Weird voltage readings on filament windings

[Bloomfield]

Hi, I've just taken a project down off the shelf and got back into it. It's a new build but the power transformer is a mystery item that I found somewhere years ago. I wired it up today and took some readings. The HT is about 320-0-320 VAC which is fine. There are two filament windings, both of which measure around 7 VAC with no tubes plugged in. That seems fine, but when I measure the ends of the windings with respect to ground I get 186 and 190 VAC on one and about 266 and 271 VAC on the other. Neither filament winding is centre-tapped. What is going on here? All of the filaments seem to be isolated from one another, ie no continuity between any of them. Just to make sure I checked another amp and each side of the filament winding read 3.25 VAC to ground, or about what I'd expect.

[Helmholtz]

That should be "ghost voltage" caused by the capacitance between heater and HT windings in the PT.

Shunt your meter with a 100k resistor and measure again. Voltages wrt ground should be much lower now.

Then ground one end of each heater winding and check again.

Heater voltage should always have a ground reference (except rectifier heaters) to lower heater hum, means either a grounded CT, one end grounded or ground reference/balancing resistors.

[Bloomfield]

Thanks Helmholtz, that shunt did the trick and the readings are more sensible now. The one winding can't be referenced to ground as it for the rectifier; had to put 0R47 in series with the 5U4 which put it at 5.2 V and I am getting 430 VDC at the first filter (unloaded) so things are looking promising for now.

[J M Fahey]

The one winding can't be referenced to ground as it for the rectifier; had to put 0R47 in series with the 5U4 which put it at 5.2 V

WHAT???????
Are you using the regular 6.3V winding for the 5U4 rectifier too?
DISCONNECT IT **NOW**!!!!!

Direct heater rectifiers WHICH CARRY DEADLY +V ON THEM were built with 5V rectifiers SO YOU CAN **NOT** CONNECT THEM TOGETHER WITH OTHERS but only to its own FLOATING HIGH VOLTAGE INSULATED WINDING.

And we are talking 1kV or better insulation here, NO 6.3V winding is built that way.

What you did is very dangerous.

Use a 6.3V indirect heated rectifier there o 1N4007, not 5U4 or any direct heater rectifier there.

Again: there is a reason they made them "incompatible" with others.

[Chuck H]

Good catch Juan! Good looking out.

[Bloomfield]

Where did I say it was connected to others? There are two independent filament windings.

[Chuck H]

Where did I say it was connected to others? There are two independent filament windings.

I think that post was just a safety precaution in case there were any misunderstandings. Juan interpreted that you may have the rectifier sharing a filament supply with other tubes. I'm sure he's glad to be wrong about that.

But I WILL add that by adding the .47r resistor in the rectifier filament supply to bring it down you have effectively added to the loading on that winding and the resistor dissipates watts in heat. That winding is supposed to be at 6.3V when loaded as rated. I may be wrong but it may be that to get it down to 5.2V you are effectively over loading that winding.

EDIT: The way to tell would be to measure the voltage across the resistor AND the rectifier filament. If it's close to 6.3 you're fine.

[Enzo]

Oh, the winding only sees its load, it has no idea if it is one resistor or five resistors or a mix of resistors and tubes.

[Chuck H]

Right. So what IS the load? I suppose I could look up the tube data and do some ohms law math, but see edit above. We know how much voltage is ACROSS the rectifier filament. So we just need to know how much is ACROSS the .47R resistor. If the sum is around 6.3V it's all good. Unless I have this wrong.

[Helmholtz]

It's not about sharing heater windings.

A rectifier heater winding that is floating on high voltage requires a higher level of PT internal insulation/isolation to/from other windings and PT core than a normal heater winding.
Means different internal construction requirements.
Requirements for HT winding and rectifier heater winding should be essentially the same.

[Enzo]

Chuck, the voltage is always going to be whatever the transformer makes. A series of more than one thing - say a tube and a resistor - will voltage-divide itself. The tube will not automatically drop 5v across itself. This is very much the same problem as running an LED from a supply and calculating the series resistor needed. SO if the tube heater needed 2A at 5v, and we have 6.3v supply, then we calculate what resistance will drop 1.3v at 2A of current. (Yes, I ignored all the little details like what real world voltages we have rather than ideal)

[J M Fahey]

Where did I say it was connected to others? There are two independent filament windings.

Separate but NOT PROPERLY INSULATED, BY FAR.

Read gain:

its own FLOATING HIGH VOLTAGE INSULATED WINDING.

Minimum insulation is 1kV DC and that is edge of the precipice ; I bet HIPOT tests use at least 1500V DC if not more.

Here, a Hammond 5V (so intended for 5U4 and such) filament tester:

https://www.tubesandmore.com/product...ent-open-5-vct

Hi-Pot test of 2,000 V RMS.

Do you even DREAM a regular 6.3V filament winding is THAT insulated?

By the way, "helmholtz" also hinted at that, even before you even mentioned a 5U4:

(except rectifier heaters)