Theoretical question

Hi all.

I gave my son a schematic of a tube preamp for a project at school…

At modest gain settings there were no problem but at high gain settings there appeared some nasty saturation from the output transformer.

I took it to the bench and with the volt meter I could measure some 4vdc after the last coupling cap… into the output transformer. This was at high gain settings. Reducing the gain the voltage came down to 0vdc.

I used the scope and found out that at high gain settings there were some serious interference on the signal at the secondary of the input transformer (he did not use shielded cables from the input transformer to the grid of the first pre am tube).

Changing the cable to a shielded one all the problems were solved. 0vdc after the last coupling cap and no saturation.

My question is how interference on the first input can cause vdc leaking through the last coupling cap into the output transformer?

Is this normal and if so why? It could easily been mistaken for a leaking cap. When measuring I could also see some strange leakage after other coupling caps within the circuit.

I am just curious about it all

Best regards

/John

hf-oscillation into some asymmetric nonlinearity? Like when the AM radio gets into audio..

Jakob E.

Well, I don't know for sure, but I guess people call it "crosstalk" (after reading about similar problems on the Net).

What I am curious about is how vdc can pass the coupling cap when crosstalk is apparent in the early stages of the circuit? And when these early stages are isolated from crosstalk the coupling cap works as normal… passing no dc?


Best regards

/John

Schematic?

Yes, coming up

/John

Here you go…

The preamp is intended for use with a DI-box for acoustic guitar.

Best regards

/John

Oh… sorry for the swedish in the schematic… forgot to translate some words.

Best regards

/John

gyraf said:

hf-oscillation into some asymmetric nonlinearity? Like when the AM radio gets into audio..

Jakob E.

Click to expand...

I am not a tube guy, but this is what I was thinking, some form of rectification causing the HF oscillation to inject a DC offset somewhere in the path.

JR

That's a very strange one. What was the Vdc of the ground bus? Measured from the ground of the output transformer to the rectifier cap ground.  A problem in the ground buss would be the first place I would look. I use a heavy copper ground buss wire (I'll use a piece of 12 gauge pulled from romex)
If it were VAC it would make sense that crosstalk were creating an AC signal, but DC is strange. Could be that your voltmeter is getting mixed up trying to measure DC when there is an AC signal there.

JohnRoberts said:

I am not a tube guy, but this is what I was thinking, some form of rectification causing the HF oscillation to inject a DC offset somewhere in the path.

JR

Click to expand...

I cannot tell for sure. But changing that 2 inch cable from unshielded to shielded stopped all problems. (the cable just after the HPF). If the rectification is causing the problem then it should be causing HF oscillation in other places too?…No?

Could it just be a "layout problem" solved by the shielded cable?


Best regards

/John

johnheath said:

I cannot tell for sure. But changing that 2 inch cable from unshielded to shielded stopped all problems. (the cable just after the HPF). If the rectification is causing the problem then it should be causing HF oscillation in other places too?…No?

Click to expand...

The DC offset is a consequence of rectifying the HF oscillation, just like rectifiers in a PS convert AC to DC.

Stop the oscillation and there is no AC to rectify.

Could it just be a "layout problem" solved by the shielded cable?

Click to expand...

Yes....  exactly

JR

Best regards

/John

Click to expand...

dmp said:

That's a very strange one. What was the Vdc of the ground bus? Measured from the ground of the output transformer to the rectifier cap ground.  A problem in the ground buss would be the first place I would look. I use a heavy copper ground buss wire (I'll use a piece of 12 gauge pulled from romex)
If it were VAC it would make sense that crosstalk were creating an AC signal, but DC is strange. Could be that your voltmeter is getting mixed up trying to measure DC when there is an AC signal there.

Click to expand...

Well, I did not measure that at the moment… and my voltmeter is fine. I haven't had any problems like this in other circuits before and I have used the same grounding of the output transformer before without problems.


Best regards

/John

JohnRoberts said:

The DC offset is a consequence of rectifying the HF oscillation, just like rectifiers in a PS convert AC to DC.

Stop the oscillation and there is no AC to rectify. Yes....  exactly

JR

Click to expand...

Nice

But can you tell me why the coupling caps doesn't block the DC? (Newbie question maybe?)

Best regards

/John

johnheath said:

Nice

But can you tell me why the coupling caps doesn't block the DC? (Newbie question maybe?)

Best regards

/John

Click to expand...

It depends on where the rectification actually occurs, might be after the cap. 

Don't kill too many brain cells worrying about, just prevent the unwanted oscillation. When circuits misbehave all kinds of weird symptoms can occur.

JR

Hehe - No worries… stupid "clients" at work kills my brain cells sometimes and occasionally booze

I agree that if problem solved… leave it there.

Thanks for the info and inputs.


Best regards

/John

Two questions. The reason for asking is to really find out what is happening.

1). Did DC appear on power up  or after turning the HPF switch?
2). How long did it stay on the transformer primary?

I have mentioned this before some time ago on another thread so, excuse the repeat but it is good to keep in mind.

Capacitor will not pass DC but there is what is called an integration effect.  If a DC static is built up on the cable due to lack of shielding this normally will not cause any problem until it is disturbed due to mechanical action of a switch (my first question). This will create a  transient (one time rise) on the left hand side of the first coupling capacitor. As this is a transient it will also appear on the right hand side of the first capacitor. It will travel in this fashion until the right hand side of the 2.2uF coupling capacitor before the transformer primary. Between the right hand side of this capacitor and the ground there is the DC resistance of the transformer primary. So, the 2.2uF capacitor will be charged. This charge will start to discharge through the primary DC resistance (my second question). So normally the DC value you read on your meter should start to drop in line with the time constant created by the 2.2uF coupling capacitor and the primary DC resistance.

Interesting analysis.

This is how it worked: I played a guitar via a DI-box and at the same time messing with a FMR comp. Trying to see that gain reduction effect on the sound I cranked the preamp fully and to my "horror" I got this terrible saturation and of course I unplugged tha comp to see if that one was the problem…but no, so I tested to adjust the gain of the preamp back and forth and this is the time I realized that it was the pre ramp that was the problem. So I opened it up and started measuring voltages all over the circuit.

At a specific point when turning the gain control the problem occurred… it was all the time on the same spot. I measured the voltage on the primary of the output transformer and when reaching this "spot" on the gain control the voltage occurred on the transformer primary… up to 4vdc at full gain and staying there.

So I used the scope and found out that the, in this case, unshielded cable from the HPF were interfered with some sort of crosstalk. Similar interference could be found on the mentioned positions in the circuit… after all sorts of coupling caps.

After changing the unshielded cable to a shielded one all of these interference were gone. Of course it puzzled me how this could happen but I guess John Roberts is right when saying that when circuits misbehave all kinds of weird symptoms can occur.

So… I did not use the HPF-switch and the VDC stayed on the output transformer primary as long as I kept tha gain settings above the "critical" point.

Best regards

/John

It is very strange.  I'll think about this while cooking my dinner.

By the way the secondary of the output transformer is drawn in the most unorthodox fashion if not wrong. That indicates short circuiting both windings.

Also pin 1 of the XLR does not need to connect to audio ground unless you have a specific need to carry it forward.

Yeah, I see it… my bad drawing there

/John

johnheath said:

Hi all.

I gave my son a schematic of a tube preamp for a project at school…

At modest gain settings there were no problem but at high gain settings there appeared some nasty saturation from the output transformer.

I took it to the bench and with the volt meter I could measure some 4vdc after the last coupling cap… into the output transformer. This was at high gain settings. Reducing the gain the voltage came down to 0vdc.

I used the scope and found out that at high gain settings there were some serious interference on the signal at the secondary of the input transformer (he did not use shielded cables from the input transformer to the grid of the first pre am tube).

Changing the cable to a shielded one all the problems were solved. 0vdc after the last coupling cap and no saturation.

My question is how interference on the first input can cause vdc leaking through the last coupling cap into the output transformer?

Is this normal and if so why? It could easily been mistaken for a leaking cap. When measuring I could also see some strange leakage after other coupling caps within the circuit.

I am just curious about it all

Best regards

/John

Click to expand...

Your voltmeter is probably rectifying. Does the oscilloscopê show this DC shift?