Yep, the separate CCSes will work only with bypass caps or zeners.
Or, with a single non-polar cap bridged between cathodes (my preference), or, as Shoog pointed out, two back-to-back el caps, but I liked it better adding 1M R reference to the ground. The sound difference with R is very subtle but exists.
The separate CCSes make very sense if the tubes can't be well-matched i.e. the supply is limited, otherwise I would not bother. Overally I liked how such setup sounds.
Using a choke instead of CCS in output stage seems very interesting, I never thought of this, any example?
Thanks for explaining.
Common choke in power stage was used in several Western Electric triode amplifiers. Here is an example:
http://www.westernelectric.com/static/library/specifications/amplifiers/59.pdf
And here is another one where common choke is used in differential driver stage:
https://archive.org/stream/am-1950-04/Audio-1950-Apr#page/n13/mode/2up
From my several months experience on testing/auditioning the CCSes in any possible combinations on my SHIII project I'd say the idea of a common cathode choke looks superb. Hell, it's gonna be weighty and take some space but the sound must be superb from the purist's point: it makes less silicone involved into the sound path.
Here is e.g. 60 mA 420E that could work for smaller bottles:
1pc AC Power Choke 15H 60mA 60Hz R=420? Tube Pre Amplifier 12AX7 12AU7 6BQ5 | eBay
I think the needed one could be ordered, some other 150 mA 110E one:
Raphaelite Choke 10H 150mA Tank seal inductance for audio use | eBay
or its cheaper version:
Raphaelite Choke 10H 150mA Clasping inductance for audio amplifier use hifi diy | eBay
Here is e.g. 60 mA 420E that could work for smaller bottles:
1pc AC Power Choke 15H 60mA 60Hz R=420? Tube Pre Amplifier 12AX7 12AU7 6BQ5 | eBay
I think the needed one could be ordered, some other 150 mA 110E one:
Raphaelite Choke 10H 150mA Tank seal inductance for audio use | eBay
or its cheaper version:
Raphaelite Choke 10H 150mA Clasping inductance for audio amplifier use hifi diy | eBay
To my ears, MOSFET CCSes also don't sound right. Simple BJT are somewhat better, but chokes are the best. Sizing a choke for this duty, one should keep in mind that inductance of a choke is not constant, but depends on frequency and flux density. Inductance at 20 Hz may be several times higher than at 100 Hz.
Old time chokes with paper between layers of wire are superior to new production ones that don't have insulation between layers. Old output transformers, both SE and PP, can be converted to audio choke duty. Re-stacking old chokes with better quality modern laminations is also a good idea.
Common choke is especially good idea for transformer-coupled differential driver stage, where such choke may have reasonable size. For power stage, a choke needs to have inductance comparable to that of 1/2 of OPT primary winding, carry full DC current of the stage, and have reasonable DC resistance. These chokes may be really large, 3-4 times larger than OPT. But, with improved Class A stage symmetry and better 2nd harmonic rejection, it is worth it.
Old time chokes with paper between layers of wire are superior to new production ones that don't have insulation between layers. Old output transformers, both SE and PP, can be converted to audio choke duty. Re-stacking old chokes with better quality modern laminations is also a good idea.
Common choke is especially good idea for transformer-coupled differential driver stage, where such choke may have reasonable size. For power stage, a choke needs to have inductance comparable to that of 1/2 of OPT primary winding, carry full DC current of the stage, and have reasonable DC resistance. These chokes may be really large, 3-4 times larger than OPT. But, with improved Class A stage symmetry and better 2nd harmonic rejection, it is worth it.
Maybe you speak for US made new production chokes?
I know of several European brands, and quite likely a few Asian brands as well, who make layer insulated chokes. A good choke is not just randomly winding an x amount of turns on a bobbin.
Superior in lower shunt capacitance. Extra insulation between layers decreases capacitance.
Burning out is a different issue. New production chokes are much better in this respect because they use wire with significantly better insulation than old time lacquer. Relying on high quality magnet wire insulation, most current manufacturers dispense with extra insulation between layers. So, these chokes are less prone to burning out, but are not as good as old ones for audio use.
I think you refer to the resistance part only. The choke's inductance is what makes it acting somewhat to a shunting el cap, for which the opposite "shunt inductance" would be a flaw. But, on my (maybe lame) understanding that choke, in fact, works more on a CCS manner, and that is perfect for PP output stage with common cathodes.
yes, the choke has a dc resistance that biases the output tubes to dc operating points....
the inductance of the choke is what makes the choke impedance look larger than 500 ohms, way much higher...
so that if that choke has say 15H, then at 20 hz., inductive reactance is 1884 ohms, the total then is 1950 ohms...
a mosfet will have much much more, so much more that perhaps that was the reason Garry Pimm used it in this amp...
the inductance of the choke is what makes the choke impedance look larger than 500 ohms, way much higher...
so that if that choke has say 15H, then at 20 hz., inductive reactance is 1884 ohms, the total then is 1950 ohms...
a mosfet will have much much more, so much more that perhaps that was the reason Garry Pimm used it in this amp...
That's right. Choke's impedance increases with frequency until shunt capacitance comes into play and impedance starts decreasing. Within its frequency range, choke acts as long tail or CCS.
Tnx for validating my points, I got all these from numerous experimenting on CCS on output stages. Nubes, do not! shunt the common cathode CCS with a cap! neither use a Zener or LEDs for it, it defeats the whole idea.
Indeed, for the big bottles the cathode choke gotta be enormous, but this idea gets more and more to my mind, I'm thinking of a PP on "those sweet TV tubes" and specifically 6KY8, of which I got some in my stash. The common cathode choke for these with 2 x 30 mA would be not crazy big (one of those evil-bay links that I posted), and R should be IFAIR about 200E, doable.
Which, in turn, makes me wondering if the Amp, only consisting of Tubes and Inductances could be built, well, with a couple of paper-oil caps in the PS. Examples? I believe they existed like 100 yrs ago, here we go: Helix of History. ..
Indeed, for the big bottles the cathode choke gotta be enormous, but this idea gets more and more to my mind, I'm thinking of a PP on "those sweet TV tubes" and specifically 6KY8, of which I got some in my stash. The common cathode choke for these with 2 x 30 mA would be not crazy big (one of those evil-bay links that I posted), and R should be IFAIR about 200E, doable.
Which, in turn, makes me wondering if the Amp, only consisting of Tubes and Inductances could be built, well, with a couple of paper-oil caps in the PS. Examples? I believe they existed like 100 yrs ago, here we go: Helix of History. ..
The old time philosophy was that for sound quality, inductors and transformers are good, whereas capacitors are bad and should be reduced to unavoidable minimum, in both numbers and values.
A caveat about common choke or CCS in power stage is that they strictly limit operation to Class A, and are incompatible with AB operation.
A caveat about common choke or CCS in power stage is that they strictly limit operation to Class A, and are incompatible with AB operation.
May be I read the schematic wrong (it's very poorly drawn)... but there appears to be two separate filament windings for each of the 252's, i.e., no common choke.
No, no, no, no, no.
If it was smoothing choke, there would be a capacitor from output transformer CT to ground, but there is no such capacitor. I am talking about lead 2 of L2 in the schematic. L2 is common choke that acts similar to CCS.
Yes.
Another way is to apply square wave, measure ringing frequency, and calculate capacitance using the formula F=1/2pi sqrt LC.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.