|
|
N-set
Gdansk, Poland
Posts 616
Joined on 01-07-2006
|
Post #:
|
61
|
|
Post ID:
|
14904
|
|
Reply to:
|
14903
|
|
|
|
|
|
fiogf49gjkf0d (Romy feel free to move this post)
 Paul S wrote: | | N-set, I understand you do not mean to hijack the thread with the Axiom, so I actually tried to make generic and/or thread-related observations. I don't know the 6E5P, but generic issues with specific DC tank circuits and driver and output tube stability therein remain. The Axiom appears to address and capitalize on specific/generic issues in very specific ways, and - not to presume - it is perhaps not a great candidate for a generic, mix-n-match approach. YMMV, etc. |
|
Paul, I'm not really sure how specific it is. I think you are mistifing it a bit ;-), but it's also quite possible I don't understand the circuit at all! At a risk of making an idiot out of myself (which is not a big effort) here is how I see it: DC coupled, choke loaded, stacked power supplies. The gross of the biasing voltage is generated as a voltage drop on the big DCR's of the loading chokes. This is something that stacked supplies allow! Very smart, but IMHO requires stability. Then, the remaining voltage is generated via autobias, but in order to diminish the effects of the cathode feedback, diodes are used instead of resistors (lower dynamical R than statical). The diodes are generic, cheap TV dampers (apart from the lowest one as it sees lower current), nothing special. Most probably they were chosen in a very simple and natural way: to give the desired voltage drop at the chosen questient current. That's all. Look in the tube manuals and pick a (good, linear) diode to do the job. There is a very educational article of Steve Bench on diode biasing:
http://greygum.net/sbench/sbench/dio_bias.html
There is no magic in that. I'm not sure why you think the diodes are so special about this design and constitute it's core and say the "diodes coming and going, to maintain balance, isolation and stability ". Perhaps I'm wrong but a cathode resistor also gives a "balance and stability". This comes from the autobias per se, not from a particular way of implementing it. I've been simulating my circuit for quite some time now. It does seem to behave...on a screen at least 
If the line voltage rises, rises both driver and out currents. More driver current=more drop on the choke DCR=more bias on the out counteracting the current rise. Simultaneously, the autobias provides a second mechanism of autobalancing in the usual way, but less efficient, since only small percentage of the bias is generated there.
What do you mean by "isolation" and why biasing diodes provide it I'm perhaps too stupid to see.
The mysterious thing I refuse to understand is whay the've put a second 6CG3 in the plate?
Then the power supplies are nothing special at all: tube rectified, generic 5U4 crap, (pseudo)-choke input, big, i.e. not 10uF, but still nothing compared to Romy, caps, reasonable volts. First choke of course has to "swallow" quite some volts, but that's normal, it's a choke-loaded supply (actually the small cap makes the job easier). What is different here from a 1000 other properly executed choke-input supplies?
Mind you I've managed to construct a 872A choke-loaded bridge supplying 1.5kV for 304tl (I'm an idiot, I know), so this looks like a toy :-) I've designed the tranny and the chokes myself. They worked. What for? To satisfy my stupid wet dreams...
I don't follow your worries of tank circuits generating lots of volts and noise. Power supply ones?
Then the tranny. Paul, what is so terryfying about it? They put all the 3 supplies on one core so it must be big if done right, but nothing abnormal. Then lots of tubes floating require lots of separate heater windings, so the heating tranny is big too, but again nothing abnormal. For my winders 100Eu is a lot. They will wind quite complicated power if I provide clear instructions. The are also old stocks of good cores, good cut core or toroid is enough, no amo crazyness, I don't want to listen to the power transformer anyway (Romy style PS, 10000uF). The careful routing (high volts) and esp. the grounding is a chalange, I agree. 1kV RF cable helps here. I have loads of it.
| Paul S wrote: | | As you obviously know, EL34 can be used in triode, and it has been my indirect understanding that its electrical parameters hold, as far as triode configuration makes use of them. I would not use the EL34 if I wanted specific sounds from the output tube, itself, but I do "recommend" it as a potentially "neutral" output tube with electrical advantages that can be electrically relied upon, if that is of use to you. "Sounds poorly" might describe the "sound of the EL-34" where it is made to have a "sound", or perhaps, even, its situational lack of a sound reveals undesirable else. In any case, I believe it can be "neutralized" by sticking to - or at least starting with and tweaking from - its optimum operating points, as listed in the manufacturer's tube manual. And, like I said in the early going, I would sooner stack tubes than stress them. Who knows, you might even re-think the going-in bias against pentode operation in this particular PP application, at least for output. |
|
Very encouraging comments on the neutrality and transparency! Thanks! I wanted to put it in the triode mode, at around 400V, 40mA. Remember it's a headphone amp I'm trying to cook not a speaker one, so I don't really need more volts, no need to stack the tubes. Max in triode mode is 425V (the screen limits here). Another candidate would be 300B, but it's stupid price and audiopedophile sound disqualifies it. Boutique tubes...let them power boutique amps, not mine.
Cheers,
N-set
Cheers,
Jarek
STACORE
|
|
|