Hello,
I would like to build my first pp; a 6as7 circuit to suit the tubes and opts I already have (opts are from vintage Philips el84 parallel PP recording studio amps, 4k to 8ohms). I have been looking around for schematics and found a few candidates but each poses a problem to my limited technical know-how:
http://www.diyparadiso.com/sk6as7pp1.htm
--> looks promising; my plan would be to leave out the PS and desing one using PSUDII to suit my old stock power trafo; is the part of the circuit on the top right meant to delay B+ for the 6as7's? Additional problem would be to achieve the high 550V B- (my trafo is about 375V@ 200mA, I never designed a PS with B+ and B- rails) + the circuit is not so simple for a beginner like me
http://www.triodeguy.com/6as7_pp.htm
--> Biggest downside is the class A operation (to suit the toroids I guess); I would prefer the 10watts output of class A/B. However, amending this circuit with simple class a/b output stage cathode bias would, if the preceding stages can drive it, seem like an easy 'beginners'-circuit. I'ld also design a different PS with the additional filtering these amendments would necessitate
http://www.dissident-audio.com/PP_6AS7/Schema.gif
--> I can't identify the phase-splitter used, looks weird to me + the output stage and PS have a few elements I don't understand (the connection starting from the power trafo's centre tap)..
I would be very happy if some of the members with more technical luggage could comment on these circuits or do some propositions; for now the second schematic with new output stage and PS looks like the best option...
Thanks a lot!!
Simon
I would like to build my first pp; a 6as7 circuit to suit the tubes and opts I already have (opts are from vintage Philips el84 parallel PP recording studio amps, 4k to 8ohms). I have been looking around for schematics and found a few candidates but each poses a problem to my limited technical know-how:
http://www.diyparadiso.com/sk6as7pp1.htm
--> looks promising; my plan would be to leave out the PS and desing one using PSUDII to suit my old stock power trafo; is the part of the circuit on the top right meant to delay B+ for the 6as7's? Additional problem would be to achieve the high 550V B- (my trafo is about 375V@ 200mA, I never designed a PS with B+ and B- rails) + the circuit is not so simple for a beginner like me
http://www.triodeguy.com/6as7_pp.htm
--> Biggest downside is the class A operation (to suit the toroids I guess); I would prefer the 10watts output of class A/B. However, amending this circuit with simple class a/b output stage cathode bias would, if the preceding stages can drive it, seem like an easy 'beginners'-circuit. I'ld also design a different PS with the additional filtering these amendments would necessitate
http://www.dissident-audio.com/PP_6AS7/Schema.gif
--> I can't identify the phase-splitter used, looks weird to me + the output stage and PS have a few elements I don't understand (the connection starting from the power trafo's centre tap)..
I would be very happy if some of the members with more technical luggage could comment on these circuits or do some propositions; for now the second schematic with new output stage and PS looks like the best option...
Thanks a lot!!
Simon
Hi,
I can at least tell something about the last design in your list.
The phase splitter, even if curiously drawn, is nothing else than a paraphase built around two SRPP.
The connection at the center tap of the PT provides a +200V B+ for the power output stage while the driver is powered by +400V.
A must for driving the lazy 6AS7 with an SRPP based driver.
Yves.
I can at least tell something about the last design in your list.
The phase splitter, even if curiously drawn, is nothing else than a paraphase built around two SRPP.
The connection at the center tap of the PT provides a +200V B+ for the power output stage while the driver is powered by +400V.
A must for driving the lazy 6AS7 with an SRPP based driver.
Yves.
Hello Yves,
Can you say a few words about the performance of the amp; e.g. how does it compare to the other PPs on your site / are you still happy with the design even after a few years of probably increasing know-how?
PS: When using a transformer without centre tap, could the "0"-point created by the "backside" of the diode bridge be used as centre tap?
I find the design very attractive but am more inclined towards LTP with ccs in the tail as I've read so much good things about this topology; unless ofcourse paraphase SRPP is an overachieving underdog.
Mille mercis!
Simon
Can you say a few words about the performance of the amp; e.g. how does it compare to the other PPs on your site / are you still happy with the design even after a few years of probably increasing know-how?
PS: When using a transformer without centre tap, could the "0"-point created by the "backside" of the diode bridge be used as centre tap?
I find the design very attractive but am more inclined towards LTP with ccs in the tail as I've read so much good things about this topology; unless ofcourse paraphase SRPP is an overachieving underdog.
Mille mercis!
Simon
Hi Simon,
This was my very first design 8 years ago for my "tube come back".
It's a very good sounding amp, delivering some 8 Watts per channel at vy low THD from 15Hz to 50Khz, thanks to he OPTs I've bought from a French winder (ACEA near Toulouse, I can find exact ccordinates if you need).
You cannot go without a true center tapped PST or you must design another power supply able to deliver 150mA at 200V per each channel (for the power stage) plus some 400V, 5 to 10mA for the phase splitter/driver.
(This is probably the key design of this amp: use low voltage for the 6AS7 because that's what they like, but high for the driver because that's what they need, voltage gain begin around 1.5 !)
LTP PI won't give gain enough and will force you to use an additional gain stage with the innevitable link caps and so on.
Could be done of course but it will be a completly different design while I've found the paraphase SRPP perfectly matched the 6AS7 drive requirements.
Indeed, the design philosophy was highly influenced by years of industrial job where you should obtain the best possible results using the simplest, the cheapest and the lowest number of components.
For exemple, no cathode by pass caps ! !
Not an high efficiency design anyway due to the huge appetite of the 6AS7 (at least for heather) and the very simple (but safe) cathode bias.
But who cares about efficiency should not build tubed amplifiers !
Try and listen, that's all I can add.
Yves.
This was my very first design 8 years ago for my "tube come back".
It's a very good sounding amp, delivering some 8 Watts per channel at vy low THD from 15Hz to 50Khz, thanks to he OPTs I've bought from a French winder (ACEA near Toulouse, I can find exact ccordinates if you need).
You cannot go without a true center tapped PST or you must design another power supply able to deliver 150mA at 200V per each channel (for the power stage) plus some 400V, 5 to 10mA for the phase splitter/driver.
(This is probably the key design of this amp: use low voltage for the 6AS7 because that's what they like, but high for the driver because that's what they need, voltage gain begin around 1.5 !)
LTP PI won't give gain enough and will force you to use an additional gain stage with the innevitable link caps and so on.
Could be done of course but it will be a completly different design while I've found the paraphase SRPP perfectly matched the 6AS7 drive requirements.
Indeed, the design philosophy was highly influenced by years of industrial job where you should obtain the best possible results using the simplest, the cheapest and the lowest number of components.
For exemple, no cathode by pass caps ! !
Not an high efficiency design anyway due to the huge appetite of the 6AS7 (at least for heather) and the very simple (but safe) cathode bias.
But who cares about efficiency should not build tubed amplifiers !
Try and listen, that's all I can add.
Yves.
I would go with the triodeguy amp. I have built something very similar but with interstage transformers. It will work with the output transformers you have. It address the key issue with the 6AS7 - which is current balance.
The only thing I don't like is the operating point as I think these tubes will perform better at lower voltage and higher current.
My experience with these tubes is that they sound sweet and powerful if driven hard.
Shoog
The only thing I don't like is the operating point as I think these tubes will perform better at lower voltage and higher current.
My experience with these tubes is that they sound sweet and powerful if driven hard.
Shoog
Thanks for the advice Shoog. I was planning to take that route and copy the whole signal circuitry including current balanced output stage. Would you know by any chance how I can calculate the necessary value of the cap between the 6as7 cathodes? I'll have to recalculate to suit the primary inductance of my opts
Cheers -- Simon
Cheers -- Simon
My technical knowledge is sadly limited. I'am a suck it and see man myself.
In my case I had 1000uf cap bypassed by 10uf caps which where originally referenced to ground. At Brian Becks advise I tied their "-" together and applied a 1mg resistor from the tie to ground to turn them fully differential. This will give an effective cross coupling of 500uf. This seems adequate, but on further thought the bass isn't quite as strong as it was originally. This could be because I have effectively halved the bypassing cap. It wouldn't be to difficult to add another 1000uf to each cathode to bring it up to 2000uf per cathode (though its getting very tight down there) to see if it brought the bass back up a little.
Because the output impedance is so very low with the 6AS7 the cathode bypassing has to be much larger than with more usual output valves.
Thats a bit rambling - but I hope it gives a few pointers.
Shoog
edit: I just looked at his schematic again. I can't quite believe that he is only using 10uf between the cathodes - but thats what he seems to be doing. My understanding is either very poor or he has compromised his design. Unfortunately its probably the former. If I knew the theory behind how to calculate that cap then I would be able to offer a more useful answer.
In my case I had 1000uf cap bypassed by 10uf caps which where originally referenced to ground. At Brian Becks advise I tied their "-" together and applied a 1mg resistor from the tie to ground to turn them fully differential. This will give an effective cross coupling of 500uf. This seems adequate, but on further thought the bass isn't quite as strong as it was originally. This could be because I have effectively halved the bypassing cap. It wouldn't be to difficult to add another 1000uf to each cathode to bring it up to 2000uf per cathode (though its getting very tight down there) to see if it brought the bass back up a little.
Because the output impedance is so very low with the 6AS7 the cathode bypassing has to be much larger than with more usual output valves.
Thats a bit rambling - but I hope it gives a few pointers.
Shoog
edit: I just looked at his schematic again. I can't quite believe that he is only using 10uf between the cathodes - but thats what he seems to be doing. My understanding is either very poor or he has compromised his design. Unfortunately its probably the former. If I knew the theory behind how to calculate that cap then I would be able to offer a more useful answer.
Hi, I'm also interested in the 6AS7, but driven by a CCS LTP 12AT7 with your 400V, with a regular grounded-cathode 12AT7 input stage. All the gain anyone could need, I hope. My question is this: With your power supply, aren't you unbalancing the power transformer by only using half a winding to feed the 150mA to each 6AS7 pair? Would a transformer with completely separate 200V windings work better? Maybe I'm just reading the schematic wrong...Yvesm said:
One more general question - is there any preference between 6080's and 6AS7's? Or the Russian 6N5S or 6N13S?
I just had a discussion with Brian Beck about that cap value. You may be interested to read the whole of the "vacuum State dpa300b" thread, if you haven't been following it. Brian recons that Triodeguy is doing something wierd with resonances and that has dictated his choice of cap. Since you are using a different transformer that cap value is definately wrong. Brian agreed with me that he couldn't see any reason why this cap value would be different than for a regular cathode bypass. In that case I would suggest something in the range of 500uf to 2000uf. I am going to up mine from 1000uf to 2000uf. This is because in my setup the two cathode cap are in series and give an effective bypass of only 1000uf. Hope that helps.
I am currently using the russian 6H13's. Can't say as they sounded much different to the 6080's I tried. The only real difference I can think of is that the Russians look the nicest. I think the driver has the most influence over the sound of the amp. I might just try some NOS Mullard 6080's in the hope that they have that Mullard magic - but I doubt it.
Shoog
I am currently using the russian 6H13's. Can't say as they sounded much different to the 6080's I tried. The only real difference I can think of is that the Russians look the nicest. I think the driver has the most influence over the sound of the amp. I might just try some NOS Mullard 6080's in the hope that they have that Mullard magic - but I doubt it.
Shoog
sorenj07 said:
Just look at the power supply as being two full waves rectifiers on a 400V CT transformer, one positive, the other negative.
You obtain two rectified voltages : +200 and -200V relative to the center tap.
Now, remove the ground from the center tap and move it to the -200 output.
The result is now 0, +200V (from the formerly grounded center tap), +400V.
The transformer is not abused in any way.
(All numbers rounded !)
Yves.
Yes, there's a paragraph about this in the audioxpress article. I'm waiting for a reply from triodeguy; naturally I'ld prefer lower value cap (all the more because I've got many of those in stock). The huge difference in capacitance between what he uses and what you use is intriguing..
Simon
Hi Cotdt,
I remember your mail but forgot to answer it, sorry for that.
You can find the schematic here: http://www.t-linespeakers.org/tubes/triode-trick.html
the original text on joelist by Bill Perkins connects the 10ohms resistor to plate, not to the screen (and so did I) but this probably doesn't matter
Eddy Vaughn uses a similar trick in his Carina and gave some advice on how to improve the schematic by Bill Perkins, see his wise words on: http://www.hawthorneaudio.com/forum/viewtopic.php?t=866&highlight=perkins
http://www.hawthorneaudio.com/forum/viewtopic.php?t=302&postdays=0&postorder=asc&start=0
As my initial impressions throughout the web indicate: I'm very enthousiastic about the tweak (in spite of the fact that very few triode-strappers seem to have implemented it and there's much scepticism) . FWIW I use el84 (philips = mullard = valvo) with ecc40 as driver (previously used e80cc, both are improved versions of ecc82 and 5965 if you will) and added small amount of anode to anode feedback (RH84 style) using a 300k resistor. This does wonders for making the bass tighter and faster without any apparent drawbacks.
Chiao,
Simon
I remember your mail but forgot to answer it, sorry for that.
You can find the schematic here: http://www.t-linespeakers.org/tubes/triode-trick.html
the original text on joelist by Bill Perkins connects the 10ohms resistor to plate, not to the screen (and so did I) but this probably doesn't matter
Eddy Vaughn uses a similar trick in his Carina and gave some advice on how to improve the schematic by Bill Perkins, see his wise words on: http://www.hawthorneaudio.com/forum/viewtopic.php?t=866&highlight=perkins
http://www.hawthorneaudio.com/forum/viewtopic.php?t=302&postdays=0&postorder=asc&start=0
As my initial impressions throughout the web indicate: I'm very enthousiastic about the tweak (in spite of the fact that very few triode-strappers seem to have implemented it and there's much scepticism) . FWIW I use el84 (philips = mullard = valvo) with ecc40 as driver (previously used e80cc, both are improved versions of ecc82 and 5965 if you will) and added small amount of anode to anode feedback (RH84 style) using a 300k resistor. This does wonders for making the bass tighter and faster without any apparent drawbacks.
Chiao,
Simon
Interpreting the part of the article where triodeguy describes the ac-current balance circuit (= the capacitor), the formula must be easy. Variables are:
L (primary inductance opt)
C
Rplate
fhz (resonant frequency of the circuit)
The author says the resonant circuit formed by L and C causes a (in his case calculated) peak at 5hz "that extends upwards into the frequency region where the reactance of the capacitor would otherwise interact with the cathode to cause an LF rolloff"
I think the simple formula for calculating the resonance frequency of a given circuit is all that it takes to find the capacitor value. I'll aim for 5hz as triodeguy does. When thinking about it, it's the same principle as a parallel zobel filter. Just fill in the values for L, C and R in a online zobel calculator and pronto! (off to the cellar to measure the primary inductance of my opts)
Simon
L (primary inductance opt)
C
Rplate
fhz (resonant frequency of the circuit)
The author says the resonant circuit formed by L and C causes a (in his case calculated) peak at 5hz "that extends upwards into the frequency region where the reactance of the capacitor would otherwise interact with the cathode to cause an LF rolloff"
I think the simple formula for calculating the resonance frequency of a given circuit is all that it takes to find the capacitor value. I'll aim for 5hz as triodeguy does. When thinking about it, it's the same principle as a parallel zobel filter. Just fill in the values for L, C and R in a online zobel calculator and pronto! (off to the cellar to measure the primary inductance of my opts
)Simon
EDIT: read notch, not zobel
The notch calculators I can find do not allow input of L, C and R so it'll have to go the manual way.
Shoog,
You're probably right that a much higher value cap is necessary. The big difference between those toroids and iron-cored opts is the primary inductance; 100H for the toroids triodeguy used, 8H for Hammond 1645 and 5.5 / 6.7 for the old opts I plan to use.
Simon
The notch calculators I can find do not allow input of L, C and R so it'll have to go the manual way.
Shoog,
You're probably right that a much higher value cap is necessary. The big difference between those toroids and iron-cored opts is the primary inductance; 100H for the toroids triodeguy used, 8H for Hammond 1645 and 5.5 / 6.7 for the old opts I plan to use.
Simon
I just realized that this amp is a perfect candidate for those big industrial control transformers. I have one that buzzes a tiny bit when plugged in, but it's 550VA and has selectable 115V "secondaries" and 230V "primaries" Basically you can end up with two ~230V windings at around 1.2 amps. That's an extreme example but a smaller one would also work well. You'd need an auxiliary filament transformer but with the greedy 6AS7's you'd probably need one anyway
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.