Im stumped!
At first try to hard connect both points I asigned red with ground - see picture. Thus we will know, if noise is coming from bad quality pot or input section. If hiss is there on, problem is at second opamp, or feedback pot - gain. Maybe caps or resistors too, but this we can see after this first test.
Quote from: Zerro on July 02, 2021, 03:44:15 PM
At first try to hard connect both points I asigned red with ground - see picture. Thus we will know, if noise is coming from bad quality pot or input section. If hiss is there on, problem is at second opamp, or feedback pot - gain. Maybe caps or resistors too, but this we can see after this first test.
Thanks a lot! I'll do it in a couple hours and will let you know
Quote from: Zerro on July 02, 2021, 03:44:15 PM
At first try to hard connect both points I asigned red with ground - see picture. Thus we will know, if noise is coming from bad quality pot or input section. If hiss is there on, problem is at second opamp, or feedback pot - gain. Maybe caps or resistors too, but this we can see after this first test.
Yep, still hissing when I ground both points.
I tried several opamps so I doubt it comes from there. And there's only a couple caps and resistors left! I really don't get it.
Now rise value of C21, at 820pF. It is only freq compensation on HF. And, block VR power with 100nF cap. There is only this way to compensate noise, if all details are ok. Otherwise I must say, that some cap or resistor at the end is wrong.
Quote from: Zerro on July 02, 2021, 06:14:52 PM
Now rise value of C21, at 820pF. It is only freq compensation on HF. And, block VR power with 100nF cap. There is only this way to compensate noise, if all details are ok. Otherwise I must say, that some cap or resistor at the end is wrong.
Yes, raising C21 helps, but the white noise is still present, and the dry sound sounds a tiny bit more muffled with 820p.
Im gonna make myself a new audio probe, see if I can figure out if the output cap or resistor are bad, which sound quite unlikely, but who knows...
Thanks again for your help, I'll let you know how it goes!!
Yet you can ground left end of C20, so at IC2 is not any signal from anywhere. And controll again hiss. We will find it. BTW allways do you compare it with bypass branche?
Quote from: Zerro on July 02, 2021, 07:32:27 PM
Yet you can ground left end of C20, so at IC2 is not any signal from anywhere. And controll again hiss. We will find it. BTW allways do you compare it with bypass branche?
When I ground the left end of C20 I get an even louder hiss! Like way louder.
I have it in a true bypass loop at the moment, so I can compare my dry sound and the clean sound on the pedal.
Of course, gain of output opamp is ratio of input and feedback resistor - so, if input resistor is null, only wire at ground, gain is teoretically 50k/0,001 = endless. It was bad trial. But we know, that problem is output amp set. I guess that problem will be in details around. If you changed opamps and nothing changed... BTW, serially with 50k pot set there minimally 1k resistor, otherwise when pot is at minimum, opamp lose meaning - signal is going out around it.
Here is another question - if ANY opamp will deliver such a low noise output, you want to get. It will allways add some noise. So, what noise is normal for that conception, and what is too high level?
Yes, I didn't think about adding that minimum resistor in series with the pot!
I think I'll have to redesign the whole thing, it doesn't work well enough as is... It works, but the noise level is a little bit much to my taste...
Thanks a lot for your patience and help!
That concept is good. Only here is source of noise - maybe opamp is not able to deliver noise level you desire (in this case none concept will work), maybe still try to block VR at noninverting (+) input with 100n cap or so, against to ground - maybe noise is going from power supply! Thats all I can offer now.
Quote from: Zerro on July 02, 2021, 09:50:29 PM
That concept is good. Only here is source of noise - maybe opamp is not able to deliver noise level you desire (in this case none concept will work), maybe still try to block VR at noninverting (+) input with 100n cap or so, against to ground - maybe noise is going from power supply! Thats all I can offer now.
Yes sorry I forgot to say I tried it! Didn't make a difference...
And you did it at pcb? If you have only prototype at breadboard, then I must warn - breadboard always add big noise to signal!
Quote from: Zerro on July 03, 2021, 01:27:29 PM
And you did it at pcb? If you have only prototype at breadboard, then I must warn - breadboard always add big noise to signal!
Yes I used a pcb directly, thinking it would be fine!
So, finally, if nothing will help (controll or exchange those caps and resistors), try super low-noise opamp: RC4580. If nothing changes in compare even with common standard-noise opamps here, it is clear that you must order better quality res and caps. Nothing bad. Concept is good. Only details will ask more.
Quote from: Zerro on July 04, 2021, 10:01:38 AM
So, finally, if nothing will help (controll or exchange those caps and resistors), try super low-noise opamp: RC4580. If nothing changes in compare even with common standard-noise opamps here, it is clear that you must order better quality res and caps. Nothing bad. Concept is good. Only details will ask more.
Thanks! Ill try a 4580, I think I have some in my drawer.
Otherwise yes I could try and change the caps, I've used mostly mlcc on this board. But Ive never had issues with them so far... resistors are pretty standard 1%.
ok. Try that 4580 and compare with common opamp, like 1458, 358 or 082...
Quote from: Zerro on July 04, 2021, 04:51:33 PM
ok. Try that 4580 and compare with common opamp, like 1458, 358 or 082...
4580 doesn't make any difference.
Maybe the noise is acceptable... But I just dont like it :D
So I'm gonna try this instead. Two opamps so a bit more space needed, but that way I'm pretty sure the clean signal is gonna stay clean, with no added noise whatsoever. And I'm sending an inverted signal to the overdrive (that reverses the signal) then returning it through a buffer. It should be clean!
Clean pot is 100kB.
(I'm pretty sure I could get rid of C7 and R14, I'll have to check)
That first model seemed to me better. Minimally you could put paralelly with gain pot 50k some little cap (470pF best) to minimize HF hiss. And serially some res cca 1kOhm. BTW many noises do sockets - at the end it is only mechanical contact! I suppose you put those opamps directly to the pcb desk?
Yes I already replaced the 100p in parallel with the gain pot (50k) with a 470p. It made a difference, but not enough. Larger values cut too much high end.
I don't think a series resistance with the gain pot would make a difference in noise...
The opamp is on a socket yes, but I've never had any issue with sockets!
I think my second version is not very elegant but I trust it to be less noisy (only buffers, no gain stage)
That serial res I recomended not to have clear wire here, when pot is at minimum :@)
I am really wonder how that second version will work.
BTW, with Gain pot at max. 50kOhm and paralelly 470pF cap, lowpass will have knee at 6,8kHz. For most gui FX top freq at EQ settings is 6,4kHz. That cap is ideal with that gain pot.
Quote from: Zerro on July 04, 2021, 09:54:44 PM
That serial res I recomended not to have clear wire here, when pot is at minimum :@)
I am really wonder how that second version will work.
BTW, with Gain pot at max. 50kOhm and paralelly 470pF cap, lowpass will have knee at 6,8kHz. For most gui FX top freq at EQ settings is 6,4kHz. That cap is ideal with that gain pot.
Hey thanks a lot for all your help and your patience! Its really great!
I'll let you know how it works! It's not gonna be before a few weeks I think (I have other projects too... and work... and a family... :D) but I will let you know for sure.
Ok, I will be wonder for it! Z