Again

[Hojvaelde]

Hello

After being called an amateur I had to pull myself together and developed this solution.
The output tubes are 1625 or 6BG6G.

Is this a workable solution?

Regards

Benny

[Hojvaelde]

Hey

I forgot to thank the Forum for answers to my questions about metal envolopes.

It is about static charge on the house appeared reasonable. Zap zap

Regards

Benny

[DougL]

Just a quick glance, but there is no reference voltage / path to ground for the "LTP".
Without that, it is going to build up a charge until it turns off.

HTH

Doug

[SY]

The 1u cap should be eliminated. You'll have to rejigger the DC voltages to make everything play together nicely.

[kevinkr]

I'd also run the 6SC7 at a 3 - 4mA total, a half mA per side seems totally inadequate for driving output tubes and given the plate load resistors will not develop anywhere near the swing required. 1mA per side would be the minimum I would consider.

With an rp of 53K at around 2mA of plate current the 6SC7 should have a plate load resistor >100K, and a much higher supply voltage than the 300V you've shown.. Should no more than 300V be available you might need to consider using a self-biasing phase splitter configuration, and possibly a CCS if you want to stick to a reasonably well balanced LTP.

Take a look at the Mullard 5-20 design for some hints on configuring the phase splitter.

This might not be the best choice of tube for this application frankly.

[Hojvaelde]

Originally Posted by SY

The 1u cap should be eliminated. You'll have to rejigger the DC voltages to make everything play together nicely.

Yes of course. Silly me.
I'm missing the DC working point. Either I remove the condenser and allow the input stage to direct drive the splitting tube, or place a reference voltage at the input of the splitting tube.


Benny

[Hojvaelde]

Originally Posted by kevinkr

I'd also run the 6SC7 at a 3 - 4mA total, a half mA per side seems totally inadequate for driving output tubes and given the plate load resistors will not develop anywhere near the swing required. 1mA per side would be the minimum I would consider.

With an rp of 53K at around 2mA of plate current the 6SC7 should have a plate load resistor >100K, and a much higher supply voltage than the 300V you've shown.. Should no more than 300V be available you might need to consider using a self-biasing phase splitter configuration, and possibly a CCS if you want to stick to a reasonably well balanced LTP.

Take a look at the Mullard 5-20 design for some hints on configuring the phase splitter.

This might not be the best choice of tube for this application frankly.

1: I can see that I have misunderstood the datasheet. It is 2 mA for each plate and not total.

I want to replace the glass tube with a 6J5 tube. Then the tubes are more alike.

2: If everything goes wrong I can just use the 6N7 tube. 6SC7 tube gives greater challenges.


As Arnold said: I will bee back.


Thanks

Benny

[Miles Prower]

There are things about this design that look decidedly odd.

I would definitely suggest getting that 6C4 out of there. It's an RF type, primarily intended for use as an oscillator/buffer/driver/frequency multiplier. Its only audio uses are as cathodyne splitters or cathode followers. Its operating conditions look odd. It usually doesn't need 2K2 in the cathode to bias it.

The 6J5 is far more linear, and gives the same level of gain.

The 6SC7 is made for making LTPs, especially. However, it will require much larger plate loads, being that this type is in the same class as a 6SL7: high plate resistance, high-u type, that operates with very small plate currents. Better to back this up with some sort of grid driver (cathode/source follower). Otherwise, you'll need something that can pull more current (12AT7, 6GK5 (requires two of these) 6BQ7) to adequately drive the grids of the finals, which are definitely not the Hi-Z, Lo-C load these VTs like to see.

[Hojvaelde]

Yes Miles

I can see now. The 6SC7 tube is a week tube with small plates. I am using tubeCAD.

Next: 6N7G. This tube has power for driving. And 6J5 as input.

Benny