On both this amp, the Citation V, as well as on the Citation II, the first stage voltage amp, in both cases a 12BY7A, utilizes about 2 ohms in the filament circuit to reduce the filament voltage. It's not a case that the filament voltage for all tubes is too high as there are other tubes on the same supply without the series filament voltage reducing resistors. In fact, the other 12BY7A tubes in the Citation II do not have this filament voltage reduction.
Was the purpose to "somehow" make the first stage more linear?
Was the purpose to "somehow" make the first stage more linear?
Can you elaborate on that? What advantage does high gm offer in this circuit?
Being one to like big pentodes for signal amplification, there are ooodles of tubes to run instead of the 12BY7. Gain will vary a bit if you choose higher gm tubes...and on that I will say *AT THE OPERATING POINT, DELIVERED GM that is. An 8233 is 45 mA/V per the data sheet, but at 5 mA I have noooo idear. The D3a is another...but at the OP, its gm will likely be less than the published spec.
On the cascode, if you're going to do that let me suggest a MOSFET instead of a second triode. The twin triode is going to have some issue delivering signal much below 2x the upper element grid voltage. A triode on the bottom( perhaps a 5687 section), with a FQP1N60 on top will swing to nearly its gate voltage. Type 6H6Pi will be similar. The real pentode, at the expense of g2 current has nooooo issue swinging its plate below the g2 voltage.
The upper triode grid voltage is what used to be labeled on maps where topography was unknown as 'Here be monsters'. Go past it knowing the circuit gets a wee bit wonky.
cheers,
Douglas
On the cascode, if you're going to do that let me suggest a MOSFET instead of a second triode. The twin triode is going to have some issue delivering signal much below 2x the upper element grid voltage. A triode on the bottom( perhaps a 5687 section), with a FQP1N60 on top will swing to nearly its gate voltage. Type 6H6Pi will be similar. The real pentode, at the expense of g2 current has nooooo issue swinging its plate below the g2 voltage.
The upper triode grid voltage is what used to be labeled on maps where topography was unknown as 'Here be monsters'. Go past it knowing the circuit gets a wee bit wonky.
cheers,
Douglas
Doug,
FWIW, I considered a hybrid cascode, but was approaching things from a perspective of what to do about keeping existing units functioning well using in production types, when the tubes they use are no longer available. The 12BY7 is fast disappearing
(New Sensor please pay attention) and I wanted to preserve whatever supply is available for the more complicated H/K Cit. 2s.
FWIW, 12BY7 out and 6922 in (IMO) makes an ECC99 LTP more attractive, as the net demand on the heater supply is slightly reduced.
Speaking of hybrid cascode high gm voltage amplifiers, what do you think of using the Russian 6C45Π (6s45p) as the triode at the bottom of the "totem pole"?
FWIW, I considered a hybrid cascode, but was approaching things from a perspective of what to do about keeping existing units functioning well using in production types, when the tubes they use are no longer available. The 12BY7 is fast disappearing
(New Sensor please pay attention) and I wanted to preserve whatever supply is available for the more complicated H/K Cit. 2s.FWIW, 12BY7 out and 6922 in (IMO) makes an ECC99 LTP more attractive, as the net demand on the heater supply is slightly reduced.
Speaking of hybrid cascode high gm voltage amplifiers, what do you think of using the Russian 6C45Π (6s45p) as the triode at the bottom of the "totem pole"?
The triode's suitability goes in about 2 pieces. Gm for gain, and low plate resistance so as to reduce the faux g2 voltage. Since mu basically eats plate resistance, I have not looked at the 6C45Pi...
It is all about gain and output voltage for me.
So far, the only shortage of 12BY7 tubes I have encountered is the ones in the cute liddle Tektronix boxes. The ones that say something like 'Laboratory Matched Pair', and across a fair g2 voltage are very nice indeed.
cheers,
Douglas
It is all about gain and output voltage for me.
So far, the only shortage of 12BY7 tubes I have encountered is the ones in the cute liddle Tektronix boxes. The ones that say something like 'Laboratory Matched Pair', and across a fair g2 voltage are very nice indeed.
cheers,
Douglas
Ok, can we talk a bit about this high gm thing?
High gm makes sense theoretically and certainly in a driver tube. But we're talking about a preamp tube operating into essentially no load. Do we have anything to substantiate that the high gm is actually useful in THIS circumstance?
Does the high gm actually result in a higher slew rate? And if it does, is the 1st stage the limiting factor? i.e. is everything else in the amp capable of higher slew factor than the 1st stage? And is the improvement meaningful? In other words, is the "unimproved" slew rate insufficient? Has anyone done any work in this regard, either simulation or actual circuit measurements?
I just read the Citation II and V manuals. They talk about extending frequency response 2 octaves above and below audible. They talk about the quality of their OPT. They mention video pentodes for wide frequency response. No mention of high gm or slew rate. They do say that their power supply with low Cu losses PT, Si diodes and sturdy electrolytics are as good as a regulated supply. Hmmm...
If someone has an accurate model of the Citation V OPT and the 12BY7 (I don't think the Ayumi is very accurate), I'd be willing to do some sims. But if work has been done in this regard already, let's share it.
High gm makes sense theoretically and certainly in a driver tube. But we're talking about a preamp tube operating into essentially no load. Do we have anything to substantiate that the high gm is actually useful in THIS circumstance?
Does the high gm actually result in a higher slew rate? And if it does, is the 1st stage the limiting factor? i.e. is everything else in the amp capable of higher slew factor than the 1st stage? And is the improvement meaningful? In other words, is the "unimproved" slew rate insufficient? Has anyone done any work in this regard, either simulation or actual circuit measurements?
I just read the Citation II and V manuals. They talk about extending frequency response 2 octaves above and below audible. They talk about the quality of their OPT. They mention video pentodes for wide frequency response. No mention of high gm or slew rate. They do say that their power supply with low Cu losses PT, Si diodes and sturdy electrolytics are as good as a regulated supply. Hmmm...
If someone has an accurate model of the Citation V OPT and the 12BY7 (I don't think the Ayumi is very accurate), I'd be willing to do some sims. But if work has been done in this regard already, let's share it.
If one of y'all wants to characterize a Cit.V output, I have a single. It should go to Heyboer for their un-wind/cloning process, and along with that one, a Cit.ii OPT that is somehow damaged for the same treatment.
The current crop of winders, Heyboer excepted is ridiculous in pricing...it is just a matter of getting them a model to tear down. With the ii OPT, just a warning that it got damaged ahead of taking it apart should be sufficient. And quick before their un-winder retires full time.
That same gent did a Peerless S265Q for me, and the significantly more complex S271S.
For the duty the 12BY7 is running, lots will serve. 6CL6 is one...I'd drop the plate load, and boost the current to make it even mo' stout. While not delivering the gain, a 6V6GTA would also serve...as would EL84 and EL86. This is not a big deal IMO, the 6CG7 is going to be a bigger limit than either of those...
cheers,
Douglas
The current crop of winders, Heyboer excepted is ridiculous in pricing...it is just a matter of getting them a model to tear down. With the ii OPT, just a warning that it got damaged ahead of taking it apart should be sufficient. And quick before their un-winder retires full time.
That same gent did a Peerless S265Q for me, and the significantly more complex S271S.
For the duty the 12BY7 is running, lots will serve. 6CL6 is one...I'd drop the plate load, and boost the current to make it even mo' stout. While not delivering the gain, a 6V6GTA would also serve...as would EL84 and EL86. This is not a big deal IMO, the 6CG7 is going to be a bigger limit than either of those...
cheers,
Douglas
I just finished running characterization of the few (used) samples of the 12HL7 and 12GN7 along with new samples of 12BY7.
From my limited testing, they are not plug in replacements. Biasing and plate load need to change for optimum operation.
I have new 12GN7 and 12HL7 on order and will perform the test again when they come in.
My interest is in using them as a driver for a 300B.
From my limited testing, they are not plug in replacements. Biasing and plate load need to change for optimum operation.
I have new 12GN7 and 12HL7 on order and will perform the test again when they come in.
My interest is in using them as a driver for a 300B.
when i say plug in, i mean they have the same pin-outs....9BF..
even the 11hm7 has same pinouts.....High Gm driver pentodes
even the 11hm7 has same pinouts.....High Gm driver pentodes
I agree. The first stage is direct coupled to the next stage so practically no load. I don't see the big fuss about it. Yes, I agree that driver tube needs muscles. But for an input tube? Plus, subjectively I never like the tone of high gm tubes, always very "hifi" sounding, a nervousness to the sound to me. I personally prefer a more euphonic sound. I want tube to sound like tubes so the "straight wire with gain" concept is not for me. I think EF86 would be perfect for me if I ever get to mod a Citation V.
how can a tube not sound like a tube? to me there are only bad sounding circuits with tubes, and many in between, and also good sounding circuits with tubes, they are all tubes...
and i am of the belief that it is not the fault of the tube but the one who designed with them...
and i am of the belief that it is not the fault of the tube but the one who designed with them...
Last edited:
A few weeks ago I was toying with the idea of using a triode strapped pentode to drive a 300B.
I dug out some candidates and ran some quick tests to see how they would do and was pleasantly surprised, particularly with the 6DX8 video amp pentode section.
This got me doing some reading which lead me to this thread.
Since several of the tubes I tested were used and I had no new samples I ordered 20 12GN7s and 10 12HL7s to test with.
Those came in recently and I have been testing them.
I dug out some candidates and ran some quick tests to see how they would do and was pleasantly surprised, particularly with the 6DX8 video amp pentode section.
This got me doing some reading which lead me to this thread.
Since several of the tubes I tested were used and I had no new samples I ordered 20 12GN7s and 10 12HL7s to test with.
Those came in recently and I have been testing them.
