You'll need a capacitor from the screen grid to ground.
Also, a 10K plate load resistor is a little low for the ECC83/12AX7
Also, a 10K plate load resistor is a little low for the ECC83/12AX7
Mr. Triatic said:
Hi Triatic !
The 832 is a dual tetrode where the cathodes are internally tied together as well as the screens.
It is difficult to use just an half because if you apply screen voltage without anode one, you'll burn out the screen !
The unused section could be "overbiased" (some -50V at its grid) to void that.
Moreover, ther is also an internal bypass cap between screens and cathode.
But data sheet says its value is some 65pF, forget it !
B+ is allowed to be as hi as 750V, but screens no more than 250V.
Rated anode dissipation is 15W per section.
It was designed to be used in PP, but both sections in parallel in an SE configuration should work.
As a (rough) starting point, I would suggest the following setting:
Va +400, Vg2 +250, IA 35mA (per section) this calls for some -25V grid bias (or a 350 ohms common cathode resistor for the 2 paralleled sections).
With your 9K tranny, this should produce some 7W PO in penthode mode.
In PP you could certainly obtain up to 20W PO.
As a stereo amp (that is one section per channel) it would be avantageous to increase the load up to 15K.
I agree than 12AX7's plate resistor should be 100 to 200K rather than 10K 🙂
Regard, Yves.
Yvesm said:
Oh dear, how could I have missed that! Yvesm is of course right, you'd probably have to go with parallel single-ended
ok, then i have to use two tubes in order to get stereo, and parallell the anodes on each tube due to the SE-mode...?!
the ECC83 can of course be single because i use one half to each channel, i have to raise the 10kohm to 100-200 kohm.
you dont think this construction would be prone to oscillate? i have a feeling that it wont, but?! -i guess i have to try it out 😎
i can arrange a B+ at 400V and filament supply from a transformer i have.. perfect!
the ECC83 can of course be single because i use one half to each channel, i have to raise the 10kohm to 100-200 kohm.
you dont think this construction would be prone to oscillate? i have a feeling that it wont, but?! -i guess i have to try it out 😎
i can arrange a B+ at 400V and filament supply from a transformer i have.. perfect!
Hi Triatic !
Not really.
It would be a very nice and cute to use one half of the 832 for each channel.
Just oversize the cathode and screen bypass caps, or even use Zeners ?
And at 35mA anode current, your tranny will be happier !
9K is not the optimum load, but they should do the job, specially if the loudspeakers are a bit above 4 Ohms !
Anyway, toobs are very tolerant to design "twists", and you should reach your " 2 to 3W " target easily.
This will give you some gain margin to apply feedback (no tetrode are highly linear).
Yes ! You MUST try ... and keep us informed !
Cheers, Yves.
Not really.
It would be a very nice and cute to use one half of the 832 for each channel.
Just oversize the cathode and screen bypass caps, or even use Zeners ?
And at 35mA anode current, your tranny will be happier !
9K is not the optimum load, but they should do the job, specially if the loudspeakers are a bit above 4 Ohms !
Anyway, toobs are very tolerant to design "twists", and you should reach your " 2 to 3W " target easily.
This will give you some gain margin to apply feedback (no tetrode are highly linear).
Yes ! You MUST try ... and keep us informed !
Cheers, Yves.
Hi Triatic,
You don't need to use two 832's for stereo provided that the cathode bypass capacitor is large enough to assure minimal cross-talk at the lowest frequency of interest.. There is precedent for this approach - just take a look at Bob Danielak's darling or dc darling to see what I mean. (Good sounding amplifier design incidentally.) IMHO I would ground the cathode and use fixed bias instead, with independent bias adjustment for each half of the tube... Make sure in either case that the screen is well bypassed right at the tube to prevent parasitic hf/vhf oscillation and a good low impedance supply seems warranted - perhaps using a gas tube as a regulator..
Kevin K
You don't need to use two 832's for stereo provided that the cathode bypass capacitor is large enough to assure minimal cross-talk at the lowest frequency of interest.. There is precedent for this approach - just take a look at Bob Danielak's darling or dc darling to see what I mean. (Good sounding amplifier design incidentally.) IMHO I would ground the cathode and use fixed bias instead, with independent bias adjustment for each half of the tube... Make sure in either case that the screen is well bypassed right at the tube to prevent parasitic hf/vhf oscillation and a good low impedance supply seems warranted - perhaps using a gas tube as a regulator..
Kevin K
9k ... probably best suited for EL84/6BQ5, EL95, 6V6, ELL80 (2 EL95s in 1 bottle), SV83, ECL86 (often called half a 12ax7 + EL84 in one bottle), ECL82/6BM8 (1/2 12AT7 + EL95 in one bottle), 6AQ5, Russian equivalents....
you need to consider how much current you can flow thru the OPT before it saturates as well.
dave
you need to consider how much current you can flow thru the OPT before it saturates as well.
dave
Hi there,
I don't think the screen supply design is going to bring out the best in your amp. Something regulated would probably work much better as suggested in Kevin K's post. A gas tube is easy to use and works well.
I don't think the screen supply design is going to bring out the best in your amp. Something regulated would probably work much better as suggested in Kevin K's post. A gas tube is easy to use and works well.
-> martinab2
tuberegulation crossed my mind, i have got one of those beautiful AZ4 tubes, unfortunately i miss a suitable socket..
http://w1.871.telia.com/~u87127077/diod/az4.htm
and one
http://w1.871.telia.com/~u87127077/diod/az1.htm
the 832A that i intend to use:
http://w1.871.telia.com/~u87127078/pentodl/832a.htm
what's the disadvantages of using a silicon fullwave rectifier other than sacrificing soundquality?
tuberegulation crossed my mind, i have got one of those beautiful AZ4 tubes, unfortunately i miss a suitable socket..
http://w1.871.telia.com/~u87127077/diod/az4.htm
and one
http://w1.871.telia.com/~u87127077/diod/az1.htm
the 832A that i intend to use:
http://w1.871.telia.com/~u87127078/pentodl/832a.htm
what's the disadvantages of using a silicon fullwave rectifier other than sacrificing soundquality?
Sorry, should have been clearer about the supply i was meaning. The regulation for the screen grid supply is important to how the amp will sound. At the moment the 4.7K resistor will allow the screen voltage to vary with the ht and any changes in screen current drawn. What I was suggesting is a gas regulator tube. These maintain a fixed voltage across a limited current range. In your application, it would be connected between the screen side of the resistor and ground.
This is the sort of gas regulator I had in mind http://hereford.ampr.org/cgi-bin/tube?tube=0D3
This is the sort of gas regulator I had in mind http://hereford.ampr.org/cgi-bin/tube?tube=0D3
I would recommend either an 0A2 or VR150 for screen regulation at about 150V..
Cathode stripping of the 832 is one possible consequence of using solid state rectifiers.. If you don't want to deal with an additional filament winding you could use damper diodes like the 6AU4 (single) or 6BY5 (dual) with 6V filaments. Their cathode insulation is designed to withstand about 400 - 500V without problems. An untapped high voltage secondary is no problem just use two semiconductor diodes for the other (negative) side of the bridge. Someone has named this configuration but I can't remember its name.. lol
Cathode stripping of the 832 is one possible consequence of using solid state rectifiers.. If you don't want to deal with an additional filament winding you could use damper diodes like the 6AU4 (single) or 6BY5 (dual) with 6V filaments. Their cathode insulation is designed to withstand about 400 - 500V without problems. An untapped high voltage secondary is no problem just use two semiconductor diodes for the other (negative) side of the bridge. Someone has named this configuration but I can't remember its name.. lol
Hi,
May I ask why?
If you mean a pair of semi-conductor diodes followed by a pair of hollow-state rectifiers in a Graetz bridge configuration then that's commonly refered to as a hybrid rectifier set-up.
Cheers, 😉
May I ask why?
If you mean a pair of semi-conductor diodes followed by a pair of hollow-state rectifiers in a Graetz bridge configuration then that's commonly refered to as a hybrid rectifier set-up.
Cheers, 😉
Yvesm said:
Triatic has no screen bypass cap at the moment.
The current draw of a tetrode/pentode is more dependent on the screen voltage than the plate voltage. A gas voltage regulator fed by a suitably sized resistor (or a solid state constant current source) is a common way to improve the performance of tetrode/pentode circuits as it keeps the screen voltage fixed (well, not exactly, but the variation will be small). Here is an example of an amp that uses this method of screen voltage regulation (among other things, it's an interesting design): Gary Pimm's PP47
The AZ4 you mention appears to be a thermionic rectifier, not a gas regulator. A gas regulator is the valve equivalent of a zener diode.
audiousername said:
AZ4? The poster probably meant to say 0Z4 which is indeed a cold cathode (no filament) gas rectifier tube in twin anode, common cathode configutation like the better known 5U4 and others. It was designed for use with vibrator power supplies in the old car radios and such, saving battery power by not needing a filament supply. It has a minimum and maximum output current range of 30-110 mA.
The AZ4 is a European-type DH full wave rectifier (4V filament, weird Euro-type base).
http://w1.871.telia.com/~u87127077/diod/az4.htm
I think the poster misinterpreted "tube regulated" as "tube rectified"
http://w1.871.telia.com/~u87127077/diod/az4.htm
I think the poster misinterpreted "tube regulated" as "tube rectified"
Fascinated !
Even if Gary describes a PP, some vy interresting ideas such as penthode driver DC coupled to a heavily "voltage feedbacked" output stage could certainly be used in an SE design.
The version with stacked supply is truly appealling !
However, in the present case and considering that Mr Triatic's project is single ended and thus pure Class A, regulating the screens is good but not a must as long as cathode auto bias is used.
The 4K7 resistor looks fine to me if a large (50 to 100µF) bypass cap is added.
Yves.
- Status
- Not open for further replies.