5 volts from a 6.3 volt tap for Rect tube

[pjd3]

So, this did seem to do the trick in case anyone was ever considering having to do this.

Originally, I was going to use SS rectification so ordered a PT with no 5 volt tap. When changing plans from KT88 to EL34's I thought a lower B+ should be in order so having three 6.3 coils on the new PT, looked into making a 5 volt supply from one ot the 6.3 coils for a GZ34 rectifier tube.

It was just some ohms law (1.3 volts divided by 1.9 amps = .68 ohms). I wired a 1 ohm and 2 ohm 5 watt wirewound in parallel which gave me close to that - .66 ohms mathematically. I was actually a bit scared not being 100% certain if this would interact with pin 8 (GZ34) being the high pulsed DC output but just wired in the resistor at pin 2 of the rectifier, away from the high V out of the tube.

No smoke, no popping with 1.3 volts across the resistor and 5.6 volts across the actual rect filament. The voltages were all notable high for all the secondary coils, even the HV coil was at or over 450 volts for a 335-0-335 rated coil (Edcor xxx96). I assumed that this was due to 124 ac going in from the wall combined with no loading from tubes and OT's as of yet.

The resistor (or double parallel resistors) did get a bit hot after a while and for a final build I would go with a single 10 watt/.68 ohm resistor but, this approach does seem to work. This topic has come up in a thread that I was a part of and got alot of good feedback that it should be a viable approach. I was left decently confident that this would work. My only warning was to keep the pin 8 rect output in mind as it is shared with the filament. So I did!

Best,
Phil Donovan

[dorrisant]

In the old AAR5 radio circuits they did this... used series resistance to get what they needed for the filaments. I don't see why your method won't work. Maybe use a 20 watt resistor, like a Dale with aluminum extrusion... same amount of heat just more robust.

[pjd3]

Hey Dorrisant, well it is good to know that others have done this as an accepted part of the plan. I hadn't actually heard that anyone else had actually done it so, it was all theory, at least to someone in my early stages of the game on amp "design" (hopefully we'll be getting to a little of that soon). It did teach me something about the changing resistance of filaments at room temp vs hot.
The filament on the GZ34 measured a dcr of about 5 ohms cold, and ohms law calculated it to be around 15 ohms once heated and running. I had to check this because the calculations of the expected filament current didn't make any sense based upon using the cold 5 ohms as the resistance. I had heard this happens but, now I see it in action.
And I'll look into those Dale 10 watt resistors you recommended. Aluminum extrusion. Hmm, interesting.

Of course, there is always a chance that I will want more output volume for this particular application should I find a more Hi-Fi circuit compliments the application. I would expect a drop on volume due perhaps to the more NFB going on in hi fi amps. And then be curious to try a pair of KT88's with solid state rectification. Could happen! But, this whole venture is to force myself into trying many different circuits with various mods and additions. Seems to be working so far.

Thanks Dorrisant,
Phil Donovan

[Phil_S]

This has been done before. There are a few problems with it and I think you already know what they are:
1) Line voltage will affect the result. With 124VAC, the result will be above 5V.
2) The solution makes a lot of head. I second Tony's recommendation for a high wattage resistor that will help you dump the heat into the chassis. Make sure there is a "zone" around that hot resistor.
I suggest, as long as you are within about 10% of 5V (4.75 to 5.25V) the series resistor solution should be OK.

Are you sure you want to go to this trouble? Why not use a 6.3V rectifier? I don't know where you are in the build, so maybe this suggestion is no longer in play. The problem is getting enough current for a pair of EL34's from a 6.3V rectifier. Without knowing what you built, I'm going to assume you'd need a pair of EZ81's or a pair of 6AX5-GT's. For a pair of EZ-81's you'd need 2.0A filament supply and for a pair of 6AX5GT's you'd need 2.4A. The 6AX5-GT's if you can find them, will drop 50V or so. The EZ81's will drop about half that.

[thetragichero]

i'm with the 6.3v rectifier solution. there's something about a big old resistor getting hot that doesn't sit with me the right way, especially if it's an imperfect solution based on voltage on the primary

[dorrisant]

i'm with the 6.3v rectifier solution. there's something about a big old resistor getting hot that doesn't sit with me the right way, especially if it's an imperfect solution based on voltage on the primary

Agreed that the 6.3v recto would be the easy way, but... that big old resistor will not get as hot as the tube itself.

Like I said, it was done for millions of little AAR5 amps. I see them come in and cringe at the widow maker aspect (with no PT), but I have seen very few if any of those dropping resistors fail, even at 5w. Check out the example below.

I'm not saying it is the preferred way to do it but it was done in production without much problem, so it would certainly be safe to try it out.

[pjd3]

I must admit, it was with some trepidation that I tried this out, and this configuration is probably only for my "tube amp design test station". When it comes time to actually build the amp for my own performance use, I will probably bite the bullet and order one or two PT's with 5V filament coils.

I did originally think that this amp was going to have SS rectification and it very well may (if I decide I want a little more kick and go KT88's, which was my original idea).

But, for experimental purposes it should be fine, and I did learn a little something, that this is a possibility, if not always a popular one. it doesn't surprise me that it has been used in the past in some commercial devices. And should I end up using this PT for the EL34 solution, I would use those nice gold metal extrusion resistors like we have in work here at my day job. And of course, there would be plenty of real estate around the resistor to let it vent some.

Thank all,

Best,
Phil Donovan