Two transformers in output stage... Help

[thehoj]

So I'm working on this old Webcor amp right now. It's got 2 transformers between the output tubes and speaker.

The amp has a push-pull 6V6 output stage. I'm not totally sure if this is the right schematic, but I believe it is very close: http://i.imgur.com/Ob9NiZz.jpg

The amp as it sits on my bench right now appears to have 2 output transformers.

First I did some measurements on the transformer that's actually connected to the 8ohm speaker,

Here's what that first output transformer looks like:


With 6VAC applied to the bare leads on the right I measure:
168VAC between Yellow and Red on the left (turns ratio = 28:1)
168VAC between Green and Red on the left (turns ratio = 28:1)
336VAC between Yellow and Green on the left (turns ratio = 56:1)


The second transformer is a little puzzling to me..
It has 3 leads on one side, and 2 on the other.
It looks like this:



If I apply 6VAC to the white and blue leads on the right side, I measure the following on the left side:
Between White and Green = 119VAC (turns ratio = 19.8:1)
Between Yellow and Green = 119VAC (turns ratio = 19.8:1)
Between White and Yellow = 71VAC (turns ratio = 11.8:1)

I assume this second transformer was used somehow to get the impedance ratio more appropriate for the output stage driving the 8ohm speaker.

Can someone help me understand how I would implement these two transformers properly using the 6V6 output stage to drive an 8ohm load? Is this interstage transformer capable of accomplishing what I think it should be able to? This two transformer output section is confusing the heck out of me.. Any help would be greatly appreciated.

[Chuck H]

I'm confused about how the second transformer could have the CT side read with either end closer to each other than the CT??? Are you sure about the diagram and readings?

I've never seen the power tube cathodes coupled through the OT secondary like that. Certainly the amp you have is different. The schematic shows a single OT CT'd on both sides. I suppose it's possible the two transformers in your amp work together to perform the same function. I would have expected a more predictable impedance pairing but I'm sure the designer had their reasons. Can you make a diagram of how the transformers were hooked up in the amp?

[jazbo8]

If I apply 6VAC to the white and blue leads on the right side, I measure the following on the left side:
Between White and Green = 119VAC (turns ratio = 19.8:1)
Between Yellow and Green = 119VAC (turns ratio = 19.8:1)
Between White and Yellow = 71VAC (turns ratio = 11.8:1)

Something is odd with the voltage measurement between the White and Yellow, it should read much higher if the green is indeed the CT... Anyway, what is the model of the Webcor? Some Wecors used cathode feedback but the windings were usually wound on the same transformer core, but it does not seem to be the case here...

[Alan0354]

Measure the resistance to make sure the primary is what you think in your diagram. It does not make sense. Also measure to the phase between the input to output with a dual trace scope and note down the phase of each test to make some sense out of it.

This seems to be a very odd design. The PI stage has positive feedback from R11 and R12. Then you have what looks like NFB of the power tubes through the OT back to the cathode of the power tube.

[teemuk]

This seems to be a very odd design. The PI stage has it's own NFB from R11 and R12.

That's not feedback. It's input to the inverterting amp section of a standard paraphase PI. The input needs to be attenuated to compensate voltage gain of the stage.

Then you have what looks like NFB of the power tubes through the OT back to the cathode of the power tube.

Yep, the secondaries are cleverly wired in series with the cathode biasing circuitry placed on the secondary's center tap. When you start to deal with other tube amps but guitar amps then nested feedback schemes like this become extremely common: Transformers employed in feedback and output tube cathodes as alternative feeback injection points.

[Alan0354]

That's not feedback. It's input to the inverterting amp section of a standard paraphase PI. The input needs to be attenuated to compensate voltage gain of the stage.



Yep, the secondaries are cleverly wired in series with the cathode biasing circuitry placed on the secondary's center tap. When you start to deal with other tube amps but guitar amps then nested feedback schemes like this become extremely common: Transformers employed in feedback and output tube cathodes as alternative feeback injection points.

I got the wrong polarity, it's actually positive feedback in the PI stage!!! A positive going signal cause the plate ( pin1) to go negative. R11 and R12 divided this signal down and feed back into pin 7 which is the negative input of the PI.

Never seen circuit like this.

[jazbo8]

It's a standard paraphase PI as teemuk said. Still curious about the two OPTs though...

[J M Fahey]

Guys, there's a whole World beyond Fender

As noted by Teemu and jazbo8, that schematic shows a paraphase inverter, where the second triode gets an attenuated sample from the first triode plate (attenuation = to tube gain, so end result is gain=1 , and phase gets inverted) and cathode feedback from the output transformer (which needs to be center tapped or have a dedicated feedback winding) straight into the 6V6 .

But I don't think the posted schematic matches this amp, so please hand draw it showing the actual output stage.

Are both transformers used?

Voltage measurements on the second one are also weird, don't trust them.

Did you measure it with all leads unconnected?

We need an actual schematic, even if not pretty.

Good luck

PS: and that's not positive feedback, signal does not go back but forward, follow the signal flow.

EDIT:

pin 7 which is the negative input of the PI.

You still think it's a long tail inverter.
It's not, the tail is very short and decoupled to boot.
No audio at those cathodes so no interaction possible through them.

[Raybob]

The only way I could think of that makes any sense to me, to hook up two, with only two push-pull pair, would be series connection on primaries. Need to do some math to see output impedance, I suppose.

[Chuck H]

Yup. A whole world beyond Fender indeed. Localized transformer FB was employed to achieve the benefits of NFB while combating phase shift errors in the transformer itself. I'd just never seen it done like this. And to this cause I would say that using two transformer to achieve it doesn't make any sense at all because it complicates the phase errors more than it could correct them. So how were these two transformers used in the actual amp? My best guess is that they are used as a load matching system to take an amp typically used for one purpose, like operating a butt load of speakers for an intercom system, and matching to another purpose like standard PA operation. But we can't know. And as mentioned three times, there's something hinky about your measurements for that second tranny. Measure everything again. Remember to take a measurement of the input voltage with the unit hooked up because the AC can load down. And please draw a diagram or report about why you can't and offer any info you can about amplifier identification.

[thehoj]

Hey all, I appreciate the info so far. I just woke up, so haven't had a chance to recheck everything.

I'll redo all my measurements, that second transformer confused the heck out of me. I don't know that that green lead is necessarily a center tap, I probably shouldn't have drawn it like that. I just know what the measurements were between each lead. I will redo those though.

Unfortunately this amp was given to me mangled up, the transformers wires were disconnected already unfortunately, it looked like someone tried to "fix" or modify this amp previously (There were some new electrolytics haphazardly soldered in a few locations). I don't have much to work with as far as original circuit. The amp itself says it's a 166-1 model on the faceplate, that's how I located that schematic online.

At this point, given the completely mangled non-functioning state of the amp, I just want to get it operational in some way. I was hoping I could re-work the circuit in SOME way, I'm not concerned about getting it back to it's original state. When I first started taking measurements I was kind of thinking I'd just do away with that second transformer, but when I calculated the turns ratio of 56:1 on that first transformer, that gives me an impedance ratio of 3136:1 (25088:8) which is too high for dual 6V6's..

[thehoj]

Okay, one mistake is that I had that second transformer measurements backwards.

I have to actually apply voltage to the left side to get a reading on the right side.

And the voltage does drop a bit when power is applied. So it's like this:

With 5.6VAC measured between white and yellow, I read 72.5VAC on the white and blue leads
With 5.4VAC measured between white and green, I read 118.1VAC on the white and blue leads
With 5.4VAC measured between green and yellow, I read 120.2VAC on the white and blue leads

[thehoj]

And I did re-measurements on the first transformer.

It's still measuring the same thing, although the voltage applied to the secondary does drop a tiny bit.

The measurements are as follows:
5.8VAC measured on secondary leads, I read 339VAC on the yellow and green leads
5.8VAC measured on secondary leads, I read 171VAC on the yellow and red leads
5.8VAC measured on secondary leads, I read 171VAC on the red and green leads


Oh also, nothing else is connected to any of the leads when I'm doing these measurements.

[Chuck H]

I'll assume that the cathodes of the power tubes aren't coupled to any secondary. Perhaps grounded?

I think it's like I said in my last post. You have an intercom amp with an auxiliary impedance matching transformer for other loads. I'm almost certain that you would hook the two bare leads of the first OT to the three lead winding of the second transformer (probably the white and yellow OR the white and green) and use THAT setup to connect to more ordinary speaker loads. Hook them up like this and do the test again using the white and blue winding of the second transformer as the "secondary", the yellow,red, green winding of first transformer as the "primary". Again, coupling the two transformers is probably bare leads to white and yellow OR bare leads to white and green. This should be where any impedance switching for more normal loads would happen.

[Alan0354]

I only read the schematic briefly, I thought it's a differential pair, now I see there's a big cathode bypass cap and essentially its a two separate stages. I don't think I like this kind of design, I only read from books before but never thought people actually doing it. Maybe it's not important for low freq in audio range, but in general, the second stage causes phase shift and extra delay, you don't get a true differential signal. On top, the gain of the second stage drift with time and temperature, you don't have good control the amplitude drift between the two stages.

I sure will never entertaining using this in any of my circuit. Want differential signal, do it the right way use a true differential pair like Fender and Marshall. Just because people make it work does not mean it's a good design. You want to have different amplitude between the two differential signals, do what Fender did, using a 100K plate resistor on one side and 82K on the other. It is not just only looking at Fender, there are good and bad designs.