Cathode follower voltage question 6SL7

[Bloomfield]

Hi,
I am working on a design right now which is largely based on the 5G9 Tremolux in the tremolo and power amp sections. For various reasons I am using all octal-based tubes. I am planning on using a 6SL7 for the tremolo tube in order to have enough gain for the oscillator. Chassis layout dictates that the other 6SL7 be used as the cathode follower in the tremolo circuit. Looking at the datasheet I see that the heater-cathode maximum is listed as only 90 volts. Elevating the heaters is no problem, but I'd be reluctant to go higher than say 60 volts. My question is would the cathode follower function adequately if I used a very low supply voltage to the cathode, say 150v or so? The 5G9 schematic shows 260v on a 12AX7 at this point.

Any ideas would be much appreciated. The fact the the 6SL7 oscillator has higher plate current than the 12AX7 makes me hopeful that a little loss of performance in the cathode follower might not be such a problem, but I don't really know.

Thanks,

Andy

[Dave H]

My question is would the cathode follower function adequately if I used a very low supply voltage to the cathode, say 150v or so? The 5G9 schematic shows 260v on a 12AX7 at this point.

Any ideas would be much appreciated.

I'm not sure I understand the question. You don't supply a voltage to the cathode. The cathode follows the voltage on its grid. You'd have to bias the driving stage plate to 150V.

[eschertron]

"voltage to the [anode of the] cathode follower [stage]" perhaps? In order to minimize cathode-heater voltage, right?

[Bloomfield]

Yeah, sorry I see I made a mistake there, I was meaning the plate supply voltage on the cathode follower. Looking at the datasheet again I see plate resistance at 44K, so at 2mA for example it would drop 88V. If I was to aim for about 260V as the plate supply, resulting in 150V or so at the cathode, should that work reasonably well?

Thanks,

Andy

[Gingertube]

Schematic Link:
Prowess Amplifiers - Fender - Schematics - Tremolux 5g9 - Schematic

The 1/2 6SL7 triode (12AX7 on the schematic) which is the trem oscillator has say +2V on the cathode. The other 1/2 6SL7 triode which is the trem drive cathode follower has +260v on the cathode.
Ideally you would want to put the heater at +130V to even out the heater to cathode voltages (about -128V and +130V for the 2 triodes) but it is outside the data sheet spec..

The oscillator triode output signal will be about +/- 100V (200V peak to peak) and the cathode follower signal output will be the same.
Hence the current value of +260V for the cathode follower triode (to accommodate the 200V pk-pk signal swing).

This output is then divided by the 1M and 250K intensity pot to a max. of 1/5th of that 200V peak to peak to be applied to the output tube grids.

So you don't need that signal swing.

I would drop the B+ voltage for the entire trem oscillator and cathode follower to say +220V.
You would then get more like +/- 80V (160V peak to peak) and can then set the cathode follower cathode voltage to say +180V.
The ideal heater voltage float is then +90V and you are within (JUST) the +/- 90V heater / cathode voltage spec.

Adjust the 1M divider to the 250K pot down accordingly to say 680K

Plan B (better) would be leave the oscillator alone but to divide the oscillator signal output down before feeding it to the cathode follower, by say X2, you can the use much lower cathode follower cathode voltage and a lower heater elevation. Compensate for that be dropping that 1M divider resistor feeding the 250K Intensity pot down to say 330K. That is: to move much of the trem signal divider from after the cathode follower to before the cathode follower so you can lower the followers cathode voltage.


Hope this makes sense to you.

Cheers,
Ian

[Chuck H]

That^^^^^^^^^^^^^^^^^^^

You can keep in mind that the cathode follower only has a gain of less than 1. The difference in gain between a high plate voltage and a relatively low voltage won't make a lot of difference to the cathode follower itself. So separating the cathode follower (which doesn't need excassive on voltage for gain) from the oscillator (which does rely on voltage for gain) as suggested makes a lot of good sense if you wish to keep the filament circuit elevation to a reasonable level. Though I don't know just what reasonable means in this application.

I did build a trem circuit a few years ago and had similar concerns. Being as there are some Rusky 12ax7's that tend to fail with a high cathode to heater differential. This was an amp for a customer so I wanted to be sure it would work no matter what tube got plugged in there. IIRC I had to elevate the filament circuit 70V. I used a PT from Edcor that had two filament windings so the power tubes were on one and the elevated preamp circuit was on the other. But...

I haven't read much about 6sl7's failing under similar circumstances and I haven't read much about older trem amps that DON'T have elevated filament circuits having any trouble either. So there's that. Maybe I (and you?) are just being fussy and it's not really a problem.?.

[Dave H]

Or you could save a tube by using a MOSFET source follower

[Chuck H]

Or you could save a tube by using a MOSFET source follower

Somebody get that guy outta here.

[Bloomfield]

Hi Ian,
Thanks for that. I saw your post just before going off to bed way too late and it was a bit too much to process at the time, but having read it again today it makes more sense. I'm leaning more towards plan B; as you and Chuck point out it allows for a lower heater elevation. I'd worry that going too high might be an issue in the other tubes, although I'll check the datasheets and see. Doing away with the direct coupling is fine; I'm not setting out to build a 5G9 but just using that tremolo circuit as a starting point as it seems to be very well thought of.

I looked around for other examples of 6SL7's as cathode followers and couldn't really find any useful examples. That 90 volt heater to cathode rating struck me as low, but looking at some other tubes it's nothing unusual, so I won't worry too much. I think I'll probably elevate the heaters 80 volts or so and it should be fine. I've got a nice old mil-spec 6SL7 so it ought to cope with the situation.

I know Dave, a MOSFET would do the job, and one of these days I'll do that, but this one is chance to use some nice old stuff I found in my grandad's basement after he died and some other nice old stuff I've scavenged over the years that goes well with it. Could be a while before I get much done on this one; there's a somewhat obscure chassis punch I need for the sockets, and I have a lot of real work to get through the next few months.

Thanks guys,

Andy

[tubeswell]

I'm with Dave H. An IRF820 source follower is a shoe in for the buffer stage of a LFO circuit. With a 100k load resistor attached to the source and a 100R to 200R limiting resistance in series with the plate of the driver stage and the MOSFET gate, dissipation will be well under 1W and you won't even need to heatsink the IRF820. Also you'd need a 12V zener between gate and source to protect the gate insulation.