Anyone experience popping sound when switching channel?

[Alan0354]

After I put the amp together and play, I notice the amp gives a slight pop when I switch channel. Here is what I test so far.

1) I isolated the problem to the power amp input relay. The relay is a DPDT, I use one side only. The relay choose between the clean and the OD channel and output to drive the PI stage. I disconnect both channels and grounded both input of the relay. I still get a pop when switching the relay.

2) I remove the input of the PI from the relay output and the pop goes away. So my conclusion is the pop is from the relay.

3) I tried putting a 1uF across the coil of the relay and it did not help or hurt at all.

4) It is very strange that if I turn off the amp for 5 seconds( not enough time for the tubes to cool down) and turn it back on, The pop is very loud when I switch channel. But it gets progressive lighter with time. After 20 seconds, it becomes light enough to be not a major issue, but it's still there.

I saw schematics of other amps and they do have relay switching before the power amp, I don't see people complaining about popping problem. Any suggestion?

I have reverse diode on the relay to suppress the HV spike already. When I heard the pop, I thought it's going to be an easy fix, but I was wrong. I spent over an hour on this and going nowhere.


Edit: I just remember, I did not use any relay drivers. I just connect one side of the coil of the relay to 6V, then I pull the other side of the coil with the switch to ground to activate the relay. Could that present problem due to bouncing of the switch? Maybe I should try using a transistor to pull down, then try using cap to filter out the base of the transistor to deglitch the signal.

[loudthud]

It should be obvious that any capacitor that goes to the relay must have a pulldown resistor (perhaps 1 Meg or higher) so that there is zero volts on all relay contacts when there is no signal. What's not so obvious is that if the B+ is moving around due to sag or tube warmup, there will be a small voltage at the relay as the coupling caps acquire the DC they are ment to block. DC conditions are changing and the coupling caps are charging or discharging to compensate. In some amps they just bite the bullet and use a muting circuit. This is the can of worms channel switching creates. LDRs have their own problems but switching noise is usually low because they switch so slowly. An architecture where all switching is accomplished by shorting or unshorting signals to ground can minimize switching noise.

[Pedro Vecino]

Capacitor between the coil of the relay usually has no noticeable effect if you have a reverse diode. But that same capacitor (1uF) between the switch pins (6V) usually have it. I use it always

[Alan0354]

It should be obvious that any capacitor that goes to the relay must have a pulldown resistor (perhaps 1 Meg or higher) so that there is zero volts on all relay contacts when there is no signal. What's not so obvious is that if the B+ is moving around due to sag or tube warmup, there will be a small voltage at the relay as the coupling caps acquire the DC they are ment to block. DC conditions are changing and the coupling caps are charging or discharging to compensate. In some amps they just bite the bullet and use a muting circuit. This is the can of worms channel switching creates. LDRs have their own problems but switching noise is usually low because they switch so slowly. An architecture where all switching is accomplished by shorting or unshorting signals to ground can minimize switching noise.

Thanks for the reply, I was very careful to make sure everything is pulled down by resistor and no open end for any cap. In fact as in #1 in my original post, I grounded both channels at the relay and just the switching cause the noise.

I found it hard to find any solid state switch for muting because at the input of the PI, I have from +20V to -40V swing. I don't know how people can use a single FET to pull the signal line to ground to mull the signal. The signal goes below ground, it's AC which swing + and -. Particular at the input of the power amp, signal is very large.

I have been searching for solid state relay also.

[Alan0354]

Capacitor between the coil of the relay usually has no noticeable effect if you have a reverse diode. But that same capacitor (1uF) between the switch pins (6V) usually have it. I use it always

Thanks. What do you mean by the switch pins?

[Alan0354]

I am looking into photoresistor or SS relay. I know Sodano use photoresistor in one of the amp. Are they quieter?

I have seen quite a few amps that use relays without any muting circuit. Do they pop also?

[Alan0354]

I just posted a question on Physics Forums in finding a device for muting the signal before the power amp. Please don't feel offended that I post there, because it is like flipping a coin either to post here and link to Physics Forum or post there and link it to here. I meant to post on both as they are more SS but they don't know tubes!!! If you have any opinion, please let me know.

Need help in muting the signal.

The drawing on 2) is actually from Fender Super Sonic. They use a FET to short out the signal.

I am only trying to use SS relay for muting because I don't want the guitar signal to pass through the MOSFETs in the SS relay. They introduce distortion and might change the sound.

Thanks

[J M Fahey]

Please don't feel offended that I post there

Donīt worry about that

As of what you wrote there:

Ideally, I would like a photoresistor type that it becomes over 5MΩ in dark and low impedance with light. But I have not seen any one with build in LED.

I guess you wrote that as a joke.

Hard to match the kind of questions you post with what would be expected from an EE with over 40 years experience designing sophisticated circuits.

Oh well.

By the way, Berkemanīs answer

Yeah, circuit #2 won't work for AC signals, at least not without several modifications.

, when referred to this:

makes me doubt a little about the Technical level at:

http //www physicsforums com

Oh well [2]

When he answer it gets even funnier:

Actually, maybe a better idea -- stay with the shorting circuit and use a zero-crossing-synchronized optocoupler to do the shorting.

Yup, just what the Doctor ordered.
Maybe thatīs why itīs the preferred channel switching method used by everybody

[Alan0354]

Well do you have a solution willing to share instead of sarcasm?

I have been searching today and have not find any optocoupled resistor. You have plenty of photoresistors that change resistance with light. Problem is you have to provide a light source. If you have a device like the SS relay, that you input a current and switch the resistor, please post it.

So you think 2) would work? Please explain. Yep, this 40 years engineer does not know how it work.

[Alan0354]

Yup, just what the Doctor ordered.
Maybe thatīs why itīs the preferred channel switching method used by everybody

Can you show me a zero-crossing-synchronized optocoupler circuit?

[nashvillebill]

...

I have been searching today and have not find any optocoupled resistor. You have plenty of photoresistors that change resistance with light. Problem is you have to provide a light source. .....

I'm no expert, but here's a bunch of Vactrol optoisolators.
Allied Electronics - Catalog

[nashvillebill]

And here's a great article on using them in guitar amp channel switching: Project 145

[Alan0354]

I'm no expert, but here's a bunch of Vactrol optoisolators.
Allied Electronics - Catalog

Yes, that's the one, you are of big help.

Thanks

[J M Fahey]

Well do you have a solution willing to share instead of sarcasm?

I have been searching today and have not find any optocoupled resistor. You have plenty of photoresistors that change resistance with light. Problem is you have to provide a light source. If you have a device like the SS relay, that you input a current and switch the resistor, please post it.

So you think 2) would work? Please explain. Yep, this 40 years engineer does not know how it work.

Google is your friend, 81400 results , including datasheets, list prices, suppliers, tech examples, how to make your own, etc. in:
https://www.google.com/search?num=20....0.AybaTlFVG3c

[J M Fahey]

Can you show me a zero-crossing-synchronized optocoupler circuit?

Google is your friend [tm] , 65400 results in:
https://www.google.com/search?q=zero...x-a&gws_rd=ssl

That said, yes, in that point I was being sarcastic, not about you , my friend, but about Physics Forum user berkeman, who first used complex sounding names (to impress???) in his answer but when you asked him for clarification he retorted:

Zero-Crossing-Detector circuit? I think both should be described fairly well at wikipedia.

No kidding !!!
And then to avoid further questions from you he added:

I have to bail out for the day.

It wasnīt me only, Physics Forum user davenn made fun of him too ... with good reason.
Besides his "bailing out" when asked for details, the *main* problem is that LDR are WAY too slow for zero crossing synchronized anything so his answer was 100% BS from the beginning.
As a side note, I never ever make fun at beginners/noobs/etc. , quite the opposite, will go to great efforts to simplify and explain things, but when somebody comes in a pompous way yet showing absolute lack of skills in the area he claims, well, heīs fair game
In a light, finger ribbing way, of course
For me, MEF is akin to a late Friday night meeting of old friends, relaxing around a few beers.
Meaning itīs a FUN place