Direct Coupling with Zener Diodes

[Steve Conner]

Well, 12AX7s aren't used in the positive grid region in guitar amps, but like any other tube, they still work. Tim de Paravicini (?) designed a hi-fi amp powered by about 12 of them in Class AB2.

The RC coupled circuits in guitar amps don't have enough balls to push the grid more than a tiny bit positive.

To the original poster: looks like you've embarked on quite an unusual journey. You may be travelling alone, as you seem to see a different problem to the rest of us. (I would say you are solving the wrong problem, but I gave up objectivism for Lent.) I think you'll end up going back to AC coupling, and using a diode bridge ala JCM 900, or a couple of zeners ala Paul Ruby/Chuck H, to keep the grid voltage bounded under overdrive. Then you can use as big coupling caps as you like with no blocking distortion.

[jmaf]

Steve, I'd be curious to learn more about these circuits work the tubes positive(power or 12ax7's). Can you point me to some schematics?

[Steve Conner]

Well, if you're looking for a circuit where the grid voltage is positive under idle conditions, there is no such thing. The closest is the so-called "zero bias" triodes, these are dirty great tubes like the 3-500Z, used in high power audio and RF amps. The bias voltage is, as the name suggests, zero, and at this condition the tube is designed to pass a modest idle current. The grid is driven positive for its entire working half-cycle. A push-pull pair with transformer drive to the grids makes a great high-power amp, as easy to drive as a Class-AB2 amp ever can be.

All other circuits only push the grid positive on signal peaks. See the Fender 300/400PS or Ken's BAGA.

With the Chuck H mod, the grid-cathode diode must pass as much current in one direction as the Zener does in the other. Therefore, it allows more grid current.

[jmaf]

Well, if you're looking for a circuit where the grid voltage is positive under idle conditions, there is no such thing. The closest is the so-called "zero bias" triodes, these are dirty great tubes like the 3-500Z, used in high power audio and RF amps. The bias voltage is, as the name suggests, zero, and at this condition the tube is designed to pass a modest idle current. The grid is driven positive for its entire working half-cycle. A push-pull pair with transformer drive to the grids makes a great high-power amp, as easy to drive as a Class-AB2 amp ever can be.

All other circuits only push the grid positive on signal peaks. See the Fender 300/400PS or Ken's BAGA.

With the Chuck H mod, the grid-cathode diode must pass as much current in one direction as the Zener does in the other. Therefore, it allows more grid current.

Thanks for the info, Steve. I'm glad that within my universe of guitar amp tubes, not RF or specialty tubes, what I said above isn't complete gibberish.

[kg]

Grids cannot be biased positive on any tube.

how on earth did you come up with that?

i just don't know how to even start to address it. it's simply completely utterly wrong.

[jmaf]

how on earth did you come up with that?

i just don't know how to even start to address it. it's simply completely utterly wrong.

kg, I don't claim to know it all, but on your first post you just dissed what I said, offering no explanation. Then now you dismissed it again, without offering an explanation. Same with redelephant, called what I said nonsense and so forth.

Please, I'm here to learn and listen as well, so tell me an example of positive biased grids, how that works and where I messed up please. Honest request and very much appreciated. Thanks very much in advance.

[Steve Conner]

Well come on then Ken, work that copy of Spangenberg!

The only scenario I can think of is some kind of extreme audiophile who wanted to use a 3-500Z in single ended Class A, at a relatively low B+. Since they don't pass enough current for Class-A operation at zero bias, the grid would have to be persuaded positive with a MOSFET or something. But anyone who built that probably needs locked up for their own safety.

As far as I see the Marshall CF, it'll draw grid current if the tone stack loads the cathode too heavily.

[kg]

kg, I don't claim to know it all, but on your first post you just dissed what I said, offering no explanation. Then now you dismissed it again, without offering an explanation. Same with redelephant, called what I said nonsense and so forth.

Please, I'm here to learn and listen as well, so tell me an example of positive biased grids, how that works and where I messed up please. Honest request and very much appreciated. Thanks very much in advance.

jmaf,

i don't mean to sound dismissive. it's just that this isn't even an matter of subjective interpretation or grey area.

the vast majority of tube circuits do not use grid current. that's because it requires actual power dissipation from the driver stage. if you're going to do that, you might as well use BJTs (only half joking). keeping the grid negative means no current, which greatly relaxes driver design, since it only has to drive electrode capacitance, and can use easy/cheap RC coupling.

as steve correctly points out there are some tubes which are better suited for grid current operation. or perhaps one could say they are tubes which are mostly useless if grid current is avoided.

however, there is absolutely nothing that prevents ANY tube from being operated that way. the idea that tubes will lose their ability to modulate plate current when Vgk>0 is just plain wrong. i can't explain it any other way. it just is not true.

[jmaf]

jmaf,

i don't mean to sound dismissive. it's just that this isn't even an matter of subjective interpretation or grey area.

the vast majority of tube circuits do not use grid current. that's because it requires actual power dissipation from the driver stage. if you're going to do that, you might as well use BJTs (only half joking). keeping the grid negative means no current, which greatly relaxes driver design, since it only has to drive electrode capacitance, and can use easy/cheap RC coupling.

as steve correctly points out there are some tubes which are better suited for grid current operation. or perhaps one could say they are tubes which are mostly useless if grid current is avoided.

however, there is absolutely nothing that prevents ANY tube from being operated that way. the idea that tubes will lose their ability to modulate plate current when Vgk>0 is just plain wrong. i can't explain it any other way. it just is not true.

OK, surely then there must be well known circuits which bias the grid positive. Can you point me to one so I can better understand this? Especially one with a positive biased preamp tube which is the discussion here. Thanks.

[kg]

i don't know if you're deliberately being obtuse, or you're having a go at me, or what.

positive quiescent Vgk bias in ubiquitous commercial amp designs is NOT the discussion here, despite your attempts to redirect it that way.

the statements that you've made that i find utterly wrong have already been quoted by me in this thread. scroll up--they're there.

if you continue to insist on adhering to them, you will continue to lose my respect. i have told you the way it is and pointed you in the right direction. whether or not you do your due diligence to self-educate is completely up to you. i do not owe anyone an explanation, reason, or example, for anything that i have claimed. frankly, i think your attitude and tone has been adversarial from the get-go and i have better things to do with my time than spend it trying to convey knowledge and wisdom to people who do not want to retain an open mind or even attempt to figure things out for themselves.

a decade ago i built myself a very large tube power amp for musical instrument use, using ONLY receiving tubes--which, by tradition, are not typically used in the positive grid bias region--and i assure you without question that 1) the output stage can and will draw grid current on positive signal peaks, 2) the amp works, to the tune of 2 x 300wrms, and sounds fantastic, 3) the same beam power tubes i installed back then are still in use today, and 4) the design was based on an amalgamation of time-tested textbook practices as well as real world examples of existing amp circuits.

i do not make shit up. i speak from experience.

[jmaf]

We didn't start by talking about output stages, you threw power tubes in later. I was talking about the preamp dc-coupled tubes, which do not modulate above zero volts, they kill the signal, it's basically a ground for a high impedance source(~ 70K ohms from the previous plate?), I may sound confusing and I am not a very good teacher. Also from bench experience, as I am no Einstein when it comes to theory but you know better than I that positive grids destroy preamp tone.

You are more senior than I you ridiculed my opinion earlier like I was making shit up, I was sort of offended by that, thus my replies. By it's ok, and I just asked you for a single circuit that I can study that has the tube grids biased positive. You were very assertive saying how stupid my opinion was, right? Surely you gotta have something there to help me learn? I am just trying to be respectful, because though I am not an encyclopedia I have never seen positive biased tubes in audio circuits....guitar tube amp circuits, to be precise.

Now, your line above, about losing your respect and such, first of all we don't need to go there.... I respect you and everyone on this forum, I hope we can continue to have these discussions without getting personal at all, I am not judging you, I haven't even met you unfortunately so I am not even close to offending you. I know you don't make shit up and if at any point I made it look that way, I apologise.

[J M Fahey]

I think two very different things are being mixed up here.
1) Are preamp tubes (to keep the discussion outside of AB2 power amplifiers) normally used with positive bias?
Not usually, although I *DO* know a couple which do (surprise).
As a side note: are these last ones exotic products of a feeble/wicked mind?
Not at all, they are respected products in their own area, often used live in great stages of the world. he he.
2) Ok, ok, besides being popular or not, back to the point: *can* tube grids be positive? or not?
OF COURSE THEY CAN. WHY NOT?. What did I say just 2 lines above?
3) Can a triode preamp be designed, with grid positive biased at, say, +1V (with cathode grounded straight to avoid any hidden self bias discussion) and have it work as any other triode, that is amplifying the signal present at its grid?
Of course.
4) Are we talking monster water cooled triodes here?
I was rather thinking of some 12A*7 ... or some pentode such as an EF86 ... or any other "receiving tube"

[Chuck H]

Let's not go into semantics... Do grids swing positive? YES. Are there amp designs which do it intentionally? NO. jmaf is talking about a practice and kg is talking about an incidental reality. And never the twain shall meet it seems. I have to side with kg in that it's a fact to be dealt with and not an avoidable ideal.

[exclamationmark]

Back on toooopic. I'm confused about these mystical constant current sources. If I were to do it say, as per redelephants schematic, wouldn't the CCS be trying to pull an additional 2mA through the plate resistor, and by virtue of ohm's law would presumably 'clamp' the voltage across the resistor to the B+ (or very near it)?

Could I place the CCS before the zener and the other end to a voltage source (or other alternate arrangement) if the above scenario does occur? I know in cathode LED bias many people stick a CCS at positive potential between the cathode and LED so the tube doesn't attempt to pull a ridiculous amount of current. This ensures that the LED is drawing enough current to operate linearly whilst the tube above doesn't really notice anything amiss.

Thanks for all your help so far!

[kg]

like i pointed out, i believe that the VERY common plate loaded triode direct coupled to cathode follower pulls grid current on EVERY large signal positive peak. as soon as you cut off that plate loaded stage, trace the resulting voltages and currents. imo it's far more common than people think.

i spent a little time this morning digging through some datasheets. there is a dearth of plate curves for positive grid voltages for the 12ax7. however, there are some quite widely used small signal/voltage amp tubes that DO have characteristic curves for this region. i will link to them below.

6sn7
http://tdsl.duncanamps.com/dcigna/tu...ge/6sn7-4h.gif
Vg up to +35v

6922
http://tdsl.duncanamps.com/dcigna/tu...ex/6922-4h.gif
http://tdsl.duncanamps.com/dcigna/tu...ex/6922-5h.gif
Vg up to +4v

12at7
http://frank.pocnet.net/sheets/087/1/12AT7WA.pdf page 5
Vg up to +6v
http://frank.pocnet.net/sheets/138/1/12AT7WA.pdf page 8
Vg up to +8v

12bh7
http://tdsl.duncanamps.com/dcigna/tu.../ge/bh7-4h.gif
Vg up to +25v

please note on ALL of the above characteristics the remarkable linearity despite positive Vgk (and positive grid current flow).