Can I phase split like this?

[Sherman]

I recently obtained (for no good reason other than the price was right and they looked cool) a couple of interstage type transformers. The person I got them from indicated they were 2.5K:10K with the primary center tapped but not the secondary. My guess is that they can't handle any DC, they are very small. I am wondering if I might be able to use them as input phase splitters in an EL84 PP amp.

My idea is to 'turn them around' and use the secondary as the primary connecting one leg to a volume pot and the other leg to ground, then connect the primary (now used as the secondary) to the grids of a driver stage, grounding the center tap and driving the EL84s directly from the drivers. The driver tubes could be 6N1P's or 6922's since I have several of each on the shelf. (I think I have some 12AX7's also.)

The attached schematic gives the basic idea. Any feedback would be appreciated.

[EC8010]

In (simple) theory, you can do this. In practice, the transformer will object to being driven from the high (and variable) output resistance of your volume control, and show this by an HF response that changes with volume setting. Reducing the volume control to 10k might help, but will load your sources more heavily.

The only real way of knowing would be to try it and measure it to death.

[pmillett]

I think I'd try using the 6922 in front of the transformer, as an SRPP stage. Cap-coupled to the transformer.

Or, add a low-gain stage in front. I've used 12B4's like this and they do well. Or a cathode follower if you don't need any gain.

I'd bet you'll get really bad (and variable with volume control position) frequency response if you try driving the transformer directly from the volume control.

Pete

[Johan Potgieter]

So would I. Also, for sufficient bass the transformer inductance would need to be high enough not to shunt the source. Otherwise good advice; I would also consider feeding from a cathode follower.

[Sherman]

Johann and Pete,

Thanks for the replies! This is one of those projects that just sort of happened since I had the parts and I want to gain a little experience with phase splitters.

Since wiser, more experienced heads than mine are of the opinion it won't work acceptably as drawn and since a cathode follower would be an easy addition, requiring only 1 more tube for a stereo amp I will probably go that route.

Here's an updated schematic. I'm leaning toward a B1+ of 325V (for the output stage) and a B2+ of 250V (for the cathode follower and drivers).

[Johan Potgieter]

Quite so, Sherman,

Except that you might have to elevate the dc level of the cathode follower g1 above common in order to get the right current through the cathode resistor (right operating point for the c.f.).

[Shoog]

I wouldn't bother with the cathode follower - its not needed to get the best out of the transformer. An anode follower will be simpler and will allow you to compensate for the step down of the transformer.

Because you are stepping down the transformer will behave massively better than normal with a much wider bandwidth,
I would consider the possibility of eliminating the 6922's as drivers by using a front end valve with higher gain in a Mu follower.

Shoog

[Sherman]

Originally posted by Shoog
... An anode follower will be simpler and will allow you to compensate for the step down of the transformer.

... consider the possibility of eliminating the 6922's as drivers by using a front end valve with higher gain in a Mu follower.

Shoog

Shoog,

I hadn't considered that the performance of the transformer might increase with the step down. That's good news.

I have some questions on this approach if you don't mind.

1- Where to put this stage?
A- before the transformer
B- afer the transformer and before the driver stage
C- after the transformer and eliminate the driver stage

2- I'm a noob when it comes to anything beyond grounded cathode, cathode follower or white cathode follower. In poking around I'm seeing plate followers (using a single triode and giving a gain of maybe 22dB with a 12AX7 and mU followers using two triodes (and to me they look like SRPP stages?) with gains of maybe 36dB with a 12AX7. I'm assuming you mean the higher gain stage?

Here's a shot at it and thanks for the input!

[Shoog]

The plan you have shown looks workable.
Two small issues, in a Mu stage you don't have the cathode resistor above the take off point on the upper triode. The Mu stage gives gain and most of the drive characturistics of a cathode follower.
Also the way you have it drawn the top triode has no grid reference and so will not work.
A good analysis of the Mu stage can be found at;

http://www7.taosnet.com/f10/mustage.html

What is the voltage ratio you will be getting with this transformer ?
You need this to work out if your Mu stage can give enough gain to drive the EL84's directly. If it is possible to achieve this then you save a valve, and also you have superior drive to the grids of the EL84's which will allow it to better transition into AB2 class.


Shoog

[Shoog]

i have been toying with a very similar idea myself using ECL82's and 807's.
However you plan as drawn has one very serious design flaw and that is its poor damping factor. It is very difficult to introduce global feedback into a design with an interstage transformer due to stability issues.
There is a clever fix to get round this. If you keep the 6922 driver stage you can introduce plate to plate feedback ala RH84.
Another subtle refinement is to turn the 6922 into a LTP with a common CCS in the tails.

Also if you do go with the Mu stage, consider the alternative of a CCS loaded anode follower. This potentially should outperform the Mu stage by a long margin.

All in all that would make an excellent little performer.

Shoog