Ultra-Linear Measurement Transformer

Design and use of the various types.
Post Reply
hifizen
Posts: 2
Joined: Tue May 28, 2013 10:24 pm

Ultra-Linear Measurement Transformer

Post by hifizen »

Hi Dave,

I'm looking for a custom push-pull tube output transformer for ultra-linear measurement purposes. However, I am not after a tapped primary. What I am looking for is a tapped tertiary winding for the screen grids, which can be connected to a lower DC supply voltage than the plate / primary winding.

So, I'd like to discuss the basic feasibility / sanity of this, and approximate cost, or any ideas to make this more feasible or economical. Here is the basic outline of what I'm thinking...

This will be for use with various power pentodes, with plate dissipations around 25W to 50W each, so the transformer should be capable of handling 80 to 100W peak. EL34, 6L6GC, 7027A, 6550, GU-50, etc. maybe some 6146 or sweep tubes, we'll see.

Primary Winding:

Plain center-tapped winding. Impedance should be around 6.5k-ish plate-plate. I will be loading the output winding with a variable load in order to test at different plate load impedances, probably over the range of 4k to 10k or so.


Screen grid winding:

Obviously will not need to handle as much current as the plate winding. Preferred taps would be:
impedance turns
CT 0
0.025 0.158
0.05 0.223
0.1 0.316
0.15 0.387
0.2 0.447
0.25 0.5
0.5 0.707
1.0 1.0


Load winding:

This can be a single 8-ohm winding, but I am open to discussing a more suitable impedance if this would ease the design constraints. While I would like to be able to test with real speakers, the main purpose is measurement, so I am first and foremost interested in using this with a variable test load on the bench. Output voltage and HV isolation will need to be compatible with a distortion analyzer etc, although I can pad down the output signal within a reasonable range. Another option would be to have a measurement winding with minimal loading, placed for optimal coupling to the primary. The load would then be applied elsewhere - perhaps directly across the primary or on some other winding.

- Chad.
hifizen
Posts: 2
Joined: Tue May 28, 2013 10:24 pm

Post by hifizen »

... and I just realized, another possible approach to this would be to make an autoformer with all the taps for the screen grid, and simply connect this across the plate winding of a standard output transformer using caps at each end and at the center tap. But I am not sure this is really any better in terms of emulating the ultimate goal, which would be a single transformer with dedicated UL screen grid tertiaries. The coupling of such an arrangement is certainly not going to be quite the same as wrapping all those windings around a common core.

I should also add that bandwidth requirements are relaxed since this is for measurement purposes. But I do intend on running SMPTE measurements, a solid 60Hz and 7kHz would be needed, more if possible.

- Chad.
Post Reply