Cathode follower inquiry

[thejaf]

I've searched a few places, and can't seem to find the direct answer that's not ubertechnical, so here's my inquiry:

In a cathode follower stage with plate connected directly to B+, how can I figure out the expected voltage at the cathode? I assume lowering the cathode resistor will also lower the cathode voltage as well, but is there a limiting voltage drop "across" the triode (12AX7)?

Circuit in question would be in the self-biasing cathode follower FX send of series loop, where plate is connected to power tube screen voltage (~440VDC), 470ohm K resistor, self-biased with 1M to grid, then 22K to ground from that connection. What voltage would I expect to see at this output junction of 470-1M-22K? Circuit calls for a 1uF blocking cap, and I'd like to know the proper voltage rating.

Thanks!

[roberto]

You'll have around 200-230V, but usually you have great swings on CF, so I suggest you to use a cap that can handle the supply voltage, so 400V.

[thejaf]

Great, thanks for letting me know! I assume it would also be a good idea to elevate the heaters @70V or so as well?

[roberto]

Yes it's a good idea. Something like a 1M 220k voltage divider from B+, with a 10-33µF in parallel with the 220k. Adjust the 220k value to your B+ to obtain around 70V.

I supposed 100k on the cathode of the CF, as usually used.

[pdf64]

'I assume lowering the cathode resistor will also lower the cathode voltage as well'
No, rather it would tend to increase cathode current.

The voltage at the cathode tends to be a volt or 2 more than the grid voltage. My reading of your arrangement is that there's a potential divider from the 440v screen node, the top resistor being 470k, bottom resistor to ground of 22k.
The output voltage of that potential divider would be 19.7v.
This then sets the grid reference voltage via a 1M grid leak resistor.
The cathode voltage will be about 21v, so for a reasonable mid biased cathode current of ~1mA, a 22k cathode resistor should be a good starting point.
That arrangement is putting a lot of voltage across the 12AX7, plenty over the limiting value of 330v, though the plate dissipation would be a managable 0.45W.
It may be beneficial to derive the plate / bias voltage from a lower voltage node.
Pete[/quote]

[thejaf]

Thanks for the reply, the top resistor is actually 470-ohm (K was for cathode), sorry for the confusion.

Attached is the FX loop circuit in question, although my B+ isn't quite so high as the example. I don't use anything in an FX loop, but am interested in how this would change the character of my current amp project (based off the Bogner Shiva). I have heard that the series FX loop gives a good amount of this amp's character, and also notice the return is used as overall master volume. As a former Shiva owner, I know that the master volume worked extremely well.

[gingertube]

The 12AX7 follower in the circuit posted will be working quite hard due to that "lower than usual" 470 Ohm bias resistor. Probably up around 2 mA. That means that you will see about 44 Volts on the top of that 22K. The normal bias across the 470 Ohms would be around 1 Volt so the cathode should be sitting about 1 volt higher than the top of the 22K. That leaves 400'ish volts across the triode which is GOOD - that high voltage will help hold the tube out of operinting in its grid current region which you would usually see with a 2mA anode current (well it is actually about 50V higher than ideal but still OK). In your own "copy" you might want to increase that 22K to say 33K to get the voltage across the triode (anode to cathode) down to 350 Volts.

The triode section will be dissipating 400V x 2mA = 0.8 Watts, that is quite high compared to most 12AX7 triode circuits you will see but is still within specs. For long life of the entire amp might be good to swap this tube with one of the others every 12 months or so.

Cheers,
Ian

[martin manning]

Attached is the FX loop circuit in question, although my B+ isn't quite so high as the example...

Do you know this schematic to be correct? IMO running the loop directly from the screen node is not the best idea- It might be noisy and it definitely puts the operating point for the cathode follower way out of place. I would add a branch from the screen supply and create two new nodes, the first at say 350V to power the recovery stage, and the second at about 250V to power the CF.

[tubeswell]

What Martin said. Lower the B+ to about 280V with a dropping supply, then use a 100k load for the CF and 1k bias resistor. JM2CW

[thejaf]

Thanks all! I am pondering the ommision of the cathode follower stage, and taking the tone stack-channel master output directly to the above posted FX recovery stage (used as global master).

Maybe also omit this stage too, and just use traditional master for now..... I'm curious the tone/feel differences.

[tubeswell]

If you want the CF to add 'flavour' to the amp's grunginess when you jumper from the FX send jack to the return jack, try setting up the CF with about 56k load line. I attached a soundbyte of a CF mash-up that I did to a C30 like this about 4 or 5 years ago (Geetar is a strat). Gives quite a satisfying late-60's grind methinks