Dropping filament voltage with diodes

[Bloomfield]

I have an old power transformer that I am using in a build. It has two filament windings (both 6.3v) and I would like to use one for a 5 volt rectifier. Searching around on this site I saw a recommendation to use a pair of silicon diodes in anti-parallel in series with the heater. I rigged up an old bridge rectifier that I had and this brought the voltage down from 6.85v to 6.0 using a 5U4. It looks like a second pair of diodes ought to get the voltage about right.

My question is what voltage and current rating do I need for the diodes? The 5U4 draws 3A so I would assume more than that but I'm not sure how the series/anti-parallel arrangement factors into that. Also the B+ voltage (350 or so) is on the filament, so should I use 1000v rated parts?

[Tony Bones]

With two diodes back to back, the PIV never gets above 1.5V or so, so that's not an issue. Just make sure the current rating is adequate. Sometimes with high currents it makes sense to buy a bridge rectifier instead of trying to use individual diodes. You can wire it to drop voltage the way you want.

Having said that, I would use a power resistor instead. The current is more or less constant, so a resistor will work fine and won't introduce all of the switching noise of silicon diodes. R = V/I = (6.3-5)/3 = 0.43 ohm. If it's an old transformer meant for lower line voltage then you might need slightly more resistance to get down to 5V.

[Bloomfield]

Thanks Tony, the resistor was my first thought but I was worried about adding extra load to the winding. Maybe it's not such a concern.

Adding a 5v filament transformer would be the best solution, but it would mean losing the choke for the screens so I'd rather not if possible.

Andy

[Helmholtz]

the resistor was my first thought but I was worried about adding extra load to the winding. Maybe it's not such a concern.

The series resistor means no extra load for the winding. The winding just feels the (3A) current and has to supply ~19W (its terminal voltage times current). Just make sure it is rated for 3A. The resistor should be rated at least 5W continuous and 50W short time, peferably wirewound.

[Bloomfield]

Thank you for clearing that up Hemholtz. I will just go with a resistor then.

[J M Fahey]

Please be careful
5V rectifier windings are not same as "a standard 6V filament winding, just putting out 5V" by any means.
Filament windings expect to live grounded or worst case elevated a few Volts; rectifier ones routinely expect up to 500V RMS applied to them, so 750V peaks on a good day, so are specially wound apart from all other windings/core/ground by at least 1000V insulation; maybe 1500V just to play it safe (and it actually costs the same).

In fact I think 5V was chosen for rectifiers on purpose so nobody even *thinks* about connecting them to the regular 6V winding and viceversa.

[Tony Bones]

Please be careful
5V rectifier windings are not same as "a standard 6V filament winding, just putting out 5V" by any means.
Filament windings expect to live grounded or worst case elevated a few Volts; rectifier ones routinely expect up to 500V RMS applied to them, so 750V peaks on a good day, so are specially wound apart from all other windings/core/ground by at least 1000V insulation; maybe 1500V just to play it safe (and it actually costs the same).

In fact I think 5V was chosen for rectifiers on purpose so nobody even *thinks* about connecting them to the regular 6V winding and viceversa.

Yeah, that's a popular theory, but I've never found any evidence to support it. "Modern" transformers need to withstand 1500V hipot between windings and core. Even if we're not sure "modern" means, 20 mils of kraft paper will withstand at least 3KV as long as it's been dunked in varnish to prevent absorption of humidity. Bottom line, there's no reason to treat one winding as special; they all need to be sufficiently insulated from any other. That's just a few wraps of kraft paper in any case.

The transformer designs that I've studied pay no particular attention to the 5V (or any other) winding. They just put "adequate" insulation between windings and core and call it good. But, if you know of exceptions, then I'm eager to know about them.