Biasing JJ 6V6’s within the Dumble like power supply and output stage?

[Mark]

I have a 5E3 amp to service and it’s one of those amps that is a culmination of mods. (All the parts are of good quality, such as Mercury Magnetics transformers and boutique caps.)

It uses a 12AX7 in lieu of the 12AY7, one of the preamp coupling caps is an 0.047uF, the P.I. stage uses the Paul C mod.

https://ampgarage.com/forum/viewtopic.php?f=8&t=38295

The power supply is different too, the amp uses a new production Tunsol 5AR4, 30uF first filter cap, a choke in lieu of a 4.7K or 5K choke resistor. There are no screen grid resistors.

I decided to start with checking the bias of the output JJ 6V6’s I thought the extra voltage from the 5AR4 coupled with the 250 ohm 8 watt Ohmite cathode bias resistor (measures 251 ohms) might mean the output valves are biased too hot.

I measured the mains voltage and it was measuring 239vac (opposed to 240vac Australian). The B+ voltage are:
B+1 = 419VDC , B+2 = 419VDC, and B+3 = 309VDC.

The voltage across the cathode resistor is 26.23VDC

The current flow through the two 6V6’s is 53.8mA and 48.7mA respectively. This is just nuts, the maximum current should be 35.5mA’s.
The percentage dissipation of the first 6V6 is 151% of the maximum valve and the second one is 137%, I don’t think this is a good environment in an output stage.

I tried putting a NOS 5Y3 rectifier in and the power supply nodes measured: B+1 = 373VDC, B+2 = 370VDC, B+3 = 274VDC.

The voltage across the cathode bias resistor is 23VDC. The current flow through the valves is 47mA’s and 43mA’s. Given the maximum current flow is 40mA’s, this means the plate dissipation is 117% and 107% respectively.

This seems a far better situation than the Dumble Tweedle Dee setup which I really question the value of at this point.

I did try simply increasing the cathode resistor value till the output valves were within acceptable current flow, however I needed a 802 ohm resistor and the rail voltage was now 462VDC which would exceed the voltage rating of 6V6 valves.

At this point I’m not seeing the value of the Dumble Tweedle Dee like power supply/output stage. I’m inclined to use a 5Y3, reduce the first filter cap to 20uF, increase the size of the cathode bias resistor, and use 1K screen grid resistors.

My question is have I missed something or is my logic sound?

Thanks for your assistance.

[sluckey]

Lower the screen voltage to bring the current down.

[Mark]

Lower the screen voltage to bring the current down.

Thanks again for your help, I will give that a try, and see what happens.

[Stevem]

As posted back to you in reply,the screen voltage is the ultimate control of the current passing thru the tube.

Also in a cathode resistor biased output stage the voltage seen across the resistor as far as the tube is concerned is subtracted from the plate voltage on the tube.

The JJ 6V6 is more of a 6L6 in many ways.
Voltage wise I have used them in Hiwatt 50 models with well over 470 volts sitting on them.

If when you have them running at max wattage with a 250 ohm resistor on them and in a dark room you do not see the plates showing any red spots starting to develop on them then you are likely ok.

[Mark]

Also in a cathode resistor biased output stage the voltage seen across the resistor as far as the tube is concerned is subtracted from the plate voltage on the tube.

The JJ 6V6 is more of a 6L6 in many ways.
Voltage wise I have used them in Hiwatt 50 models with well over 470 volts sitting on them.

If when you have them running at max wattage with a 250 ohm resistor on them and in a dark room you do not see the plates showing any red spots starting to develop on them then you are likely ok.

Thanks for taking time to reply to my question.

The calculations for bias were carried out using plate to cathode voltages.

I had heard that the JJ 6V6 is quite rugged, however the plate of one valve is dissipating 21 watts. That is very close to 5881 territory and the JJ datasheet does state 14 watts as the maximum plate dissipation. I’m not entirely comfortable with using something that is an unknown quantity. How long do the valves last at 151% plate dissipation?

EDIT:
https://ampgarage.com/forum/viewtopic.php?t=13887

This link alludes to the plate dissipation being as high as 16 watts which is still a long way short of 21 watts. I have read a few posts and replies, but as stated earlier nothing was particularly definitive. Most were concerned with the 500VDC maximum voltage rating.

[martin manning]

The JJ 6V6 is more of a 6L6 in many ways.
Voltage wise I have used them in Hiwatt 50 models with well over 470 volts sitting on them.

They are capable of higher plate dissipation, but still rated at ~half of 6L6GC rating. Characteristics are very close to 6V6GT. See: https://ampgarage.com/forum/viewtopic.p ... 79#p293279

[martin manning]

Lower the screen voltage to bring the current down.

You could bring the idle dissipation down by doing this, but you would also bring the knee of the Vg1=0 curve down, potentially ending up with a poorly positioned load line.

[Mark]

The JJ 6V6 is more of a 6L6 in many ways.
Voltage wise I have used them in Hiwatt 50 models with well over 470 volts sitting on them.

They are capable of higher plate dissipation, but still rated at ~half of 6L6GC rating. Characteristics are very close to 6V6GT. See: https://ampgarage.com/forum/viewtopic.p ... 79#p293279

Thanks Martin, I appreciate your input. Your conclusions were as I suspected that the 6V6S doesn’t have the capability of a 5881.

[Mark]

Here is the update on the amp.

Update required due to found error.