801a maximised class A2 excursion.

[Paul Barker]

Presently planning all solid state voltage amplification/buffering/power drive. Just as one version, with option to swop mosfet and jfet tasting.

Just idly looking through a bag of valves old and came across four EF37a which might work like 6j5 as input valve but with advantage of trick construction for quiet heater supply.

As quoted from Valve Museum.

“Mullard designate the EF37A as a high gain, high impedance, pentode for use in AF pre-amplifier stages. The heater system is of anti-noise construction, specially designed bifilar wound heater reduces hum to negligible proportions.
High quality amplifiers have used the EF37A for many decades. The ME1400 electrometer pentode is also rated for high quality audio use and is a direct replacement for the EF37A upon which it was based. The EF86 was its successor.
The data sheets also give parameters for use as a triode.
See also 6J7 almost the same valve in operation.”

Also found a pair of 6C5G a 1932 predecessor to the 6J5 except for the diodes.

[pre65]

Watching with interest.

If you get stuck on any parts Paul, post a list and I'll see if I can help.

[Paul Barker]

Thank you

[Paul Barker]

OK , decided to drop the idea of building it first for 10y and then upgrade power supply for 801a. Though I only have two 801a’s theyll maybe see me out.

Not set in stone but a likely operating point 430v 45mA -20v with maximum grid voltage +47 at which point the last vestiges of half of the sound wave would vanish (the other half of sound with won’t clip it has more headroom). But up to that moment power output is 6.6 watts of clean sound with 1% 2nd harmonic.

To drive the grid current we require 0.9 watts peak current 20mA and for max power 45 volts rms.

The plan for amplification was jfet cascode which gives 34db acording to the author but if source is sufficient level the db gain won’t affect the 801a if the headroom required is above jfet max volts. Efforts to find an alternative proving unproductive so far.

I was working backways from the mosfet source follower, studying the which I stumbled on an Audioxpress article in which I found the jfet cascode circuit, stay with me.

The upper jfet is dc to the mosfet source follower with level shifting (to be decided on which form, but the bipolar mosfet negative rail is obviously involved) the bias voltage of the output valve (after allowing for the g:s voltage difference, in the event where the bias voltage is picked off the resistor voltage divider) will be via a multiturn 10k pot so easy to quickly home in if necesary.

But; now I cant get the voltage swing to the 801a for the full 6.5 watts. So shall I build the solid state stage? Yes, I’ll do it as is, and as higher voltage jfet becomes available that limitation is breached. So because its new and Im learning and do it like this first. It could easily end up full power for the sake of the required jfet.

But It’ll not end there.

[IslandPink]

Of course there is the circuit that Steve Bench did - is that not of interest ?

[Paul Barker]

his 841 circuit is great for inspiration, ive lived in it for 22 years.

[IslandPink]

Yes but there's an 801a A2 circuit delivering about 8W - let me see if I can find it.

[steve s]

I've got some 801, 801a and 10y,
One or two of each think...
I'll never use them Paul. they need testing but I can do that.
if you ever need them...

[Paul Barker]

Cool, Ill see if my 801 pair are good first. I have sufficent 10 watt jobbies.

[Paul Barker]

Yes but there's an 801a A2 circuit delivering about 8W - let me see if I can find it.

Yes, thanks for the reminder. He also did a spreadsheet to put youre fgures into and it displade chart and all distortions power etc. But I forgot long ago where that was hosted. The JAC music mirror of his site doesn’t include it, I don’t think.

[IslandPink]

I might have it saved at work. It's not easy to find, I remember than now - I was on the JAC mirror earlier, you're right. But there definitely was an 801 circuit.

[Paul Barker]

I might have it saved at work. But there definitely was an 801 circuit.

Definately, but Im not sure if it was ever on his web pages, might have come via another route, but Ive forgoten so much since those days, I cant get the memory to filter up anything on the matter.

[IslandPink]

I didn't find the whole page, but I did find two crucial pictures :

d_801a2a.jpg (31.78 KiB) Viewed 5045 times

[andrew Ivimey]

This reminds me.... I've rigged up a setup for 801s etc ( as I have a few if this family) ef37s at the ready but started with 6sn7srpp driver to get going in A1. Then the plan was a bias on the grid and bingo, into A2 I was to go.... but I didn't. It's a thought though.

I cant help thinking that Paul's endeavour will be much more interesting and successful. Onwards!

[Paul Barker]

Thanks Mark, the circuit is good to have back.

If I remember rightly the whole spread sheet he’d writen to automate picking operating points and displaying them as if you’d fawn them up on a characteristic curve graph. Pretty sure he sent the file to me because I definately had it and played with it for hours.

Now Im back to ruler and pencil on photocopy, and calculating the distortion manually (very long winded).