• WARNING: Tube/Valve amplifiers use potentially LETHAL HIGH VOLTAGES.
    Building, troubleshooting and testing of these amplifiers should only be
    performed by someone who is thoroughly familiar with
    the safety precautions around high voltages.

Directly heated Cathode/insulation question

I propose* to insert a 5R4GBW diode in series after a bridge rectifier.
The line in which the diode is to be inserted is at a potential of +1200V approx (rectified from an 800Vac secondary winding). As you may know the cathode is directly heated by 5 volts ac/dc.

I will be using another 5V winding to supply the filament. This winding will then float at 1200Vdc. My concern is this; will the internal insulation of the transformer be able to cope? Does anyone have experience on the subject??

*FYI I want to use the warm-up time of the 5R4.. to act as a delay mechanism before HT is applied to the amplifying triode plates.



2007-07-12 1:42 am

The winding voltage insulation is a transformer spec... look at what you're buying and see what the rating is. Note that the 5R4Gxx being directly heated is not going to have much delay at all in voltage output. If you want to delay the voltage using a rectifier tube, look for an indirectly-heated cathode type.

Regards, KM
Thanks both Nick and KM.
I'll look for an indirectly heated cathode.
The diode will need to operate at high voltage, although the PIV will not be an issue, as it will be installed after the choke, in fact between the choke and one end of the OPT. The other end of the OPT is connected to the anode of the output triode (845).

Any idea on what I might use instead of the 5R4... given that I'm looking for a long warm-up delay?

If there is 1kv on the cathode and you ground the filament then you would probably have a failure.

Your filament supply would be about the same for both indirect and directly heated tubes.

You would still need a good filament transformer that could stand off a couple kv in my mind.



2007-07-12 1:42 am
Most 5V indirectly heated rectifiers have the cathode tied to one of the filament lines in, like the 5AR4, 5V4G, etc. so you can't have that situation. Also, most filament transformers are rated at 2000 VRMS insulation... even the Hammonds meet this spec. What current range are you requiring for the B+?

Regards, KM
I was just looking at the 5AR4 data-sheet online, and that is shown as strapped internally.

So it would appear that this is common practice.

But it does seem to violate safety, by bringing high dc voltages back into the transformer? Which was the basis of my concern when I started to wire things up to include the diode .

I expect the 845s to draw about 120 ma each so total 240ma, max fused at 200ma per channel just in case I lose bias
Attached retry


  • crop _diode tube in power supply circuit.jpg
    crop _diode tube in power supply circuit.jpg
    9.7 KB · Views: 131
kmaier said:

If you want to delay the voltage using a rectifier tube, look for an indirectly-heated cathode type.

Regards, KM

Take a look at "TV damper diode" tubes. They have indirectly heated cathodes, and more specifically, the H-K insulation is extra thick, to take voltages up into the kilovolt range. Tubes like the 6CG3.

link to Frank's tube data site 6CG3 pdf

It says +900VDC cathode to heater, +5KV on peaks. Using a separate heater transformer, bias the heater winding around +400VDC with 3.3 meg or so resistors. Or a voltage divider hooked to the 1.2KV B+, something like 1 meg to ground, 2.2 meg to B+, the mid point to the heater. This would also help drain off residual charge off the filter caps after you switch the power off. A safety feature. Use 1 watt resistors.


2007-07-12 1:42 am
Good idea on the damper diodes.... however, once you start looking at the specs, a few things start becoming clear...

1- average current is generally 200ma or less
2- heater-cathode voltage ratings are +100 to +300 and -900 to -4500 (don't ask me how the tube knows)
3- the PIV rating is generally for 15% pulse rating of a horizontal scan cycle, around 10 microseconds
4- plate dissipation is usually less than 10 watts

I've seen these do everything from a simple glowing plate to internal fireworks when things start breaking down. Not sure I want to use these in continuous voltage applications.

Again, typical Plate/Filament transformers (and single filament transformers) are rated at 2000VRMS (multiply by 2.8 for DC) so having the filament winding hanging at 1200- 1500VDC is fine.

I would be inclined to use either a 5AR4 or back to the original 5R4GY with a separate filament transformer. In the case of a 5R4GY, use a simple 555 timer driving a relay for the required delay for turn-on.

Regards, KM
Good stuff here, many thanks to all.

FYI The 555 timer was where I started.. The timer circuit looked a bit complex and flimsy to me, with relay contacts in the 1200V and 400v supply lines. My thought was to replace both with the diodes using the heater warm-up as the delay. Now it seems the warm-up delay won't be long enough.

So do you think I could put a 555 driven relay in the Diode filaments
path?? That way I keep well away from the high voltages.

The idea of a separate filament Trans. makes a lot of sense and I'll certainly do that together with the resistor bias ladder.

I already have the 5R4WGB's so I'd like to stay with them.