Horz Sweep amp Pent or screen drive

[mashaffer]

As I am in the habit of doing I am looking ahead a project or two. I have read a bunch of threads on horizontal sweep tubes and I have yet to figure out which method of use would be the most desirable WRT ease of building with minimal test equipment and best performance.

I have on hand one each of
6JB6
6GE5
6JN6
6JS6
17JM6
17JN6
22JF6
33GY7A
36KD6
38HE7

and three
6LQ6

Given that the only one I have multiple copies of is the 6LQ6 I thought I might make a PP subwoofer amp to drive a tapped horn sub or something like that. A couple of these tubes seem to be virtually identical except for the socket used or wiring thereof but that seems like a really kludgy way to build a PP amp. Maybe a couple of SE mono blocks or something down the road.

The question is do I go pentode mode with regulated screen supply (Maybe a gas regulator tube) or screen drive with FET drivers? What are the advantages and disadvantages of each?

Since the 6LQ6s are so valuable I suppose I could sell them and buy a bunch of one of the other tubes. Or selling all of the other sweep tubes might net me enough to buy one more 6LQ6.

Another possibility might be some swapping with others who have a mismatched tube collection so that we can all end up with some matching sets.

 

[Miles Prower]

The question is do I go pentode mode with regulated screen supply (Maybe a gas regulator tube) or screen drive with FET drivers? What are the advantages and disadvantages of each?

Screen drive can give better linearity. However, it also has a higher drive requirement, and you may simply lose back in driver distortion what you gained in final linearity. I did an amp with the 6BQ6GTB horizontal deflection VT. Used PP pentode mode with active screen regulation. Worked great, and the 6BQ6GTB turned out to have a friendly harmonic signature: Mostly h3 with but a trace of h5. It didn't require much help from gNFB to clean up whatever open loop pentode nastiness was there.

Maybe a couple of SE mono blocks or something down the road.

Doubtful that'll work. These HD VTs like to pull some big currents, and aren't likely to operate Class A1 without red plate destruction to hit the most linear part of the characteristic.

 

[mashaffer]

Thanks Miles.

Screen drive can give better linearity. However, it also has a higher drive requirement, and you may simply lose back in driver distortion what you gained in final linearity. I did an amp with the 6BQ6GTB horizontal deflection VT. Used PP pentode mode with active screen regulation. Worked great, and the 6BQ6GTB turned out to have a friendly harmonic signature: Mostly h3 with but a trace of h5. It didn't require much help from gNFB to clean up whatever open loop pentode nastiness was there.

Interesting to hear of your success. I note that McIntosh made a couple of well respected power amps using 6LQ6 in pentode mode so there is another data point favoring the pentode mode. I don't quite understand their driver and bias setup. As I read the schematic they are using something akin to a CF to drive the grids of the 6LQ6s but the drivers look a little funny. The plates of the two drivers are connected to a 470V B+ at the CT of an OPT winding and the cathodes to the the ends of the winding. I don't follow that at all. I don't get the ground reference for the cathode.

I had assumed either a very high performance CF or a SS follower for adequate current drive but I infer from what you are saying that voltage swing will need to be substantial as well due to lowish mu. Is that correct.

Doubtful that'll work. These HD VTs like to pull some big currents, and aren't likely to operate Class A1 without red plate destruction to hit the most linear part of the characteristic.

Sounds like HDTs like PP. That's OK I guess. There are other candidates for midrange/fullrange amps. Use these guys for bass where they naturally fit the bill.

 

[SY]

Mac used a unity-coupled output stage, a pretty special circuit. Bootstrapping, special transformers, you name it. Draw some arrows and you'll see that all the DC currents in the transformer primary cancel out.

edit: There was a fine article about this topology in November 1959 "Audio." The author was Norman Crowhurst.

 

[mashaffer]

Thanks SY. Pretty tricky are they hmmm...

BTW I seem to remember that you have a bit of experience with these things. Any thoughts to add.

 

[SY]

I use p-p screen drive on my amp, run very close to class B. Distortion is surprisingly low; the tubes act very much like zero-bias triodes. SE is a lot trickier because you'd need to run A2 if you want to screen drive.

Pentode mode for HD tubes tends to be pretty nonlinear, but as Miles said, using screen drive makes the driver stage a challenge. It's a fun challenge.

 

[mashaffer]

Thanks SY. For a subwoofer or musical instrument amp it seems like a bit of non-linearity might not be a big problem as some GNF is not such a problem in those applications. Subs being driven by SS seems to work pretty well so Pentode mode with GNF should work pretty well also I would guess. It might also be fine for a full range amp in non-critical listening systems such as a garage or work room system.

What do you think of using gas regulator tubes to control the screen voltage in pentode mode?

mike

 

[SY]

You're asking someone who thoroughly believes that tubes shouldn't be used for subwoofers, and especially not pentodes.

 

[kenpeter]

What do you think of using gas regulator tubes to control the screen voltage?

Screen drive? A2? Gas Tubes??? What, are you crazy?!?

I wasn't very far along with this one when I set it aside...
Its got errors I am sure. Something I been meaning to debug
and finish. I have actual intentions of building, go figure...

Input stage sorta Lofty, intended to float up on to B+ noise.
Goin back down, its a FETron P-CH triode emulator, abusing
a screen driven Pentode as if it were an enhancement FET.

 

[mashaffer]

Screen drive? A2? Gas Tubes??? What, are you crazy?!?

Actually I was thinking gas regulator for screen supply in pentode mode.

SY, I would be interested in your thoughts on why pentodes are especially inappropriate for subwoofer duty. I am assuming a reasonable amount of GNF for the sake of Zout. Also do you not like tubes for Sub duty because SS does it easier or because tubes are in some way unable to do the job properly.

So if I do use them for subwoofer duty screen drive (which I take it is pretty much a triode) is what you would suggest?

As far as power goes my current 8 ohm 12" sealed sub is working fine with plenty of output with a plate amp rated at about 120W into 4 ohms and the next sub should be more efficient yet (tapped horn) so power shouldn't be an issue.

Would need an OPT with plenty of inductance on the primary but that should be doable especially since highs don't matter.

As far as it goes I haven't committed to doing sub with these. I could also get another 6LQ6 (or three of one of the cheaper ones) and use it for the regular woofer amp (60-200Hz).

P.S. SY, didn't you post a schematic of source follower driver for screen drive somewhere here?

 

[SY]

I probably posted the driver at some point, but can't remember where. Tubelab's circuit is pretty much the same thing; we seem to have independently arrived at the same answer. Nothing complex, just an IRF820 as a plain vanilla source follower, with the drain connected to +200, and the source connected to -200 via a 100k resistor, then direct coupled to the screen. Bias supply feeds the gate via a large resistor (I think it was 1M). The driver is not class A, strictly, but the distortion is low nonetheless.

The problem with deep bass is the output transformer. And when the OPT is driven from a high source impedance (like a pentode), the LF problem gets much, much worse.

 

[mashaffer]

Thanks SY, I had forgotten that it was bi-polar (manic-depressive?) power supply and fixed bias. Will do some searching and have a peek.

 

[kenpeter]

Actually I was thinking gas regulator for screen supply in pentode mode.

A well bypassed 0C2 might drop a good 75 volts from plate to
screen in a triode strap. As long as screen current stays within
the 5-30mA range. And never drops low enough to let the pilot
light go out... Small current source from screen to to ground
could insure the gas regulator always remained lit.

My screen saver 0A2 above, intended to be almost always off.
Remove itself from the audio circuit as safe operating voltages
for G2 are realized... Definately not the normal way to abuse
such a regulator tube.

Important to keep in mind that gas tubes are noisy, have internal
resistance, and imperfect regulators of voltage where current varies.
All quite manageable in the right circuit.

 

[mashaffer]

Well I wasn't thinking of triode mode but a regulated screen supply for pentode mode. So a screen supply of say 175 volts would be fed to 150V worth of regulator tubes and then to the screens. Of course that would demand a regulated B+ too I suppose. At that point one has to start thinking about SS regulators.

Is it normal practice to shunt gas tubes with a cap to kill the noise?

 

[mashaffer]

I know it is expensive and so I would be unlikely to use it in the near future but has anyone tried or seen inter-stage tranny used as a driver for screen drive? I know that tubelab tried one briefly in his power drive work but I not found anything in googling about that shows it's use for screen drive.

 

[kenpeter]

How's this Screen Driven Cascodlington for low parts count?

Triode in "4th circuit" should be regulating the Sandy/Pent cascode,
making all behave as-if one really big triode? Or at least I think so...

Its got your 4th circuit. Its got your gas tube. Its got your LED.
Its got yer screen drive. Its got your A2. Its got your Cascode.
Its got your current multiplier. Its got your herbs and spices...
Its in there...

If it ain't a spaghetti sauce, its Circuit Succotash for sure.

 

[SY]

Is it normal practice to shunt gas tubes with a cap to kill the noise?

Only small values, not enough to make a big difference. Otherwise, Google "relaxation oscillator."

 

[kenpeter]

That could be a problem....

But only if the current isn't high enough to keep it lit continuously.
There should be no negative resistance concerns above 5mA...
A positive resistance won't oscillate... Well, at least it shouldn't...

 

[SY]

I dunno. How serious is the noise at these levels, taking into account the various shunting impedances?

 

[kenpeter]

I am an idiot!

Screen tied to Plate. Grid tied to Cathode.
Not an active amplifying device wired this way.
Its nothing but a big power wasting Diode.

It probably still works, but not as advertised.

-------------------------------------------------------

If I were to ground the Pentode's grid though a
suitable resistor. Then perhaps it does something
meaningful again.

The mode would be Triode, and not even A2...

Though the screen is still delivering a current,
it can't be "screen driven" when G2 has clearly
been bypassed and strapped to the plate...

I'll re-draw with a correction in a minute.