Cathode follower

[Diabolical Artificer]

Out of curiousity can I bias a 6N7 using a positive voltage? I want to use as a cathode follower. See graph. The 6N7 is an odd valve having really a low 0v grid line for a triode.

Andy.

[Herald1360]

Provided the drive to it can cope with the low impedance, there's no reason why not.

To bias it to the blob point on your curves with 200V Vak you'll need to put about +65V on the grid from an HT of about 250V- allow for the 5mA of grid current- and with a 1K cathode resistor (2.5W dissipated there) the cathode should sit at about +50V.

It's a strange valve designed for zero bias B2 push pull operation- why use it as a cathode follower?

[G8HQP Dave]

If the drive can cope with low (and varying) impedance then why use a CF at all?

[Diabolical Artificer]

Think I might be missing something here, why will it have a low and varying Z? I admit, I hav'nt calculated the input Z, but assumed that like most cathode followers it would be high.A cathode follower AFAIK is effectively a buffer/current amp, with high impedance in - low out. I'm toying with the idea of using it as part of the driver to drive the 807 amp I'm doing, as it has high current capability.

The problem as you say because it's an odd valve is the shared cathode, which means I cant bias the grid by using a cathode resistor, but will either have to bias by using a potential divider or separate bias supply.

The 6N7 probably isn't suitable to be honest, but as I had one I though I'd experiment.

Andy.

[G8HQP Dave]

A valve only has high input impedance when no grid current is present. Positive grid bias will guarantee significant grid current, and this will not vary linearly with signal. Fortunately the cathode follower will bootstrap this low input impedance and so raise it, but it could still be a cause of distortion if the signal source has high output impedance. If the source has low output impedance then you don't need a CF.

[Herald1360]

Originally Posted by Dr Wobble

The 6N7 probably isn't suitable to be honest, but as I had one I though I'd experiment.

Andy.

Why not use it as intended and knock up a 10W p-p zero bias class B2 amplifier? That sounds like an interesting challenge (and you'll need cathode followers to drive the 6N7 if you don't use a transformer for the phase splitter).

[Hartley118]

The 6N7 is an interesting triode output valve, largely undiscovered by those seeking 'audio magic', perhaps because of its intended use as a class B P-P output. You can, though bias it perfectly satisfactorily into class A. I use one as the speaker output stage in my R1155; both halves are connected in parallel. I chose a triode stage because it doesn't then matter if I unplug the external speaker with its integral OPT, whereas a tetrode or pentode would overstress its screen grid. And, like you Andy, I happened to have one!

It should work as a cathode follower, but I'd bias it to avoid those positive grid voltages because grid current does rather negate the high input impedance purpose of the circuit!

Martin

[G6Tanuki]

The 6N7 is indeed interesting - I once used one as the modulator in a little transmitter with a 2E26 as the PA.

Using a carbon fist-mic passing plenty of current, and a step-up centre-tapped transformer feeding the 6N7 grids was all I needed to get plenty of 'communications quality' audio.

Not sure about the benefits of using one as a cathode-follower though; the 6N7 expects to be driven into grid-current so it'll present a low (and significantly varying) impedance to the previous stage.

[tri-comp]

Hi,

Something is quite wrong here.
I'm all for the 6N7 as a CF-driver for the 807's, just parallel the halves of the 6N7. You really don't have any choice with the shared Cathode. It will be an enormous overkill, but since you have the valve on hand, fine.
You will absolutely NOT think about driving the 6N7 into Grid-1 current; the input-Z will drop to just a few 100 Ohms and how will you drive that with the preceding P/I ? It expects to see a load not much below 500KOhm !!
When you set the 6N7 up for a CF-drive for the 807's the cathode will be a the negative G1 potential required for the 807's. The 6N7 G1 will be even lower; just set to the wanted G1 voltage appears on the 807.
I suggest you take another look at the CF drive I did for my 2C34 project.
It will be very similar when used for the 807's.
Go for a CF-drive w/o CCS for starters; just connect a 20-40KOhm/Lots of Watts (10W possibly) resistor between each CCS points 3 & 2 and forget about CCS point 1, which is not needed.
Obviously quiescent negative 807 G1 voltage should be set much lower than -15V, but that's just a case of adapting the voltage divider resistors around the bias setting pot's. I would suggest raising the 100KOhm resistor from 2 x 2K7 resistors to ground. Maybe 330KOhm will be suitable.

[Diabolical Artificer]

Thanks all. I kept 6N7 out of the title in a probably vane attempt to keep it's existence quiet. I thought about using it because I've run out of 6SN7's (6H8C) having dropped one, but thought I was missing something, still struggling with impedance a bit and some of the aspects of cathode followers.

The data sheets show figures for 6N7 Class A operation, and as you say Martin, it should/does make a half decent low output, output stage.

As I'm kicking my heels waiting for some bits for the amp to come, I'll have a play and practice Z calculations and try out out your driver T.

Andy.

[ms660]

Some stuff on cathode followers here if it's any use, Scroggie, Tektroniks, Blencowe:

http://www.keith-snook.info/wireless...20Follower.pdf

http://w140.com/tube_circuits_in_tek.pdf

https://books.google.co.uk/books?id=...encowe&f=false

Lawrence.

[FERNSEH]

This page from the Radiomuseum shows the original application for the types 53, 6A6 and 6N7 as a 10watt push-pull amplifier.
http://www.radiomuseum.org/tubes/tube_53.html
The 6N7 was used in the timebases of American TV receivers.
Never seen it used as a cathode follower.

DFWB.

[tri-comp]

Originally Posted by FERNSEH

This page from the Radiomuseum shows the original application for the types 53, 6A6 and 6N7 as a 10watt push-pull amplifier.
http://www.radiomuseum.org/tubes/tube_53.html

Not a very good idea driving the 6N7 into class B2 from a transformer.
Bad for feed-back and bad for notch-distortion.
See the attached document for explanation on why and how to CF-drive correctly.
CF-drive works and works very well for Hi-Fi even when supplying gridcurrent to the output valves.
It will also work admirably when running the output stage in class AB1.
Imho the 6N7 will do nicely with a mu high enough to be able to calculate a suitable operating point iaw the attached.
Check it out Andy even if paralleling sections is a huge overkill.
...or, perhaps not so much overkill after all considering you'll end up with 3 hungry 807 controls-grids to supply from each 6N7.

[G8HQP Dave]

Transformer drive was the standard method for Class B2. This was generally used for high power PA and AM modulators, not hi-fi.

[Diabolical Artificer]

Thanks for all those links and info. I had a copy of the Tek circuits, very useful.

One question; When designing normal common cathode stages we have to make sure that the preceding valves AC output "fits" within the following valves G1 scale, EG, if valve 1 has an output of 20vAC P-P, then looking at V2's load line we see that the G1 curves only go from 0v to 6v, we would either have to use different valve as V2 or reduce the voltage output/swing of V1. This isn't, AFAIK, the case with a cathode follower, as the OP of a preceding phase inverter with gain is likely to be 100v or more. Why?

Andy.

[Hartley118]

Originally Posted by G8HQP Dave

Transformer drive was the standard method for Class B2. This was generally used for high power PA and AM modulators, not hi-fi.

Nothing wrong with transformers as long as they're well designed. The audio from many of the most highly regarded recording studios will have travelled through a dozen or more transformers by the time it reaches us.

[Hartley118]

Originally Posted by Dr Wobble

One question; When designing normal common cathode stages we have to make sure that the preceding valves AC output "fits" within the following valves G1 scale, EG, if valve 1 has an output of 20vAC P-P, then looking at V2's load line we see that the G1 curves only go from 0v to 6v, we would either have to use different valve as V2 or reduce the voltage output/swing of V1. This isn't, AFAIK, the case with a cathode follower, as the OP of a preceding phase inverter with gain is likely to be 100v or more. Why?

I'm not sure I'm understanding what your question is here. A cathode follower is only a means of increasing the current drive with a voltage gain approaching unity.

It's always good practice to ensure a bit of headroom in earlier stages so that overload occurs first in the output stage. So, e.g. with 807s (gm of 6) if you want a peak anode current of, say, 180mA, that will need a theoretical ideal minimum peak to peak drive of +30 to -30V. Best to have something in hand for practical circumstances with lower gm etc.

Martin

[G8HQP Dave]

Originally Posted by Hartley118

Nothing wrong with transformers as long as they're well designed.

My point was not that transformers are not hi-fi, but that Class B2 is not (usually) hi-fi. Also, that when B2 was used it was traditionally via transformers not cathode followers.

[Hartley118]

I guess it's easy to get class B2 wrong and not have enough open loop gain at low signal levels for negative feedback to be effective. We had the same problem with some early transistor amps, but we don't have much of a hangup on the topic these days.

Martin