Zenith radio filament question

[Randall]

I am working on a Zenith H 732 Z radio that shorted a couple of caps and burned up some of the PS. Got that working when I noticed someone had modded the filament circuit. The selenium rectifier had been disconnected on the + side and the wire that was there is now connected to the - side. Also, a large 33K ohm resistor is connected from filament pin 4 of the 35C5 output tube to ground. The radio works, but I notice that 33K is getting warm to almost 200 degrees after being left on for 15 - 20 mins or so.

So I ask is it OK to feed unrectified voltage to the (newly replaced) filter caps? Or should I put a modern day rectifier in there? What is the 33K to ground all about? An attempt to lower filament voltages?

Here is the link, schematic is on page 4. Second one is my drawing of how it is now.

https://www.doctsf.com/documents/aff...8581&num_fic=1

[The Dude]

I've found this to be a useful article on the subject.

http://w3hwj.com/index_files/RBSelenium2.pdf

[Tom Phillips]

...I ask is it OK to feed unrectified voltage to the (newly replaced) filter caps? Or should I put a modern day rectifier in there?...

You definitely need a rectifier. Yes use a Si rectifier and reference the article that the Dude posted for overall guidance.

...What is the 33K to ground all about? An attempt to lower filament voltages?...

Makes no sense at all to me. What are the measured heater voltages at the tubes?

[Randall]

I have read the article, which makes sense to me. I don't have a 2-3 watt resistor in that range to put in series, so one is on order. The measured heater voltages as is are anywhere from 33vac at the output tube to 15vac on some and I believe I saw as low as 7 vac on one. This does not make sense to me. Should all of these tubes draw the same? And if so, why? And if they don't, excepting for the dual triode... why?

I favor the idea of swapping out the selenium rectifier for a terminal strip with a diode and a 3W resistor.

[g1]

The schematic file didn't work without adding the pdf suffix so I'll attach it here.
This is a 'widow-maker' type radio so exercise extreme caution. The heaters are all in series with the line voltage across the series string. This is why you are measuring different voltages for each tube.

[The Dude]

......This is a 'widow-maker' type radio so exercise extreme caution......

Yep. An isolation transformer would be advised.

[Randall]

I already put an iso transformer on it when I did a re-cap months ago. I will take out the selenium rectifier and wire up per the schematic with a 1N2007 and a 3 watt 150 ohm resistor.

But to answer my question, with all but a pair of tubes being different, one would expect varying voltages across the respective filaments as long as they total to something close to the line voltage? Correct?

[The Dude]

Good on you for putting the isolation in there! I'm not entirely sure I understand your question, but all of the filaments are in series across the AC in (+ R21 39 ohm resistor), so you needn't really concern yourself with each. If any filament or the resistor is open , none will work. If I understand your question, yes- the total of all of the voltage drops across each filament + the resistor should total the AC in. Really, it has to if they are lit,...... I guess unless one of those .01UF caps to ground shorts.

Edit: Probably a typo, but you meant 1N4007?

[Randall]

Thank you. One of those 0.01 caps to ground did short, and another was leaky, which smoked a few things in the power supply, so I replaced them all. Some were even wrong value. And yes I meant 1N4007.

[The Dude]

Oh, wow. Hopefully none of your filaments turned into fuses!

[Randall]

Forgive me, the one that shorted was not in the filament circuit, it was at 12AU6 pin 6, except it was a Zenith branded 0.004 600v, looks like an old blocking cap, not the 0.01 ceramic disk the schematic calls for. Also, the 15K dropping resistor was only 470 ohms, and it was charred. Some old timer was in here years ago it seems.

[Tom Phillips]

Randall,
One of the design rules for heaters in series is that the current specifications each of the individual heaters must be the same. This is because the current through a single series string of devices does not change. If you look up the current specification for the heaters of the oddball tubes in your radio you will find that it is 0.15 A for the 19T8 and for the 35C5. This is exactly the same as the heater current specs for the other tubes in the radio. Only the heater voltages are different. If we add up all the heater voltages in the string we get 35+6.3+12.6+12.6+12.6+12.6+18.9=110.6V. R21 is added as a final adjustment. With 0.15A flowing in the heater string circuit the calculated voltage drop across R21 is 5.85V. Adding that to the 110.6V we get a total of ~116.45V across the whole heater circuit which is pretty close to the specified 117V line voltage listed on the schematic.

When I asked about the measured heater voltages at the tubes I was curious weather each of the individual tubes were getting their proper heater voltage. It must be off with that added resistor to ground.

… Some old timer was in here years ago it seems.

The resistor to ground in the heater supply is a total screw up so the best approach is to remove it and restore the heater circuit to that shown in the schematic.
If you need to lower the heater voltages when the radio is operated from today’s higher line voltages then you would increase the value of R21 as required to bring them in line.
Tom

[Enzo]

TVs used to have series string heaters, so they were full of odd voltage tubes like 19v 23v, whatever. There were also 117v tube heaters.

Even today, Peavey's Valve King amps have 0.9A 6L6 heaters in series dropping 6v each, plus additionally in series three 12AX7s (wired as a group in parallel 6v. 0.3A each times three) for an additional 0.9A step.. One lower current winding at 32v instead of several amps at 6.

[Randall]

Makes sense. I removed the 33K to ground resistor. Now to make matters more complicated, I measure 131.3vac across the filament line. Today I have 123.7vac at the outlet, and am using a Triad N68X iso transformer.

N68X - TRIAD MAGNETICS - Isolation Transformer, 50 VA, 2 x 115V, 115V, 435 mA | Newark element14.

Would this transformer be raising the secondary 7.6V? And if so, what do I do about it? Is increasing R21 enough, or should I be concerned that the plate voltages and rest of the radio is running too high?

My filament string measures 35.7+20.9+14.4+13.4+15.2+12.7+12.2=124.5vac. R21 measures 45 ohms and drops 6.6v. 124.5+6.6=131.1vac. If my calculations are correct, bumping up R21 to a 100 ohm 2w would drop 15v, bringing the string back down to 116vac. Is this the way to go?

[Jazz P Bass]

Sounds like a plan.