MOD RESTING CURRENT ON K1JJ 813 X 813 RIG

[W2PFY]

Does anyone have their values of resting current on the above JJ transmitter. I am thinking there must be some lower limit of plate voltage where the modulator tubes will work and be close to class B? I addition, there must be a limit on how high the plate voltage can be raised to at zero bias before the plates melt?

I am trying to help a friend who constructed a JJ xmitter and I don't have time to run a setup to get the parameters I am looking for.

Thanks in advance for those who have experience with 813's at zero bias in an audio circuit.

Terry

[K1JJ]

Does anyone have their values of resting current on the above JJ transmitter. I am thinking there must be some lower limit of plate voltage where the modulator tubes will work and be close to class B? I addition, there must be a limit on how high the plate voltage can be raised to at zero bias before the plates melt?

I am trying to help a friend who constructed a JJ xmitter and I don't have time to run a setup to get the parameters I am looking for.

Thanks in advance for those who have experience with 813's at zero bias in an audio circuit.

Terry

Hi Terry,


Coincidently, I'm working on and testing my new 813 rig now and playing with the GFZ MOSFET board.  Usually the 813 modulators idle around 50 mA per tube is my goal, depending on voltage.  At 1500V it takes about  +8 volts of positive bias to idle.  (Bias adjusted by the GFZ board)

At zero bias, you can ramp the HV up to about 2200V before they draw too much current and even start to show color which is too hot for carbon plate tubes.

But if you are using a variable HV supply, why let the tubes dictate where you set the HV?  Just add a few diodes in the fil CT lead and each diode will add about 0.6V of bias.

2500V is a great HV level for 813s X 813s and will mean some negative bias. But to answer your question, I found that 2200V max was about right for zero bias.

Just multiply the voltage X the idle current per tube and shoot for a dissipation of 1/2 the tube rating. So, 2200V X 28 mA each = 60 watts each.  Or add some diodes to reduce the idle.

The best way is to estimate the idle like this and then run some tones to see what the IMD and waveforms look like. You may find it makes little difference to beat the tubes when you can back the idle off a little.

BTW, I have the new 813 rig and the GFZ MOSFET board working with on-board negative feedback now. I had problems from ultra-sonics which I discovered today were caused by large wire wound resistors used in the NFB ladder.  I clip leaded in some 1/2 watt non-inductive resistors and my problems went away.

Time to take Yaz out for a long walk and a swim in the Salmon River.

T

[W2PFY]

Time to take Yaz out for a long walk and a swim in the Salmon River.

Thanks Tom, I think you answered every question except one, do you have a Salmon River in CT or are you way up north in NYS near the Saint Lawrence Seaway at our Salmon River?

Another things that shows holes in my knowledge is positive bias on tubes. It was when I started looking at single ended audio amplifiers where for example, a person wanted to use a 100TL triode in class A and wanted to do it with a voltage that might give him a second chance at life if he came across the anode, that I first seen that application. Now my question is, with a high positive voltage on the grid, is the tube now pulsed more positive to give the tube monkey swing? I remember that in class A, most parameters are rather steady.

Thanks Terry

[K1JJ]

Hi Terry -

There are probably several Salmon Rivers in every state...     Yaz had the time of his life running up and down the trails and diving into the river after sticks. He's three years old now and in his prime.

I've been working on the same 813 rig design as you are helping your friend with. I just finished up the negative feedback fine tuning and think it may be the best results I've seen. I have a tone at 1 KHz looking the same as one at 10KHz.  It sweeps cleanly from 15 Hz to 12 KHz now.   The GFZ MOSFET audio driver board is a major part of the rig for these kind of results.

I want this rig to be a showpiece to how loud and clean a pair of 813s modulated by a triode-connected pair can sound. This one is almost there.

Anyway, a positive bias will not change anything except put the tube on a different point on its transfer curve. Hopefully the line is still linear/straight there. The monkey swing will be the same once it is idled properly at that voltage. I have used positive bias on the 3-500Z amplifier at 1500 volts to get a super clean -45 DB 3rd order stage.  It's just a tool to use for more flexibility.

T

[W2PFY]

Thanks Tom for all your help and inspiration.