Transformer Suppliers in the UK / EU

[pdf64]

I'm wondering about your statement from the other thread/forum:

>It's worth noting that with the RS Deluxe, the reflected load changes from about 6k4 with a 16ohm load and the secondaries wired for 16ohms, to about 5k5 with a 8ohm load and the secondaries wiried accordingly.<

Is this based on impedance measurements on an original RS Deluxe?

Yes, I liberated an old ‘Amperial’ valve amp from a cellar at work. It might have been a 60s or 70s kit build, perhaps used for the site’s PA / announcement system. I harvested the fantastic pair of Mullard XF4 EL34s from it, and it sat in my shed for a decade or 2. When I finally took the OT off the chassis, it was revealed to be an fabled RS Deluxe.
When comparing signals at the primary and secondary, the transformer has a significant resonance and phase shift between 40 and 50kHz. That's the case both using a sig gen directly as excitation, and when used in a flat-ish response, open loop amp circuit.
The JTM45 sounds great with it, but with the presence turned down and a speaker load, a lot of tweaking was required to get it stable in there, the 47pF cap between the LTP anodes was totally useless, if anything it seems to reduce stability. Eventually settled on 100pF between g1 and k of V4 KT66, and another 100pF across the 27k series NFB resistor.
Dunno how representative it is of the type, but if it is, it makes me wonder if the original JTM45s with it had a stability issue.

I found it a bit stressful to crank the amp with it in, as it’s worth a few hundred quid but is only rated for 30W and 125mA primary current, both of which are exceeded by a fair bit when a JTM45 type amp is overdriven.

[Helmholtz]

. When I finally took the OT off the chassis, it was revealed to be an fabled RS Deluxe.

You seem to prefer 6.6k over 8k, right?
Is it because of sound or output power?

No stability problems with your custom made 6.6k OT? It should show a leakage resonance as well.

Did you see instability with the RS also with speakers or only resistive load?

BTW, do you use the common 1k screen resistor?

[pdf64]

I’m a fan of the 1k shared plus 470 individual screen grid resistor arrangement with the JTM45.
The amp seemed a tiny bit weedy with the 8k 784-103 IG OT. It was totally stable though, no stabilising caps necessary.
The RS Deluxe seemed a tiny bit fatter with 8ohm cabs, a tiny bit more harmonicy when set to 16ohms.
So I had the custom OT made with a 6k primary, as a compromise. I’m pleased to say it sounds great, as good as the RS. And is unconditionally stable, though I’ve left the V4 g1-k 100pF cap in place, just to be sure.
Along with the custom OT, I had a 20H choke made, with about 600ohm resistance, with the hope that it would help to get screen grid dissipation a bit lower. Compared to the 3H 120ohm choke from IG. Which it seems to do, without power output being affected.

The RS Deluxe was borderline unstable with a resistive load, ie just about stable at 8ohms, unstable at 16ohms. That’s with presence at min, as it was advanced stability improved.
Set at or near max, which is how I use it, it was unconditionally stable, the stabilisation tweaks were just me being pernickety.

[Helmholtz]

Along with the custom OT, I had a 20H choke made, with about 600ohm resistance, with the hope that it would help to get screen grid dissipation a bit lower. Compared to the 3H 120ohm choke from IG. Which it seems to do, without power output being affected.

Thanks for your detailled answer.

I like the 20H/600R choke, but if it's not necessary for ripple reduction with small value filter caps, a 3H/120R choke in series with a 470R resistor wouild have the same screen protection effect.

[pdf64]

FYI the RS Deluxe has 3 secondary windings, with respect to the 6k6 primary they’re 1 x 20:1, 2 x 40:1.
16 ohms is all in series, 8 is the two 40:1s in parallel, in series with the 20:1.

[Helmholtz]

FYI the RS Deluxe has 3 secondary windings, with respect to the 6k6 primary they’re 1 x 20:1, 2 x 40:1.
16 ohms is all in series, 8 is the two 40:1s in parallel, in series with the 20:1.

Thanks, I have a Marstran RS clone with secondary wiring diagrams. Haven't tried it yet in my JTM 50.

[loudthud]

FYI the RS Deluxe has 3 secondary windings, with respect to the 6k6 primary they’re 1 x 20:1, 2 x 40:1.
16 ohms is all in series, 8 is the two 40:1s in parallel, in series with the 20:1.

The "8 Ohm" arrangement actually works out to 9 Ohms. See this thread for more info: https://music-electronics-forum.com/...ares-deluxe-ot

[Helmholtz]

FYI the RS Deluxe has 3 secondary windings, with respect to the 6k6 primary they’re 1 x 20:1, 2 x 40:1.

These should be the voltage or turns ratios for one half-primary (1600 Ohm).

16 ohms is all in series, 8 is the two 40:1s in parallel, in series with the 20:1

I think an RS OT has one 4 Ohm and two 1 Ohm secondaries.

BTW, if you calculate primary impedance from the voltage ratios, the result will be a little too low.
The difference is the sum of primary and reflected secondary DCRs (maybe 200R for the RS) which adds to the load as seen by the tubes.

I prefer to measure primary AC resistance (secondary loaded with rated resistance) with my LCR meter

[pdf64]

These should be the voltage or turns ratios for one half-primary (1600 Ohm).




I think an RS OT has one 4 Ohm and two 1 Ohm secondaries.

BTW, if you calculate primary impedance from the voltage ratios, the result will be a little too low.
The difference is the sum of primary and reflected secondary DCRs (maybe 200R for the RS) which adds to the load as seen by the tubes.

I prefer to measure primary AC resistance (secondary loaded with rated resistance) with my LCR meter

Oops, apologies, you’re correct of course, I wrote previously from memory.
From my notes, the voltage ratios are 80.22:1 and 40.11:1.
The ‘1 - 9’ 6k6 primary resistance is 161ohms.
The ‘A - F’ 40.11:1 (4ohm) secondary resistance is 0.48,

[Helmholtz]

The ‘1 - 9’ 6k6 primary resistance is 161ohms.

Makes sense. The original RS spec sheet specifies 175R max, probably relating to the 8k primary.

The ‘A - F’ 40.11:1 (4ohm) secondary resistance is 0.48,

Now that looks rather high even considering, that the 4Ohm secondary won't be used alone.
If we assume a total secondary DCR of 0.24R in 4 Ohm configuration, this would result in additional 384R at the primary.

Resistance below 1R is tricky to measure and requires calibration and 4-wire (Kelvin) measurement.

EDIT: I just measured the DCR of the single 4Ohm secondary of my Marstran RS clone as 0.45R. The two 1 Ohm secondaries measure 0.22R each.
So your DCR is realistic (sorry).
Using my DCRs, the 16 Ohm configuration has a DCR of 0.89R, reflecting to 356R at the full 6.4k primary or 89R at each primary half.
This together with the primary DCR means a lot of copper losses, making the low efficiency specified by inMADout a bit more reasonable than I thought.

[pdf64]

Sorry, something cropped up and I posted before I'd quite finished up last lime.
The 'B-C' and 'D-E' 80.22:1 (1 ohm) secondaries each measure about 0.24 ohms.

Measurements taken with my Fluke 189, using the 'REL' function to tare out residual lead and probe resistance.

Many thanks for the points about primary and reflected secondary resistances adding to the total effective primary impedance - I'd overlooked that

... I prefer to measure primary AC resistance (secondary loaded with rated resistance) with my LCR meter

Wow, I didn't appreciate it would read that ok.