DC on toroid PT windings

[audiowize]

Hello all,
I am trying to build a high voltage doubler by using half of the primary of a toroidal transformer with the secondaries in parallel. The transformer I purchased says no more than 250v DC is allowable between windings (B+ final is around 315v, and the doubler puts half that into the winding, so it seems OK...).

Anyway, I wired everything up and the circuit "works" as wired, but the transformer is noisy and seems to be radiating unreasonable amounts of noise. Am I being overly optimistic?

Any ideas???

[jechentau]

if the whole primary is rated for the mains voltage local to you, the transformer saturates. idk how to say better than that the core of the transformer cannot get more magnetic to make the swing of the input voltage, thus is saturated, and then primary inductance drops sharply, and quite large peak currents flow, with accompanied dissipation, and noise.

afaik

full primary and both secondaries in series should give about same voltage, and within range of the transformer.

[disco]

Toroids rarely behave noislessly due to loose windings. Like all transformers its permissable power is calculated to stay away from core saturation.

[jechentau]

well, there is a relation of voltage swing, and frequency.

voltage to high, or frequency too low, and saturation occurs on the peaks.

some transformers rated for a given voltage at 60 Hz will run into trouble operated at 50 Hz.

in terms of current: if the frequency is too low, or the voltage too high, the reactive current through the inductor exceeds the cores capability, above it, one may say the current becomes resistive and "real". the capability to transfer additional useable current has then reached 0.

hence on ships (mainly in pre-smps age), they use 400 Hz, for for a given power, one needs way smaller transformers.

[Miles Prower]

Originally Posted by audiowize

Hello all,
I am trying to build a high voltage doubler by using half of the primary of a toroidal transformer with the secondaries in parallel. The transformer I purchased says no more than 250v DC is allowable between windings (B+ final is around 315v, and the doubler puts half that into the winding, so it seems OK...).

Anyway, I wired everything up and the circuit "works" as wired, but the transformer is noisy and seems to be radiating unreasonable amounts of noise. Am I being overly optimistic?

Any ideas???

What kind of voltage doubler? If it's a Walton/Cockroft, that's a half wave type, and you're putting DC on the coil. Toroidals really do not like DC, as there are no incidental air gaps to lessen the likelyhood of DC magnetization, and therefore core saturation.

[audiowize]

You've got it miles, the doubler puts DC on the primary winding that I am using with the doubler. If I switched to an EI type transformer (simple isolation transformer), am I likely to run into the same problem?

[Eli Duttman]

Keep the toroidal power trafo. Switch to Greinacher, AKA full wave, doubler topology. Not only is DC kept out of the trafo, the PSU ripple freq. is 2X the mains freq.

[Miles Prower]

Originally Posted by audiowize

You've got it miles, the doubler puts DC on the primary winding that I am using with the doubler. If I switched to an EI type transformer (simple isolation transformer), am I likely to run into the same problem?

Maybe yes, maybe no. If you need to double the voltage, your best bet would be a full wave doubler that doesn't draw DC through the coil. W/C multipliers are usually used where the current demand isn't all that much. Still, I've seen these things saturate E/I xfmrs as well. Had that very problem when doing a project that included a W/C tripler with a "doorbell" xfmr. Even though this was for a bias supply, and didn't pull more than a milliamp, there was still DC core saturation that pulled the output voltage down to about 5.0V, instead of the ~36V I was expecting. The only way to make that work was to pull out the lams, and gap the core. Of course, being such a cheap little xfmr, it was NBD to do that, as there was no potting used in the construction.

Or take a look at TV flyback xfmrs: these are invariably gapped core devices since the HW or W/C rectifiers will pull DC. Even if it's just a fraction of a milliamp, it's flowing through thousands of turns, and there will be enough magnetization to saturate the ferrite cores.

[audiowize]

Originally Posted by Eli Duttman

Keep the toroidal power trafo. Switch to Greinacher, AKA full wave, doubler topology. Not only is DC kept out of the trafo, the PSU ripple freq. is 2X the mains freq.

I need to ground the V- of the doubler, so I end up with 1/2 of the B+ on the transformer...

[jechentau]

do you happen to have a schematic of your circuit?

and how did you mean half the primary?

about voltage multipliers; how can there be DC current if it is capacitor coupled;
transformers only need to be gapped if the current is not averaged 0, in both directions which in the case of capacitor coupling is not so.

when i used a doubler i placed one end of the transformer on the center point of 2
caps in series, and the other end with one diode to ground, and the other to B+.

that way, the caps each get charged with a half cycle, and since they are in series, ripple is still 2x mains frequency, and in effect is full wave rectification. (the caps in series i had anyway, because they are only 200V each)

(added link) http://en.wikipedia.org/wiki/Voltage_doubler

my actual high voltage transformer currently is actually a isolation transformer, but with a full wave bridge this time.