UTC o-16 Transformer High End Slump

[Dainiusb]

Hi everyone, I've just about wrapped up converting a Berlant tube preamp to stereo with 2 o-16 transformers on the inputs. This thing sounds great! Clean, took the PSU off the chassis, BUT my transformered mic input channels are slumping a bit more than i'd like in the high and low end and I just can't quite wrap my head around why given that these are supposed to actually have a peak in the highs. I've tried putting caps and resistors in different configurations on the secondaries, high shelf, resistance on transformer back end, not even sure at this point just kinda throwing ideas around. When i used the lower ratio input tap the problems get worse, it's almost like the input tube is reading it's impedance as too high? The grid leak resistor has been my main query lately, since there's far less resistance coming back through the transformer winding, why wouldn't that be defining the impedance of the circuit?

I've been running test from my scarlet out which is 96ohms impedance, and I know this thing is tiny so it saturates at 8db, but it does seem to be saturating a bit too early for that.

The UTC o-16 is running 600-50k, the original input transformer was 600-80k, and the grid leak is 1m.

Anyhoo, here's a few images of results and spec sheet info. Let me know if y'all have any input on why this could be playing out like this. Orange line is line level input, purple is transformered input.

 

[abbey road d enfer]

Orange line is line level input, purple is transformered input.

If you look at the line input, it has a capacitors that boosts HF, the 0,00025 across the 100k resistor.
+3.3dB@10kHz, +5dB@20kHz.
The xfmr frequency response quoted in specs is probably measured with a very low capacitive loading, typically 10-30pF.
Here the secondary sees the Miller capacitance of the input tube, about 50pF.
How did you measure frequency response? In particular with what source impedance?

 

[Dainiusb]

Interesting! What kind of filter/boost would that be considered and could I apply that same technique to the input from the transformer?

I measured using REW, with the output of my Scarlett going into the mic in on the Berlant it states it’s output impedance is 94ohms. I had the level set around -50dbm coming out of rew so that it didn’t distort.

I tried measuring the inductance of the secondaries with a weird little LC meter I just got but it kept saying “out of range” which would be above 100H does that seem in the realm of possibility for this transformer?

 

[CJ]

Those Ouncers have never been known for their stellar performance, if anything , people like them for their defects,

 

[abbey road d enfer]

Interesting! What kind of filter/boost would that be considered and could I apply that same technique to the input from the transformer?

I believe I had already answered that, but for some reason it has disappeared. Weird...
This type of circuit is just a 1st-order shelf.
It doesn't actually "boost" HF, it just attenuates the rest, at the detriment of signal amplitude. Applied to a transformer (or just any input) it would impair the noise performance.

I measured using REW, with the output of my Scarlett going into the mic in on the Berlant it states it’s output impedance is 94ohms. I had the level set around -50dbm coming out of rew so that it didn’t distort.

Since you're using it to drive a xfmr designed for a 150-200 ohm source, you would add two 47r resistors, one in each leg, in series. It makes a difference.

I tried measuring the inductance of the secondaries with a weird little LC meter I just got but it kept saying “out of range” which would be above 100H does that seem in the realm of possibility for this transformer?

This xfmr has a quite high ratio (1:14) so the secondary inductance is probably around 500-1000H.