Williamson amplifier. Winding an output transformer.

[trobbins]

Is your Goodsell OPT like the attached image of a double C-core from a GW-18 ?

I also have a GW-18 manual that indicates a standard 15 ohm output setup. The photo shows eight secondary sections linked as in a Partridge OPT for 15 ohm with 0.95 ohm sections. The other wiring for primary side is somewhat indistinct but appears to show P,B+,P wires and possibly two other wires for screens.

The GW-18 product manual appears to be from after May 1952, which would align with the use of UL screens (ie. 1952 onwards) and double C-cores (given Partridge CFB was introduced circa late 1950, and the UL version T/CFB with 43% taps was out by 1954 - I can't date its introduction).

[thermion]

Yes, that’s very similar, I recognize the gold colour too.
Now I’ve taken some measurements this afternoon, & the results below.
Resistance tests;
Anode to Anode 203r
CT to anodes equal @ 105r
Screen to screen 94r
CT to screen taps unequal. 1@54r the other is 40r
Voltage test. 10V AC on secondaries. With 3 sets of windings in series, 3+2+3 parallel sets;
Anode to anode 340v balanced ok @ 170v per side
Screen to screen 147v balanced ok @ 73.5v per side
With 5v ac anode to anode leakage current 0.08 mA.
So the turns ratio with 3 groups of secondaries in series is 34:1, squared is 1156, into a 10k load gives 8.6 ohms. Is that right?

[thermion]

I would say there pretty much the same, though the tagboard arrangement is different. Tried to take an impedance reading across the whole primary, A to A but couldn’t get a null. It sends my fluke nuts to & wont auto range. Here’s the transformer on the amp I was measuring earlier.

[trobbins]

I concur that the 3+2+3 secondary configuration aligns with 8.5 ohm impedance for 0.95 ohm sections of a Partridge.

Your pri and sec voltages align with 10k PP and 8.5 secondary. The screens align with 43% turns percentage. The primary inductance aligns with 200H +/-25H at 5Vac and 50Hz, which is a good result, and shows the advantage of having C-cores available.

You may get better turns ratio accuracy by driving part of a primary section such as A to Sc, and measuring the other A to CT, and the sec voltage, as the measured voltages don't then include the generator driving current - but may depend on the voltage measurement resolution you have available.

[thermion]

I think that’s all we need for now, transformer is with Ed being rewound as per Williamson spec with additional screen taps. I’m trying to remember what the original goodsell amp looked like before I broke it up. I remember it had another lower powered mains transformer supplying an output for something else. The only surviving part now are the transformers and the metal cover over the back . Send me a pic of the whole amp, mine had been hacked about before I got it.

[trobbins]

I'm not sure where I grabbed these GW-18 photos from back in 2016. Maybe the serial number will confirm for someone.

PM me for higher res, as these have been reduced from 1440x1080.