More Phase Splitting Questions

[Sherman]

I tried several configurations of circuits to use some small interstage transformers I have as phase splitters. A couple of them actually worked... except the frequency response was terrible in every configuration. Zero bass below maybe 200 Hz plus sounding like the sound is coming out of a cardboard box. (See my question and some kind responses in another thread "Can I Phase Split Like This".)

So I was sitting at the bench when my eye fell on an Edcor XPP-8K-8 push-pull output transformer. Hmmm... I connected the CT of the primary to ground, connected one of the plate tabs to both the the output of my 6N1P input stage through a 1uF cap and to the grid of one of the EL84s in the PP pair and the other plate tab to the grid of other EL84. Nothing at all was connected to the primary.

The result was very good. The amp now had bass and the overall balance seemed quite nice.

So I pulled out some other trafos and plugged them into the circuit to try them out. I didn't change the circuit to optimize it in any way, I just swapped trafos around.

* 6.6K from Triode Electronics- worked OK but seemed a bit bass-shy

* 5K SE from Edcor (XSE15-8-5K)- much more bass shy

* Edcor 10K:10K interstage (hooked up only the secondary)- even more lacking in bass

* 48VCT EI power transformer (extremely low output, no bass)

* 28VCT toroidal power transformer (same result as 48VCT)

All-in-all the Edcor 8K Push-Pull worked the best and its only $20.


Now for the questions:

1- What's wrong with this approach other than the cost and space of the extra iron (which doesn't cost all that much at $20)? I figure there has to be a drawback or everybody would be doing it!

2- Is frequency response of the trafo any better (or worse) since there is no DC in it?

3- Would a higher wattage Edcor have a better frequency response?

4- What is the best way to optimize the load for the 6N1P (or whatever)?

Thanks for the help! I'm slowly learning this push-pull stuff and I'm starting to get my copy of 'Valve Amplifiers' a bit dog-earred.

[SY]

Dumb question: why not phase-split at the input using a 1:1 with a couple precision resistors or a 1:1+1? No DC, lower signal levels, true balanced/push-pull all the way through.

Bass lack = insufficient inductance.

[Sherman]

Originally posted by SY
Dumb question: why not phase-split at the input using a 1:1 with a couple precision resistors or a 1:1+1? No DC, lower signal levels, true balanced/push-pull all the way through.

Bass lack = insufficient inductance.

That's definitely not a dumb question! I don't know enough to know what I don't know, you know?

I hadn't even thought of phase splitting with resistors though the input trafo had occured to me.

When the little guys didn't work out the other trafos were just sitting on the shelf and it only took a minute per to swap them so I though why not?

(I also built a concertina phase splitter which works fine. I'm just trying to learn something instead of blindly copying schematics.)

Splitting w/ Resistors
Two resistors forming a potential divider, input to the "top" resistor and on to the "top" driver, junction connected to ground and the "lower" resistor out to the "lower" driver. Resistor values chosen to give acceptable input impedance in parallel with the volume pot. Am I on the right track here?

[SY]

If I understand you right, then yes. If a 1:1 transformer wants to see a 10k load on the secondary, connect two 5k precision resistors in series across the winding, then ground their junction. Pick the now-balanced signal off each end of the secondary.

With a 1:1+1, it's even easier. Ground the + end of one winding, the - end of the other. Load each one with its optimal resistance. The resistors should be close, but the balance is not as critically dependent on the matching.

[Sherman]

Thanks Stuart!

As you can see I've a bit to learn yet. I'm not trying to get to guru status, I just want to be able to make some intelligent choices and have some fun along the way. Ending up with a good-sounding amp would be nice as well.

I just wish I had an easier way to set up a test bed. some 6SN7 and 6SL7 valves so I can use my octal relay sockets to more quickly change things without soldering. Wish I could find some noval sockets like that!

[tubelab.com]

That's definitely not a dumb question! I don't know enough to know what I don't know, you know?

Isn't that true of all of us?

I have discovered several working circuits because I didn't know enough to know that it wouldn't work. The bosses at work have decided not to tell me that something won't work, because I have been sucessful with things that "won't work". Sometimes it pays to just do it!

In fact, I am being sent to the Chicago area tomorrow to discuss some technology that "won't work."

Triode Electronics makes a "Universal Driver Transformer" which has 5 windings on it. There is a center tapped primary and 4 independent secondaries that can be wired up in series, parallel, or used seperately. I got a pair thinking of running two seperate push pull output stages off of a common driver. Then the load is tied across the two outputs, like a vacuum tube H bridge. I haven't had time to measure the response of the transformer yet, but I have discovered that it really doesn't like DC.

[Sherman]

I think in the world of tubes I can only dream of stumbling onto something relatively novel. However even the everyday stuff has a certain magic to me. I'm still a bit amazed that I can solder some resistors, capacitors and glowing glass tubes and have beautiful sound come out!

Maybe I'm easily amused; I've been a photographer for more than 30 years and still get a thrill when I open the developing drum and discover that there are indeed images recorded on the film. (Yeah I use film. I also use a 4x5 view camera.)

I have to say that I'm pretty impressed with the OPT as phase splitter though. Frequency response is great and it sounds good. This afternoon I got a chance to plug in a potted Avel-Lindberg 120-120 to 15V-15V (dual primaries and dual secondaries). Using the primaries as the phase splitter seems to work nearly as well as the Edcor OPT at about 1/2 the size. More fun stuff!

[Shoog]

The thing with using a lot of transformers as interstage splitters is indeed the low inductance. This can be overcome, but at a high price. They need large bias currents on the driver tube and super low driver impedance. Only a few valves fit the bill and it seems a waste to throw so much heat at using a suboptimal transformer when good ones are availble.

Shoog

[DigitalJunkie]

I think you could get rid of the 100K resistors on the output tube grids,since the grids are DC-grounded through the transformer.

(just a quick 0.02)

[Sherman]

Originally posted by Shoog
... it seems a waste to throw so much heat at using a suboptimal transformer when good ones are availble.

Shoog

I imagine there are really good phase splitter transformers out there. I don't have any and I'm not sure I'm going to be purchasing any. I'm just experimenting here and trying to learn something. But if the Edcor sounds good and is $20 why not use it?

Originally posted by DigitalJunkie
I think you could get rid of the 100K resistors on the output tube grids,since the grids are DC-grounded through the transformer.

(just a quick 0.02)

Never occured to me. I originally put together the output section thinking I'd use a concertina phase splitter (and actually did for the first test). Then I started experimenting with trafos but never took another look at the output stage!