6EM7 and deciding what to with second triode

[olivergardiner]

I've got a build in progress for a preamp that is deliberately retro, so based on the likes of an EF37A (in the clean channel) and 6SL7GTs. The issue I discovered is that there seem to be very few triodes or double triodes in Octal/Loctal format that can tolerate any viable heater-cathode voltage needed for the cathode follower in stage 2 - I only found one, the 6EM7, which has very dissimilar triodes. The first unit I can bias into the "current stealing" mode merlinb describes and I was planning to use the second unit (the high current device) as a driver for the effects loop. However, when looking at the design more closely, because it's a big beast it's going to need a load resistor on the anode to drop the plate voltage down below 150v (plus the cathode voltage of around 20v) and it'll draw a sizeable current unless the load line is practically horizontal (way below the kind of currents it's clearly designed to supply) and this also means that the power dissipation in the load resistor could be high. My rough calculations suggest that I'd be at around 0.5W dissipation with a cathode resistor of 10k - 15k and a load resistor of 47k - 82k, but the load line would sit below the curves shown on the plate characteristics and presumably this would significantly affect linearity.

The first stage is currently intended to be a parallel 6SL7GT which gives me a few options:
1. Bias the second unit of the 6EM7, as above, as the loop driver
2. Make the first stage a single triode, use the resulting spare triode as the loop driver, and leave the second unit of the 6EM7 unused
3. Just use 6SL7GTs throughout, elevate the heater to around 60v and live with around 110v quiescent Vhk on the cathode follower

NB: I've punched the chassis and would like to avoid using an adapter socket so I'm discounting the noval double triodes (which would obviously be an easy solution).

I would welcome any thoughts!

[olivergardiner]

Having re-checked the 6EM7 datasheet, I see that the max Vak for both units is 330v (the 150v plate voltage is the typical operating point) which does provide a bit more flexibility but I think still leaves a few issues:

The loop driver is fed from the tone stacks and so is already ground referenced and so I was planning to adopt a simple cathode biased gain stage with the output take off the cathode (as per merlinb's preferred arrangement with a 12AX7). That gives headaches (as far as I can see) with high dissipation in the load resistors. An alternative would be to AC couple the follower with the plate connected to the HT. This would introduce another unnecessary pole at a few hertz but that shouldn't really matter. There'd be no inherent limiting but could then bias it to around 7mA (or maybe a bit less) with a cathode load of say 10k sitting the output at around 70v. Quiescent dissipation in the tube would then be around 1.5W and 0.5W in the cathode load resistor

[PRR]

Welcome!

> I've got a build in progress for a preamp that is ... ... ... ... ... ... ... ... ...  cathode load resistor

Not clear. Draw picture.

[anotherjim]

Isn't ECC88/6DJ8 a possibility? I think they are intended to allow for use in cathode follower. However, I suspect they may be microphonic as not intended for audio.
http://diyaudioprojects.com/Schematics/6DJ8-SRPP-Tube-Preamplifier.htm
I've got a bunch of them, but not tried anything yet. Annoyingly, their pinout not compatible with 12AX7 etc...
To explain the note on that schematic a little clearer, both heaters share pins 4&5 and can only use 6.3v. Pin 9 is a screen between the triodes and isolated from everything else.

[olivergardiner]

Welcome!

> I've got a build in progress for a preamp that is ... ... ... ... ... ... ... ... ...  cathode load resistor

Not clear. Draw picture.

Apologies - will post a full circuit in due course but the first part of the problem is explained here:

http://www.valvewizard.co.uk/dccf.html

As merlinb says, this circuit "exclusively" uses 12AX7s in order to put the bias point of the DC coupled cathode follower above the 0v grid curve as this generates the characteristic 2nd harmonic distortion. As I have set myself the task of using older, octal valves, the 12AX7 et al aren't in the running. The 12A#7 series have a maximum Vhk of 200v so they don't generally complain about operating with the cathode at around 200v, particularly with a bit of heater elevation. However, the 6SL7 and 6SN7s have a max Vhk of 70v. The only octal triode/double triodeI have found with a Vhk max of 200v is the 6EM7 (intended as a TV valve with one half being a high gain low current triode for the frame oscillator and the second half a low gain high current triode for driving the deflection coils.

The first half of the 6EM7 is not dissimilar in characteristic to the 6SL7 and the combination of 6SL7 and 6EM7 can be biased into the same region as merlinb's DC coupled cathode follower circuit relatively easily. I can upload the grid curves to show this if that helps.

So far so good. The problem is what to do with the second half of the 6EM7 as it's a 10W device designed to put several 10s of mA into field coils. If I can make it into a sensible effects loop driver then that would at least put the triode to good use but biasing so that it's in a viable linear operating region without requiring the biasing components to be high power resistors is more of a challenge.

Again, I will upload the loop driver circuit in due course (merlinb discusses the topic of effects loops in his Guitar preamp book but not on the website) and I have have used his circuit very effectively with a 12AT7 for loop drive and recovery. However, I can't see that working with the power triode in the 6EM7. My current plan which I think will work is to revert to an AC coupled cathode follower for the loop drive stage.

Merlinb's circuit inherently limits the loop output signal to around +/- 2v while still offering low output impedance. Using a more conventional AC Cathode follower won't limit the swing at all but will still provide ample(!) drive.

Circuits to follow...

[olivergardiner]

Isn't ECC88/6DJ8 a possibility? I think they are intended to allow for use in cathode follower. However, I suspect they may be microphonic as not intended for audio.
http://diyaudioprojects.com/Schematics/6DJ8-SRPP-Tube-Preamplifier.htm
I've got a bunch of them, but not tried anything yet. Annoyingly, their pinout not compatible with 12AX7 etc...
To explain the note on that schematic a little clearer, both heaters share pins 4&5 and can only use 6.3v. Pin 9 is a screen between the triodes and isolated from everything else.

Many thanks for the suggestion - the reply is much appreciated!

Unfortunately, the ECC88 is small 9 pin base and part of the challenge here is to try and build the preamp using older octal valves. I can see the ECC88s working fine (not that I've ever tried it) in DC cathode follower or cascade circuits as their max Va is quite low (around 130v) but you get a bit more rail to play with if they're stacked.

Thanks again

[amptramp]

The 6DJ8 / ECC88 is a frame grid tube used as a low-noise cascode RF amplifier for TV and FM receivers where the cathode of the upper triode sits at the plate voltage of the lower triode.  The pin 3 heater-to-cathode voltage is 50 volts maximum but the pin 8 cathode can go to 130 average and 150 peak so the upper triode in a cascode is going to be pin 8 cathode.

[PRR]

I suspect "all" the 6SN7/6SL7s you can get today are fairly modern (TV-era) production and have the same H-K insulation and rating as the minis you are looking at.

There is usually some way to either run the cathode a little lower (you do not need 50V of signal!) or add an elevated heater supply.

The 6EM7 is great for what it is but NOT a 1st or 2nd choice for "preamp". (I also doubt it has fundamentally different H-K insulation, but maybe better inspection for flaws.)

[olivergardiner]

I suspect "all" the 6SN7/6SL7s you can get today are fairly modern (TV-era) production and have the same H-K insulation and rating as the minis you are looking at.

There is usually some way to either run the cathode a little lower (you do not need 50V of signal!) or add an elevated heater supply.

The 6EM7 is great for what it is but NOT a 1st or 2nd choice for "preamp". (I also doubt it has fundamentally different H-K insulation, but maybe better inspection for flaws.)

This is the current state of play - by using 200k load on the first stage of the DC Cathode Follower I can push Va down to 150v and with the heater up at 60v it's not too far away from the 70v datasheet limit for Vhk. I will have a play with whether or not to use the bootstrap capacitor. I'm using NOS Russian 6H9C tubes for the 6SL7GTs.