Big voltage swing tube recommendation please

[hemgjord]

Hello all

I need a tube that can swing approx 120-130Vrms with good linearity.
It will be driving a CF with fixed bias and a grid circuit resistance of approx 100k.
Since the drive requirements apart from voltage swing is modest I would like to stay away from the usual power tube suspects.
A 6sn7 with active load or a anode choke will do the trick, but I would like some alternatives.

So please give me some suggestions from our vast collective experience pool

/Olof

[rdf]

Do you mean a CF with 100K input impedance? Why so low? I've had spectacular measured results driving a beam triode CF with a CCS-loaded 6sf5gt for example, but the CF input impedance was much higher than 100k.

[tubelab.com]

I have been experimenting with cathode follower and screen driven output stages using 6336A and some big sweep tubes. You need tons of drive voltage to do this. To get a high drive voltage there are 3 common methods, and one relatively unknown method.

#1, step up transformer. Good ones are out of my price range.

#2, choke loaded driver stage with a moderate B+ voltage. This allows the output voltage to swing above the B+ voltage. Again good chokes are not cheap, and cheap ones do not lead to a flat frequency or phase response.

#3 CCS loaded triode with a bunch of B+ voltage. This is the approach that I have been using lately. I use a voltage doubler from the main B+ transformer to run the driver since current requirements are low.

The first myth is that you need a tube with a high plate voltage rating. Most tubes that were ever intended to drive an inductive load can swing a plate voltage up to twice its continuous plate voltage rating.

At first I was using the dual dissimilar triodes intended for TV vertical (frame output) sweep use (6EM7, 6DN7). These work well but are inconsistent and sometimes you need to try a few to find one with low distortion.

Then I started using the 9 pin dual triodes. 6CG7, 12BH7, 5965, and a few others. These work better than the 6EM7. Look for tubes with a 300 or 330 volt max plate voltage rating.

My last amp uses (you guessed it) the 6SN7. It does work the best. I have been using two of them. The first is an LTP splitter, and the second is a pair of CCS loaded drivers sharing a common unbypassed cathode resistor. I have used it to drive cathode followers and screen driven sweep tubes. 400 volts P-P is easy. I run the CCS's with 550 volts on them and the plate voltage is about 300 volts. This is the best driver yet.

Of course my old 845SE needs about 350 to 400 volts P-P. It uses a CCS loaded 45 and works extremely well. The CCS is fed with 460 volts and the plate voltage on the 45 is set to 275 volts.

The 4th possibility is to use the "augmented cathode follower" circuit. Feedback allows a "cathode follower with gain". I set mine up for a gain of 2 so the drive requirements are low. I get about 12 WPC using a 6336A with a 600 ohm load. This is limited by the power supply.

augmented cathode follower?

[astouffer]

http://www.pmillett.com/pentodes.htm

[Boris_The_Blade]

Pentodes seconded

How much B+ do you have to work with?

[Fuling]

I´ve used choke loaded 6SN7 and 12B4A in my cathode follower power amps with good results. 12B4A can deliver 500V p-p from a 300V rail when choke loaded.

[hemgjord]

Thanks a lot for the suggestions.

rdf - CF input impedance is set by by the grid resistor that is connected to a fixed bias circuit and is used to bias the DC coupled output tubes.

Pentodes, yes I have thought about that as well and will probably try it some time. I am a little sceptic about those for all the usual non rational reasons



Tubelab: as I am a little cheap I share your view regarding the iron in point 1 and 2. I'll probably stay with the 6sn7 or try a 12b4, make a ccs and bump up the B+ somewhat. Do you have a good proven CCS design that is good for these kind of voltages? (b+ 550-600V)

/Olof

[ErikdeBest]

Hi Olof

I bet Tubelab will tell you to use the 10M90S, available from Mouser!

http://se.mouser.com/Search/Refine.aspx?Keyword=10m90s

free shipping if your order is small volume and above 75 euros! With all beauties available at Mouser you have to try hard to keep below that

http://se.mouser.com/

[SY]

Olof, if the 100k is in the grid circuit of the CF, it can be made much larger, like 1M. That will make the swing an easier thing to do since it reduces the power requirements by an order of magnitude. 6SN7 is a good choice, but I think a 6S4 is even better.

[hemgjord]

Sy, your comment got me thinking, I just took the max recommended grid circuit resistance from the data sheet. But since a CF has 100% cathode feedback it should be extremly self stabilizing, hence allowing a bigger grid resistor? Am I right so far?
How do you estimate the max grid resistance for a given tube in a CF configuration.
In this case I am using the CF direct coupled to the grid of a trioded 813 so I really want the CF operating point to be stable and reliable.

/Olof