DC on cap with signal

[GainFreak]

Hi,

I have a four stage preamp with this strange problem. With no signal DC voltage after all caps are floating between 0 and 5mV. With a signal applied however I get ~5VDC after 3d stage and ~6.5V after 4th stage. After the 4th stage cap I have a tone stack for the drive channel and a voltage divider for the clean channel that go to a relay. After the drive channel tonestack I have 0mV DC, but since there are ~6.5 VDC on that cap I get some DC on the voltage divider which produces a loud pop when switching from drive to clean channel. No pops when switching from clean to drive.
I tried different caps but I still get that DC. Any ideas what might be the problem?

[R.G.]

There's no good way to say without looking at the schematic. It is possibly a bias shift caused by driving a grid into conduction, but with no more info, I can't tell.

[GainFreak]

It's based on the schematic below with some mods (part values) and some switching omitted:

http://www.yesterdaysrevolt.com/sloc...=264&mode=view

Relays SW3 and SW2 are omitted. After last cap instead of 2M2 resistor it's a 470k/470k voltage divider used for clean out going to channel relay SW1.2

[Chuck H]

As R.G. pointed out, there is often some DC component at the following stage grid. Testing the caps in circuit it's common IME to read some DC. To properly test a cap for leakage you do need to lift the downstream end and test for DC there.

Chuck

[GainFreak]

As I already mentioned I tried several different good quality caps and even added another cap after the last with the same results. A cap after the voltage divider (clean channel) didn't help either. I'm wondering how come there's no DC after the tonestack though?

[Chuck H]

It doesn't matter what cap you use. The DC is on the following grid in the circuit. The reason you don't have DC after the tone stack is because there is another cap right before the PI circuit that blocks DC on the PI grid from getting to your meter. This is the DC (from the following stages grid) that your reading in the other circuits. The extra cap in the circuit between the tonestack and the PI is blocking that from your meter.

Chuck

[GainFreak]

Since this DC is normal I guess there's no other way to avoid the pop but implement the mute circuit from the original E530 preamp?

[Chuck H]

I can't view your link but I'll guess that 'yes' you need a mute circuit. If it was part of the original design then there IS a reason. But perhaps not to avoid pop, more likely to avoid channel bleed or to keep a grid from floating (having no bias via a 0V reference). If you believe it's the DC component causing the pop then you could try the circuit you have but add another cap after the switch to block the DC from the following stages grid. Be sure this doesn't isolate that grid from a 0V reference (usually a gain control or a resistor to ground). If adding a cap does isolate that grid you'll need to add another 0V reference. A 1M resistor to ground right off the grid would probably work fine.

Chuck

[Enzo]

And tell us what sort of meter you are using.

I can't open your link, but unterminated caps are a major cause of switching pops.

[GainFreak]

Here's another link to the schematic:

http://www.aznaur.spb.ru/azgschemes/azg-engl-mod2.gif

Let me know if this one works.

My DMM is this one:

PeakTech® 2010 - PeakTech® - Convincing Top Technology

Below is what I did so far:



Added a cap after the voltage divider but still getting peak DC ( DMM shows up to 1V) and loud pop. Added a 2M2 resistor to ground before the PI still getting loud pop. ---------------Then decided to assemble the mute circuit from the original preamp:



And it took care of the problem almost completely but only after some mods to it. Instead of 10k collector resistor I used 1M as Mesa does in their mute circuit. I guess it changes some time constant because the higher value the resistor (or cap) the more time the signal is muted until it's too much for normal switching.
I'm still curious however about this DC problem and if it can be avoided otherwise? Maybe if clean channel TS is after the last stage for example?

[Chuck H]

If you have a cap on either end of a test point along a signal path and you read DC then one of those caps is leaking unless there is some other peripheral circuit that is inserting DC at your test point between the two caps.

Chuck

[GainFreak]

I checked some other forums and people there are also complaining about the (drive to clean) relay pop in this preamp. I tried the crunch/lead mode as well and there's almost no switching noise even without the mute circuit. The problem is only when going from drive to clean.
Anyway thanks to all for your help.