• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Surplus toroid transformers as OPTs...with a little twist.

Thank you George, that is an interesting thread! I just remembered that Nelson Pass used big dual primary toroids as "plate chokes" in one of his Son of Zen versions. IIRC that was an LPT with a CCS in the tail, working as an balanced or self split output stage. Perhaps it could be done in a thirties style with a huge choke instead of the CCS... :)
I bought an E-I isolation xfmr from Jameco some years ago, made in China, cheap. I was doing some C measurements on some xfmrs then to determine usefulness for floating Circlotron supplies. I was alarmed at the high primary to secondary capacitance this xfmr had (like 1500 pf, can't exactly remember), so I took the end bells off to look. No insulation between primary and secondary, wire on wire. Aside from the safety issue for which one uses an isolation xfmr, one expects some capacitive isolation too. I measured a Magnetec N90MD medical isolation xfmr and got just 121 pF primary to secondary. At least an order of magnetude less. A Magnetec-Triad N-66A Isolation xfmr measures 320 pF (made in China). All models 250 VA.

I checked an Avel-Lindberg 135 VA toroid and got 20,000 pF from pri to pri (obviously bifilar primaries). And 878 pF from both primaries to both secondaries.

I also checked a Stancor TGC175-230 E-I. Four 120V windings but a split bobbin, 175 VA. 322 pF from primary to primary, on same bobbin side (stack wound). 82 pF from primaries to both secondaries, across bobbin sides. And 18700 pF from sec1 to sec2 on 2nd bobbin side (bifilar wind). With only 82 pF from secondaries to primaries, this model (2X) got selected for the Circlotron floating supplies.

I also measured an Edcor 100 Watt CXPP100-MS-1.7K OT and got 3450 pF from primary to secondary. I guess all the interleaving does that.

Also checked a Toroidy TTG-KT88PP 80 Watt 4K toroidal OT, 4000 pF from primaries to secondaries.
And a Toroidy TG-EL34PP 50/60 Watt 6.6K toroidal OT. 2680 pF from primaries to secondaries. Both models on the same size core.
Last edited:
I measured the HF resonance for the Toroidy 4K and 6.6K PP OTs, both similar at 100 KHz. Using a Signal Gen with 10 K Ohms in series and a scope to look for the HF AC voltage dipping to a minimum. Measured between B+ lead and a primary end. (measured the same for either end)

I then checked the Edcor CXPP60-MS-4.2K and found the HF resonance at 45 KHz.

Also checked that Signal Corp DCP E-I 24 Watt xfmr 50/60Hz for HF resonance, and found 105 KHz. That only has enough turns for 6 Watts at 30Hz however.

On the low frequency end, the Toroidy 4K does 46.6 VAC across the 8 Ohm winding (using a
Variac and current probe) for an arbitrary acceptable level of saturation current spiking. For the same level of current spiking, the Edcor does 53 VAC across the 8 Ohm winding.

That would translate to LF of 32.3 Hz for the Toroidy (at 80 Watts) and 25 Hz for the Edcor (at 60 Watts). If the Edcor VAC is (arbitrarily, but bigger spiking ) increased to 65.3 VAC to get the specked 20 Hz at 60 Watts, and the Toroidy is increased by the same V ratio, the Toroidy would be at 26 Hz LF (80 Watt). And if the 80 W Toroidy were derated to 52 Watts it would reach the same 20 Hz LF limit as the Edcor. So roughly identical OTs, except for the doubled HF resonance for the Toroidy and the Edcor handling the LF end better.

My opinion. for acceptable current spiking at 20 Hz LF, both of these OTs are more like 40/50 Watt rating. For 25 Hz LF, both are roughly 60/75 Watt rating. Which just goes to show that you should always overate the OT Watt requirement some if you want real HiFi performance at the LF end.

Putting in the current derived Positive FDK fix (removing primary resistance and magnetizing current distortion feed-thru) could remove the Watt restrictions.
Last edited:
  • Like
Reactions: 1 users
Conventional N Fdbk (V derived, local or global) would also allow using the full Watt ratings at the LF end too.

The Toroidy 4K PP (KT88) could also be looked at as a 3K to 6 Ohm OT for 80 Watt at 28 Hz.
The Toroidy 4K is clearly somewhat short on turns compared to the Edcor for the LF end. Like 0.88 the turns judging by the 8 Ohm Voltage readings for similar current spiking.
The Toroidy 4K full primary measures at 62 Ohms DC and the Edcor 4K OT at 260 Ohms for the full primary. The Toroidy still has plenty of primary inductance though due to the higher effective core permeability.

The 80 Watt rating versus the 60 Watt rating would also be expected to lower the primary Ohms by 3/4.
So taking into account the 0.88 V and 3/4 W factors: Apples to apples comparison would be 94 Ohms Toroidy versus 260 Ohms Edcor for exact same OT. The toroid config obviously saving copper, although the toroid seems a little oversized too, which would save more copper (square edges on the Toroidy, not a lot of wire there). Designs can trade off copper versus steel material.
Last edited:
  • Like
Reactions: 1 user
late edit
the 0.88 V ratio above should be amended to 0.76 due to the higher V needed for 80 W versus 60 Watt at the LF end. So the Apples to apples compare would be 0.76 (V) x .75 (I) = 0.57. Giving 108.5 Ohms Toroid versus 260 Ohms E-I for equivalent OTs.

I would guess that most power type toroid xfmrs are not so heavily iron versus copper oriented as the Toroidy seem to be, They usually have a more round cross-section. You might see more like 1/2 the Ohms versus the E-I equivalent using them.
Last edited: