Hi,
About pentode-triode combos,
someone knows if there exist Russian equivalents of 6EA8/6U8A/6KD8 ?
Thanks
JJT
About pentode-triode combos,
someone knows if there exist Russian equivalents of 6EA8/6U8A/6KD8 ?
Thanks
JJT
Christmas day is quite the wrong time to point to shortcomings (if there be ever a right day ...), but I would like to point out that with many phase splitters using pentode-triodes, one must take care not to be driving power tubes from too different impedances. The circuit shown by member Hahn (#11) features a pentode driving one end and a feedbacked triode the other. There will clearly be a difference in driving impedances here with probable phase differences above 8 kHz. (No criticism to member Hahn; he showed someone else's design.)
About the only balanced phase inverter convenient for a pentode-triode, both amplitude and phase-wise, is the cathodyne or concertina following a straight pentode. Then one is also wondering about the often un-pentode like bypassing of the screen (g2) to ground, when there is signal on the cathode. It must be kept in mind that true pentode operation is only achieved with the screen bypassed to the cathode. Otherwise it assumes some sort of UL characteristic, depending on component values. I have never seen an analysis of how bad or otherwise this is; which makes one suspect that there was no such intention in the design. Neither have I done/seen distortion or spectrum analyses of such, so I must be careful of 'arm-chair' predictions. Still ...
About the only balanced phase inverter convenient for a pentode-triode, both amplitude and phase-wise, is the cathodyne or concertina following a straight pentode. Then one is also wondering about the often un-pentode like bypassing of the screen (g2) to ground, when there is signal on the cathode. It must be kept in mind that true pentode operation is only achieved with the screen bypassed to the cathode. Otherwise it assumes some sort of UL characteristic, depending on component values. I have never seen an analysis of how bad or otherwise this is; which makes one suspect that there was no such intention in the design. Neither have I done/seen distortion or spectrum analyses of such, so I must be careful of 'arm-chair' predictions. Still ...
This is the universal problem with all the paraphase splitters, and why I consider them to be a "quick 'n' dirty" solution. You'd use something like that where you weren't concerned about fidelity. These triode / pentode types are best used for voltage regulator error amps, mu-stages, or you could use the triode half for a voltage amp driving an LTP splitter, with the pentode working as an active tail load, or use the pentode as a front end amp (either as a pentode or a pseudotriode) with the triode half forming a cathodyne.
I personally like 6X8s as voltage regulator error amps.
What makes you think the UL characteristic is bad? With small signal pentode you only get a small fraction of a volt AC on the cathode. The behavior is actually extremely close to pure pentode and makes no significant difference. Your knit picking.
I've been thinking of throwing together an amp similar to that shown in post #9, using the 6X9/ECF200 and 6P3S-E, using iron (power and output) scavenged from a Fisher 400 receiver. I went as far as to pick a chassis and play with layout. A rival concept using a hefty pentode in triode mode (12HG7comes to mind) and an autotransformer phase splitter, similar to a design from Electra-Print, may knock the 6X8 out of the running. If I can find something among my various transformers that will satisfy the autotransformer function without me having to spend a lot of money, I'll try that first - otherwise, it's the 6X9, funky 10-pin socket and all.
I have used a small Power toroid as a interstage autotransformer phase splitter. Not ideal as the capacitance is quite high, but if you have a beefy enough driver it works very well. Worth trying before splashing any serious cash.
Shoog
6F1P and 6F12P should be somewhere in the vicinity of those, I don't think there is a direct substitute that you could just drop into existing circuit though (slightly different bias will be required and gain might differ anyhow).
Arnulf, many thanks. I'll check those tubes specs to see how it compares with 6EA8s.
Really convenient prized too!
Regards
JJT
Really convenient prized too!
Regards
JJT
Last edited:
I am at a disadvantage here, having already deviated from the topic by a side remark. So just very briefly, I did not say that UL is bad per se, just that it seems that designers grounding G2 instead of returning the bypass C to the cathode (in the case of an unbypassed cathode resistor), don't realise that that changes the pentode behaviour, meaning its characteristics more than basic pentode operation per se - perhaps I was not clear about that. About it making no significant difference (to the circuit) - and again not going into specifics as said - perhaps a quick example, TheGimp allowing.
I use an ECF80 pentode at Ia = about 1 mA and other components adjusted for Va = 120V with an RL of 100K. Consulting the graphs one finds ra = about 2,5 meg and S = about 1,2 mA/V, giving a theoretical µ = about 2000. Using an unbypassed cathode bias resistor of 680 ohms, one will find that with G2 bypassed to earth instead of the cathode, ra now becomes about 160K, etc, apart from changes in other parameters. The unbypassed cathode feedback will move that up again - not to elaborate.
But the bigger change comes when there is also global NFB going to the same unbypassed cathode, as is usual. If one now returns the screen bypass to earth instead of cathode (as I have seen), behaviour differs significantly from the 'pure pentode operation' one originally started off with - no longer 'a fraction of a volt AC on the cathode' (I take it you meant referred to 1V input signal). Again I do not claim disadvantage per se (although the drop in real amplification factor will need to be observed); I am simply saying that some designers seem to do this intuitively 'because screens are earthed', unaware that the circuit is no longer simple common cathode. (I see no good reason to complicate calculations by deviating from classic topology - unless others have done measurements and can broaden my horizons.)
Back to topic, with apology to TheGimp.
Is that comparing unbypassed cathode resistor + G2 bypassed to cathode VS. unbypassed cathode resistor + G2 bypassed to ground? It seems that the initial example is using a bypassed cathode resistor. Bypassing or unbypassing the cathode resistor makes a larger difference than where g2 is bypassed to.
Regardless though, I do agree that if a design is definitely going to use RC decoupling for the screen it should be referenced to cathode. The main advantage of bypassing it to ground is that it gives the user the flexibility to interchange decoupling networks and regulators on the screen without having to change other parts of the circuitry. As long as the rest of the circuitry has been tuned with G2 referenced to ground it's all good.
Out of curiosity though, I'm going to try this amp with the cap tied to cathode (as pictured in the link) vs. as it is now(tied to ground). The cap is currently tied to ground because the amp was initially using a zener on the screen.
http://img.photobucket.com/albums/v685/JebDePaiva/FinalSch.jpg
Apologize for the diversion - however, Miles can you point to an example or please post a schematic for an voltage regulator error amp?? Also would appreciate a short description of their advantages. thanks.
6X8? I love 6X8. If you are short on 6J6's, you can triode-strap the pentode -- look at the construction, it's pretty much the same, and characteristics seem to be pretty close at a glance. Good enough for an LTP when you don't need balanced outputs (or don't mind the distortion, but which since we're talking tubes, we welcome distortion).
But using the pentode as a pentode is even better. Run the triode plate and pentode screen to B+ and take signal out with a resistor from the pentode plate: no miller effect anywhere, so bandwidth is high, easily suitable for a tube PWM generator.
For the same reason, the 6X8 will indeed work well for a voltage regulator: differential input, single ended constant current output.
Personally, when it comes to regulators, I'd rather use one dissimilar dual triode or triode/pentode, since it's not really worth getting out a diffamp -- audio amps aren't Tektronix gear.
I like using triode/pentodes (6U8, etc.) for gain and phase splitter. Frankenhouse uses the pentode for gain, triode as cathodyne. Just barely enough gain driving 6V6's to close the NFB loop. Doing it with triodes would require two gain stages and potentially lead to stability problems. Pentode noise doesn't matter in this application, because it's not a sensitive, low level circuit.
Tim
But using the pentode as a pentode is even better. Run the triode plate and pentode screen to B+ and take signal out with a resistor from the pentode plate: no miller effect anywhere, so bandwidth is high, easily suitable for a tube PWM generator.
For the same reason, the 6X8 will indeed work well for a voltage regulator: differential input, single ended constant current output.
Personally, when it comes to regulators, I'd rather use one dissimilar dual triode or triode/pentode, since it's not really worth getting out a diffamp -- audio amps aren't Tektronix gear.
I like using triode/pentodes (6U8, etc.) for gain and phase splitter. Frankenhouse uses the pentode for gain, triode as cathodyne. Just barely enough gain driving 6V6's to close the NFB loop. Doing it with triodes would require two gain stages and potentially lead to stability problems. Pentode noise doesn't matter in this application, because it's not a sensitive, low level circuit.
Tim
Jeb-D,
Sorry if there is confusion. I did state in my post #22 that one is talking about an unbypassed cathode resistor (not referring to any circuit in particular), and so I assumed did you when stating in your post #24 'a small fraction of a volt AC on the cathode'. Otherwise, obviously as you say.
(But again, if you will allow, even with bypassed cathode one can encounter very low frequency stability problems in a main amplifier with global NFB, if one is unfortunate in the choice of the above regarding phase shift. Nyquist can be brutal!)
Ask and ye shall receive:
The main advantage here is that there is no load on the VR tube. Since the reference voltage is applied to the triode control grid, and cathode coupled to the pentode stage, you can include a first order LPF to help reduce the noise coming from the VR tube.
Another good thing about the 6X8 is its unusually complete spec sheet. This allows making adjustments to the screen voltage. The 6X8 pentode section is also quite a high gain section, making for a more accurate error amp.
You're also going to be connecting the cathodes together anyway, so why not just let the manufacturer do that for you? Besides, since no one wants these "ugly ducklings", they're very inexpensive.
Attachments
the 6bm8 was used in many tape and record players as it made for a cheap way to build an amplifier- i am looking to use this for a guitar amp project...
- Status
- Not open for further replies.