Triode connecting multi-grid tubes
Date: Mon, 08 May 2000 10:26:19 -0600
Subject: Re: Sv: [JN] Triode Mode
From: “PEARL Cust Serv”
Kurt Th. Steffensen wrote:
> I do not know how Bill does the trick , and I respect that he wants to keep that to himself.
> However I described such a circuit on this list , maybe12-18 months ago.
==================
Hi Kurt, Joes:
I don’t care to keep this any great secret . . I just worked hard to get the theory into my head – much digging thru old texts on tetrode theory – and I didn’t feel like giving it away . . . just wanted to have some fun is all, and maybe stimulate a little thinking and re-evaluation of some tired old “standard procedures.”
The way I do this little trick – as a couple of folks already knew – is to bypass a zener with a cap. But, not just any zener and not just any cap.
My favorite zener is a 1N821, a 15cent, 1/2 watt, 6.2 volt part that is both very quiet and of nearly zero tempco. I forget why 6.2V is the “magic” voltage for zeners but it is, the noise falls thru the floor and the low tempco yields a very stable voltage source . Maybe someone could post on this 6.2V business . . .
To build this:
– string 5 of the 1N821s in series and set them aside.
– take two, 16V, 470uF Sanyo OSCon caps and put a 10k resistor across each one, then hook the two RC combos in series; then connect that assy. across the zener string being sure to observe the polarity.
– then put a ten ohm, one quarter watt resistor on the cathode end of the string and a resistive ferrite bead on the other.
– connect the other end of the ten ohm resistor to the plate of the tube to be “triode connected” and the ferrite end of the assembly to the screen, and there you have it.
Note that if you are using true, three grid EL34s that the suppressor is best connected to about -5V or so. Just voltage divide off your bias supply. Biasing that grid slightly negative reduces its electron interception to essentially zero.
Have fun discovering what a triode connected tetrode/pentode really sounds like.
Bill – PEARL, Inc.
—————————————————–
Date: Wed, 10 May 2000 13:51:27 -0600
Subject: [JN] Pet/Tets in”Triode” – was Quiet???
From: “PEARL Cust Serv”
Hi:
> OK Hugh.
>
> Will someone PLEASE explain what the zener thing for strapping pentodes as
> triodes is all about?
>
> I asked once and so far, everyone’s ignored me.
>
> Regards,
> Paul
Hi Paul, All:
Basically, this is just a method of “level shifting” the screen to a lower DC potential than the plate and doing so thru a network with *very* low Z at audio frequencies.
Why is this a good thing?
1 – Well if you look at the plate curves for pentode/beam tetrode operation you see that plate current is pretty much a function of screen voltage and that’s why most amplification stages that employ these devices also regulate the screen supply. Otherwise plate current can be affected by the changes in screen voltage that will inevitably occur with the changes in screen current that accompany signal. The object being of course to provide a low-Z or “stiff” screen supply.
2 – While it’s the desired case for the turns of the screen grid lie within the “electron shadow” created by the negative charge on the control grid, this seldom turns out to be achieved in practice. Metal ceramic transmitting tetrodes being a notable and *very* interesting exception ie. 4CX-125, -250, -350 and their ilk.
When the screen is less than perfectly shadowed, its ability to attract electrons passing thru its accelerating, positive field on their way to the plate increases dramatically. Current at voltage equals power and consequently the screen must be able to effectively dissipate this, otherwise it will melt down eventually.
3 – The more nearly equal the screen and plate voltages, the greater the screen current, and the greater the changes in its value with signal.
4 – In “triode connected” pet/tets, this issue is typically “addressed” by the insertion of 100 to 1K ohms in series with the screen. While this *will* reduce the screen’s voltage wrt to the plate and with that, the screen’s current and consequent requirement for dissipation, it is far from an ideal fix.
5 – Referring back to (1), it will be seen that plate current is a function of screen voltage. If the screen is intercepting electronsdestined for the plate and if, triode-like, its current flow varies with it voltage it can be seen that that the potential on the screen will vary with the voltage drop across the resistor connecting it to the plate.
6 – By this action, the effective impedance of the entire B+ supply is driven up by the insertion of a screen resistor. This action is the cause of the “softish,” “mellow” sound universally reported when pet/tet/UL amps are “triode-connected”
7 – The solution for this is to “level-shift” the screen to a lower DC level – thus limiting its dissipation – and to bypass the level-shifting circuitry with a *low* impedance ie. a *very* good, high value, low voltage cap of the BlackGate, Cerafine, OsCon variety.
8 – 1N821s work well for this as they are sonically reasonably benign and completely bypassed down to very low audio frequencies.
So, there you go Paul. I hope this is satisfactory . . . .
That will be $US250.00 . . . . Cash, VISA or MC :>)
Best regards,
Bill – PEARL, Inc.
