Reducing POP when changing settings in GT2

[potul]

Hi All

I'm testing my latest build, a sansamp GT2 from tonepad that I modified by adding a "channel switch". Basically it's a stomp switch to have additional Drive and Volume knobs. This way I can switch between 2 gain/volume settings.

They way I implemented it is by putting the 2 pots "in parallel", and then use the switch to chose what center lug to use, the one from pot1 or pot2.

It works like a charm and I can now have 2 different gain/volume settings and easily switch between them, the only issue I have is that when doing the change there is quite a loud POP ... If you are in the middle of a heavy riff you don't notice, but if you change while you are not playing it's quite annoying.

This is how the drive knob is connected. The volume knob is done in the same way.



Any suggestion on what could be modified to prevent/reduce the pop?

Regards

Mat

[antonis]

As far as I can imagine, there should be a coupling capacitor leftwards (before lug 3) with one side "floating" during switching...

[potul]

thanks for your answer...

Yes, there is a capacitor right before the "S3" switch on the left.
Can something  be done without modifying drastically the sound of the pedal?
Can I add some sort of pulldown resitor here?

Mat

[antonis]

Can I add some sort of pulldown resitor here?

That's exactly why I mentioned the capacitor..
(a bigger part of circuit should be helpful..)

[potul]

mmmm... looking at the schematic, I think  the capacitor is never floating, as there are the 2 pots connected to Vb... and the switching happens after that (the switch I refer to is the red one, labeled SW in my picture, between the pot and the opamp). I don't care about S3 switch, I never use it in the middle of a song...

What is floating during the switching is the non-inverting inpug of the opamp. Maybe I should add a pulldown to Vb here?

here you have a bigger circuit portion

[potul]

bummer...

I just tested my probably insane proposal of adding a pulldown at the input of the opamp... and it didn't work. Pop is still there.

any other option?

[R.G.]

Pops are always a result of the switch suddenly changing the DC conditions in the circuit. This can be tiny changes ("ticks"), short, small changes ("clicks"), bigger, longer-duration changes ("pops") or slower, but large changes ("thumps").

The commonest way to get transients like this is to switch the end of a capacitor. Capacitors block DC, but leak down if one end is left open. The solution to this one way that transients happen is to terminate each end of the cap in the DC condition that it will have when connected, so it cannot leak down when open. The pulldown resistor is a special case of this, where the "inside" terminal of the cap is held at a DC voltage, the "outside" terminal is held at 0V, so no leakdown happens.

In midair switching, it's common to add a pull-somewhere resistor to hold both ends of the cap at a non-changing DC voltage. Sometimes you can do this by putting a large valued resistor in series with the cap and using the switch to "short out" the resistor.

Your schematic does not show in your post so I can't advise on specifics. If you can provide a link, I can help more on this.

One thing that might help is to get out your meter or oscilloscope and watch the output of the first opamp after the switched capacitor for what it actually does when the switch it changed; does its DC level change or does it just do a blip-and-return? That might provide clues.

[potul]

Isn't the schematic shown? I can see it...   
Ok, I will repost it from another source, maybe this way it's visible.

The circuit is Tonepad's GT-2, but with a modification to have 2 different swithable gain pots.
This is tonepad's project:
http://www.tonepad.com/getFile.asp?id=112


This is the portion of the circuit I modified.



The issue is popping noise when actuating the red SW (S3 pops as well,.. but I don't care)

I tried to add a 1M resistor from IC2a pin 3 (non inverting input) to Vb in order to avoid this input to be floating during the switching, as I thought maybe this was causing the noise... But this didn't help.

Thanks

Mat

[deadastronaut]

hi potul, IIRC, i had a very similar problem, and no pulldowns etc would cure it..

i'm a fan of FET switching now, its a few more components sure,

but well worth it for the 'silence'....


take a look at merlins small time delay as an example..

if your on breadboard, it might be worth giving it a go.....just my 2p...

[R.G.]

OK, I can see it now.
What opamp are you using? The NE5532 for example will have a much bigger input bias current than a JFET amp like the TL07x series, and in this circuit the DC bias current for the opamp flows through the pot and wiper, so changing suddenly from one to the other will cause a sudden change in the bias current, making for a pop.

Input bias current will be an issue in any circuit like this. You can test this by putting a capacitor after the volume control and a resistor from the opamp + input to Vb. You can use a resistor of 100K to 1M, and a cap of 0.047 up to 0.1uF to cover the entire audio range.

There are other possibilities, but a sudden shift in bias will cause the opamp to pop.

[potul]

Ok, thanks.... I was suspecting it was something quite inherent to the circuit.

I already have a 1M resistor from the opamp+ to Vb.  I will give a try and add the proposed capacitor and see how this affects the behaviour of the circuit

I will try to change the opamp as well... The one in the circuit is a TLC2262CP,  maybe I'm lucky with a tl072 and at least the effect is less noticeable.

thanks a lot

Mat

[R.G.]

Bad guess my part then. The TLC.... opamps are CMOS, and so will have even lower bias current than JFETs.

You can try inserting the cap. It may help for other reasons.

[potul]

Bad luck... the coupling cap + resistor to Vb didn't  improve the POP.

I noticed that even if the 2 pots are at the same position (let's say they are bot at the minimum position), the POP happens. So it looks like the issue is not related to the change in DC voltage, but the transition when switching. I might explore FET switching.

I guess I will have to live with it by now...while I think about another solution.

Thanks for your support anyway.

Mat

[Cozybuilder]

Isn't the schematic shown? I can see it...   
Ok, I will repost it from another source, maybe this way it's visible.

The circuit is Tonepad's GT-2, but with a modification to have 2 different swithable gain pots.
This is tonepad's project:
http://www.tonepad.com/getFile.asp?id=112


This is the portion of the circuit I modified.



The issue is popping noise when actuating the red SW (S3 pops as well,.. but I don't care)

I tried to add a 1M resistor from IC2a pin 3 (non inverting input) to Vb in order to avoid this input to be floating during the switching, as I thought maybe this was causing the noise... But this didn't help.

Thanks

Mat

Is there some reason why you need 2 pots in this position? Perhaps you could simplify by just paralleling the pot with a resistor you can switch in or out, as the following schematic snippet shows:

[potul]

Hi cozybuilder, thanks for your inputs.

The idea is to have 2 separated independent pots and a switch to use one or the other. This way I can have 2 different gain settings available with a foot step.
If I can't have 2 pots I would be ok with having one setup with the cleanest possible gain with one footstep, to wtich between clean and dirty.  (I've seen some ideas here https://translate.google.co.uk/translate?hl=en&sl=ru&tl=en&u=http%3A%2F%2Fwww.sugardas.lt%2F~igoramps%2Farticle65.htm  but I'm not sure I understood most of them  )

I don't think the mod you propose will do anything... just adding a parallel resistor will not affect much the voltage divider formed by the pot.

Mat

[Phoenix]

Most of the problem is because you're probably using a break-before-make switch, rather than the less common make-before-break. This means that there is an instant where the non-inverting input of IC2A is left floating. You should add a high value ground (in this case Vb is ground) reference resistor between the non-inverting input and Vb. I'd try 10Meg to start with, but even 1Meg shouldn't cause significant change in cutoff frequencies. I'd also suggest that rather than use two 100k pots in parallel, you use two 200k pots in parallel instead, as this will maintain the total load (the pots are always in parallel), so that the cutoff frequencies of the previous stages are not altered. You could use two 250k pots with an 820k resistor in parallel too, or two 500k pots and a 180k resistor, this will get you very close to the nominal load impedance.

[potul]

Hi Phoenix,

I already tried putting a 1M resistor from IC2A non-inverting input to Vb. It didn't help.

I was thinking about the make-before-break topic earlier this morning.... The issue is I'm not aware of any stompswitch like this. Do they exist?


Mateu

[Phoenix]

I already tried putting a 1M resistor from IC2A non-inverting input to Vb. It didn't help.

Ok, sorry, I missed you mentioning you'd tried that already. Did it help at all?

I was thinking about the make-before-break topic earlier this morning.... The issue is I'm not aware of any stompswitch like this. Do they exist?

I've never seen a make-before-break stomp switch before, I doubt you could find one off-the-shelf. If you can find another style of switch such as a rotary (make-before-break seems to be most common in this format) you could try it for testing to confirm or disprove this theory.

I also had a closer look at a schematic (appears to be the same one you're using), and there is quite a lot of gain after the gain control, so it could just be switch bounce, in which case the only solution would be to use electronic switching. You could use a momentary stomp switch to activate a Boss-style soft switch. Check out RG's article on Jfet switching if you want some guidance. http://www.geofex.com/Article_Folders/bosstech.pdf

[ggedamed]

You could check for leaky capacitors as well.



Does S3 position make any difference for the loudness of the pop?

[Phoenix]

You could check for leaky capacitors as well.
Does S3 position make any difference for the loudness of the pop?

That's a thought I had too, but quickly realized that even if any of those capacitors were leaky, with the way Mateu is switching between the gain pots, the DC load following those capacitors will not change, so any DC offset caused by leaky caps would only be between the different voltage divider ratios set by the differences in the pots settings, which should be effectively nothing given that both sides of all the caps are referenced to the same potential (ignoring small input and output offsets of the two opamps).
Won't hurt to try swapping out the caps, but I would be very surprised if that turned out to be the cause.