Interstage curious

[emrr]

Current in a PP configuration is balanced between the two halves and in opposite directions. So the core will see net DC of zero.

If current draw is balanced.....LS-50 in PP works but more power tube Ω territory, and that's definitely more current than anticipated, even if balanced. 

 

[CurtZHP]

If current draw is balanced.....LS-50 in PP works but more power tube Ω territory, and that's definitely more current than anticipated, even if balanced.

And one does not want to risk cooking such an expensive and hard-to-come-by transformer.  :'(

 

[emrr]

And one does not want to risk cooking such an expensive and hard-to-come-by transformer.  :'(

Yeah, for DC there's the LS-27, which you'll notice has lesser specs BECAUSE of DC. 

LS-50, go parafeed with a good choke to get the headroom benefits of direct coupling. 

 

[CurtZHP]

Chewed on this over the weekend.  Still weighing the pros and cons of even bothering with it.  I'd have to whip up a power supply before I can start piecing it together anyway.

So far, about the only real difference I can come up with between this and the last preamp I built is an extra transformer and an extra tube (push-pull) stage before the output.  So, basically, more gain; probably a ridiculous amount of gain for most real-world purposes.  I suppose that would give me the leeway to add a tone stage somewhere, since those tend to eat up some gain anyway.

Since I only have enough iron for one channel, I was thinking along the lines of a mic processor as opposed to just a preamp.  But that would call for some sort of compressor circuit as well, and then I'm REALLY in over my head!  Then again, part of the learning process is sometimes getting in over your head.

 

[emrr]

Really only 10dB more gain.  Higher max output.  Less 2nd harmonic because of the PP stage.  Sounds different. 

 

[CurtZHP]

Really only 10dB more gain.  Higher max output.  Less 2nd harmonic because of the PP stage.  Sounds different. 

Yes, the IS transformer will add 10dB, right?  But what about the PP stage?

I like "sounds different."  Another color in crayon box!

 

[emrr]

Yes, the IS transformer will add 10dB, right?  But what about the PP stage?

Not appreciably.  It's still one gain stage, whether SE or PP. 

 

[CurtZHP]

So, I threw this together on the bench over the weekend....

So far, it seems to work very well.  I can't remember for the life of me where I found the idea for the way the push-pull output stage is set up. 

The only testing I've done so far is with a tone generator on the input and a scope on the output, measuring gain through the various stages, and seeing how it behaves with different frequencies, as well as at what point is starts to get crunchy.

Near as I can tell, it's pretty flat response-wise, and with the gain pot all the way up, the sine wave still looks like a sine wave.  (Starts to flatten out ever so slightly at the bottom...)  All told, it gives about 60-62dB of gain, if my math is right.

I haven't checked for noise or actually listened to it yet, but plan to do that this week.

 

[CurtZHP]

I made a few minor changes and put the circuit through its paces.

I'm running into a couple problems that are seemingly unrelated.

First....
The test signal I was using for most tests was a 20mV 1kHz sine wave.  Looking at that on the scope at various points gave me a good idea what the circuit was doing gain-wise.  It also told me that things were staying clean throughout.  The problem started when I bumped up the input signal.  Anything above 25mV would distort on all but the lowest gain settings.  I started back-tracking through the circuit to find at which stage the distortion started.  It starts to appear in the first tube stage, upstream of VR1.  I tried a couple different plate resistances, and even swapped that 12AU7 with a 12DW7 (effectively giving that first stage a 12AX7), and it still starts to distort on any input above 25mV.  What I mean is that the sine wave on the scope flattens out on top and/or bottom.

For comparison's sake, I brought out my last tube build, which has an almost identical input stage (LS-10X into a 12AX7 into a pot...), and that circuit was able to take signals in excess of 100mV without distorting .

I realize a 12AU7 is going to have less gain than a 12AX7, but am I missing something else?


Second issue....
The output stage of this circuit seems to work just fine, but the tube gets EXTREMELY hot!  I expect tubes to get warm, and the first one in this gadget does; but poor little V2 gets too hot to touch.  I wish I could remember where I found this particular layout for this stage, because I might have a better idea of what was intended.  I guess I liked it because it looked so simple .  A tube and a couple resistors.

I've never done a push-pull output stage, or used an interstage transformer before, so I'm still trying to get my head around it.  I understand the basic idea behind it, but clearly my implementation is flawed.  Studying this more carefully, I can see where I pretty likely have a significant amount of DC across the primary of the output transformer, a big no-no with this type.  (Is that what's heating things up??)  I can't imagine just throwing a couple coupling caps in there would solve that issue.  Nothing's ever THAT easy...

Another possibility?
The tubes in this build are used.  Found them in the same box of junk where I found the transformers, so I can't entirely vouch for their condition.  But then, when I swapped out that first stage with a 12AX7, that tube was a known good tube.

 

[CurtZHP]

Yeah, the more I look at it, the more that output stage is a disaster.

I did find the original circuit I borrowed the idea from.  Not surprisingly, that used an output transformer that could handle DC, and there's a lot of it there.  Surprised I haven't melted the transformer.  :-[

And simply slapping a couple caps in series with the resistors won't work, obviously.  There'd be no ground reference for them, right?

Maybe a differential stage?  Long-tailed pair?

 

[CurtZHP]

Something like this....?

 

[emrr]

Anything above 25mV would distort on all but the lowest gain settings.  I started back-tracking through the circuit to find at which stage the distortion started.  It starts to appear in the first tube stage, upstream of VR1. 

all but the lowest gain settings and  It starts to appear in the first tube stage, upstream of VR1 sounds contradictory.  I don't know if I'm reading that correctly. 

I realize a 12AU7 is going to have less gain than a 12AX7, but am I missing something else?

That might be a reason for distorting the following stage, not so much on the input side, and those plate resistor values are going to swamp most differences between an X and a U, with gain and distortion on an X being much higher than with typical plate R.

The output stage of this circuit seems to work just fine, but the tube gets EXTREMELY hot!  I expect tubes to get warm, and the first one in this gadget does; but poor little V2 gets too hot to touch.
Studying this more carefully, I can see where I pretty likely have a significant amount of DC across the primary of the output transformer, a big no-no with this type.  (Is that what's heating things up??)

You are still missing the part where BALANCED DC current is OK, and that's what you have.  It's an output stage with no plate resistors and small value cathode resistors, I would expect it to get very hot.  Measure current. 

I'm wondering why you went with a cathode follower output stage over a traditional plate coupled stage, especially given the transformer in question, which has gobs of inductance for use in plate circuits.   

The tubes in this build are used.  Found them in the same box of junk where I found the transformers, so I can't entirely vouch for their condition.  But then, when I swapped out that first stage with a 12AX7, that tube was a known good tube.

Could be, definitely eliminate apples to apples. 

I don't see mention of the power supply voltage or current draw.  As drawn there's no isolation at all between the PSU and the plates of the output stage.    A traditional plate coupled stage at least would act as isolation, since it's essentially a pair of inductors coupling the plates. 

 

[emrr]

Something like this....?

no, sh*tcan that idea. 

Why are R9/10 different values?  What in the world is R11-14 about?  Why would the grid resistors be so low?  Why would you even need any given the LS-21? 

 

[CurtZHP]

all but the lowest gain settings and  It starts to appear in the first tube stage, upstream of VR1 sounds contradictory.  I don't know if I'm reading that correctly. 

Sorry.  Should have clarified.  I started out looking at the final output on the scope, and when I saw things getting ugly, I turned down the gain and observed that I had to turn it down pretty low to get rid of the distortion.  It was then that I started to work my way backwards through the circuit to find the source of the distortion.  I got as far back as right before VR1.

You are still missing the part where BALANCED DC current is OK, and that's what you have.

Balanced DC is even OK across a transformer winding that isn't supposed have any DC across it?

It's an output stage with no plate resistors and small value cathode resistors, I would expect it to get very hot.  Measure current. 

Now that you mention it, that makes perfect sense.  Will definitely measure current.  Would it make sense to either add plate resistors or increase the size of the cathode resistors?

I'm wondering why you went with a cathode follower output stage over a traditional plate coupled stage, especially given the transformer in question, which has gobs of inductance for use in plate circuits. 

As I mentioned, I stumbled across this hack on the intertubes and it intrigued me.  Thought I'd give it a try.  Not too surprised that it's shaping up to be a dud.

I don't see mention of the power supply voltage or current draw.  As drawn there's no isolation at all between the PSU and the plates of the output stage.    A traditional plate coupled stage at least would act as isolation, since it's essentially a pair of inductors coupling the plates.

Don't have my notes in front of me at the moment, but the plate voltage is around 250-280VDC.

In response to your following comment regarding that other dog of an idea I submitted...
Noted, and sh*tcanned!
The values stated were bodged together using the ampbooks.com online calculator for LTPs as well as some notes from Aiken Amps' site.  I must remind myself that those all pertain to guitar amps, so "apples and oranges."


The hardest part of this project (for me) is figuring out how to implement the interstage transformer, which is why I dove into it in the first place (and landed right on my head!).  It seems almost everyone uses IS transformers for power amps, and not much else, if the internet is any indication.

Obviously, the intent is to feed the grids on a push-pull pair, so now I'm having to educate myself (always a good thing!) on the intricacies of push-pull and how to get from there to an output transformer.

One more question....
Is there any possibility that the distortion in the first stage is a result of some sort of undesirable loading from the mess going on downstream?

 

[emrr]

Balanced DC is even OK across a transformer winding that isn't supposed have any DC across it?

yes, BALANCED DC.  I thought we'd covered that already, and discussed what the catalog says.  0mA UNBALANCED.

Don't have my notes in front of me at the moment, but the plate voltage is around 250-280VDC.

You need current draw too. 

The hardest part of this project (for me) is figuring out how to implement the interstage transformer, which is why I dove into it in the first place (and landed right on my head!).  It seems almost everyone uses IS transformers for power amps, and not much else, if the internet is any indication.

Looking at the RCA 76 program amp again, it's not a power amp.  You're building a 'classic' circuit with 'classic' transformers, you have to look back to when these things were done.  RCA, WE, etc.  That circuit could not be more simple, but it's hard to see in the way it's drawn.  Ignore all the metering taps in the cathodes and treat those pairs of resistors as single resistors. 

Is there any possibility that the distortion in the first stage is a result of some sort of undesirable loading from the mess going on downstream?

The pot is the loading, no problem there.  You need to measure voltages at all plates and cathodes, and note gain measurement at each undistorted location.  If the pot has to be way down, measure what the voltage divider resistances are.  Only this way will you educate yourself about what's going on and why. 

 

[emrr]

this is the most basic form, with addition of a current balancing cathode pot.  You will notice that pot on many tube compressors. 

 

[CurtZHP]

yes, BALANCED DC.  I thought we'd covered that already, and discussed what the catalog says.  0mA UNBALANCED.

You're right, we did cover that.  I clearly misunderstood.

Looking at the RCA 76 program amp again, it's not a power amp.  You're building a 'classic' circuit with 'classic' transformers, you have to look back to when these things were done.  RCA, WE, etc.  That circuit could not be more simple, but it's hard to see in the way it's drawn.  Ignore all the metering taps in the cathodes and treat those pairs of resistors as single resistors. 

Forgot all about the 76.  Like I said, it seemed like everything else out there assumed guitar amps or antique radios.  I'll do some more specific digging. 

Knowing that I'm OK on the output primary is half the battle.  I can stop chasing a problem that isn't there and concentrate on what is.

A question regarding the isolation of the plate supply....
On this circuit, I borrowed heavily from my previous build on the first two stages, and as you can see, I have the 2W resistor followed by the capacitor to ground right before the plate resistor of the first stage.  Granted, on the original build, that first stage was a 12AX7, so much higher gain.  I assumed the isolation should be the same here.  Should I be doing that on the subsequent stages as well?  It didn't seem necessary when it was just one 12AX7 stage into a 12AU7 stage followed by the output iron.

 

[CurtZHP]

this is the most basic form, with addition of a current balancing cathode pot.  You will notice that pot on many tube compressors.

I have seen that in numerous places, but didn't think I could get away with running the plate supply through the transformer in my case.  Leaving that aside, could I still balance it that way?

 

[emrr]

I have seen that in numerous places, but didn't think I could get away with running the plate supply through the transformer in my case.  Leaving that aside, could I still balance it that way?

Yes.

What's different about the transformer being on plate side versus cathode side as you have it?  In terms of current, nothing at all. 

 

[CurtZHP]

Yes.

What's different about the transformer being on plate side versus cathode side as you have it?  In terms of current, nothing at all.

....aside from my own misbegotten visions of smoke billowing from a very expensive transformer, apparently!

So it would work in my case to have the plate supply going to the PP plates via the center of the LS-50 primary, provided both tubes are drawing the same current?

This is where the cathode balancing pot comes in, yes?

Looks like the whole mess stems from my simply missing the point on balanced vs. unbalanced current on the primary.  I assumed there could be NO DC current whatsoever across that winding.