Hi everyone,
I'm looking for some thoughts and ideas from the collective minds here. I've been working on a 6SN7 tube preamp design for little bit now and have it performing very well using a HV bench supply.
Where I've found myself deliberating so much is on the PS design as I'd like a really quite preamp but trying to keep it somewhat simple at the same time.
First a few things about the preamp to set context, not that I'm doing anything too off the beaten path:
- uses 4 6SN7 tubes (2 per channel)
- has cathode follower input buffer
- then SRPP stage
- then output buffer stage
- need to float the heater voltage for SRPP stage (or all)
- need 250-300v B+ (would like to split rail)
One big factor is that I have quite a few 325-0-325v@70ma, 5v@3A, 6.3v CT@2.25A transformers at hand. While a bit high on the HV, it would be great to use since I have them. I don't mind burning off a little HV as heat
The transformers were leading me down a simple tube rectified path but I've also been pondering more complex solid state ideas such as UF style diodes, all poly cap design, buddha snubbers, etc. Guessing I'm really pondering the age old debate of SS vs Tube for this amp.
I'd love to get your thoughts and input and/or suggested designs/schematics you have had success with. Also, if I went with these transformers, any wisdom on dropping the voltage a bit other than traditional power resistors in the chain and where is best to place them in the PS design?
Thanks in advance for reading and your thoughts.
Mark
I'm looking for some thoughts and ideas from the collective minds here. I've been working on a 6SN7 tube preamp design for little bit now and have it performing very well using a HV bench supply.
Where I've found myself deliberating so much is on the PS design as I'd like a really quite preamp but trying to keep it somewhat simple at the same time.
First a few things about the preamp to set context, not that I'm doing anything too off the beaten path:
- uses 4 6SN7 tubes (2 per channel)
- has cathode follower input buffer
- then SRPP stage
- then output buffer stage
- need to float the heater voltage for SRPP stage (or all)
- need 250-300v B+ (would like to split rail)
One big factor is that I have quite a few 325-0-325v@70ma, 5v@3A, 6.3v CT@2.25A transformers at hand. While a bit high on the HV, it would be great to use since I have them. I don't mind burning off a little HV as heat
The transformers were leading me down a simple tube rectified path but I've also been pondering more complex solid state ideas such as UF style diodes, all poly cap design, buddha snubbers, etc. Guessing I'm really pondering the age old debate of SS vs Tube for this amp.
I'd love to get your thoughts and input and/or suggested designs/schematics you have had success with. Also, if I went with these transformers, any wisdom on dropping the voltage a bit other than traditional power resistors in the chain and where is best to place them in the PS design?
Thanks in advance for reading and your thoughts.
Mark
The 6SN7 has a -100V+200V HK limit, no need to elevate heaters.
If you use a choke input instead of a cap input your voltage should be good to go using the transformer you have.
I think you can put the choke in between centre tap and ground to use one choke for both rails. Correct me if I'm wrong.
Personally I use a 12V SMPS to power heaters and a HV boost converter for the B+.
If you use a choke input instead of a cap input your voltage should be good to go using the transformer you have.
I think you can put the choke in between centre tap and ground to use one choke for both rails. Correct me if I'm wrong.
Personally I use a 12V SMPS to power heaters and a HV boost converter for the B+.
With the transformers you have on hand the extra voltage won't hurt at all, and if anything will increase gain and linearity. Choke input isn't my style but it should be pretty easy to knock something up in PSUD2 using easily available parts. Use a fancy tube rectifier if you must, at such currents their downsides are negligible. Otherwise I would look to a simple Mosfet ripple filter, with a decent size cap from gate to ground to give you a nice soft start. UF4007 diodes for rectification, nice and cheap, while still providing formidable performance rivaling some of the more costly options.
Personally, I would elevate the heaters, just to be safe. It'll give you a little better noise rejection on the lower cathode of your SRPP stage anyway, and only costs a few resistors and a cap to make the reference (you need a bleeder in your PSU anyway, so it can to double duty as one) and then a couple across the heater winding. Easy peasy.
I try not to nitpick or get too critical of other people's designs, but I find an SRPP inside a buffer sandwich a bit odd for a linestage. I'm not a big fan of SRPP (due to their want of a specific load to perform their best) but it shouldn't really need a buffer on either side for most same loads. But, I'm not trying to derail your thread! Do your thing man. Post some diagrams and pics. Always neat to see different builds.
Personally, I would elevate the heaters, just to be safe. It'll give you a little better noise rejection on the lower cathode of your SRPP stage anyway, and only costs a few resistors and a cap to make the reference (you need a bleeder in your PSU anyway, so it can to double duty as one) and then a couple across the heater winding. Easy peasy.
I try not to nitpick or get too critical of other people's designs, but I find an SRPP inside a buffer sandwich a bit odd for a linestage. I'm not a big fan of SRPP (due to their want of a specific load to perform their best) but it shouldn't really need a buffer on either side for most same loads. But, I'm not trying to derail your thread! Do your thing man. Post some diagrams and pics. Always neat to see different builds.
Last edited:
SRPP is the best topology IMHO, it sounds really transparent, I use to use about 350V with an ECC82, do not forget that distortion decreases with an increase of +B.
Elevating the heaters to about +B/4 is always a good habit.
Poor PSRR demands a very clean +B, I always use a regulated PSU.
Now I use shunt regulators, but mine is too complicated, Salas HV shunt regulator should be OK.
A long ago I did use this
Valve rectification with EZ80, D3 and D4 are only for the simulation, I did use a noisy LM385 1.2 (reverse the polarity!), almost any power MOSFET should be OK, I did use STP6N60 from TV PSUs.
R9 is a trimpot of about 1Kohm.
This circuit was intended to replace a TL783C, unobtainable here at that time.
All Op Amp sucks, so use your favourite one.
A long ago I did use this
Valve rectification with EZ80, D3 and D4 are only for the simulation, I did use a noisy LM385 1.2 (reverse the polarity!), almost any power MOSFET should be OK, I did use STP6N60 from TV PSUs.
R9 is a trimpot of about 1Kohm.
This circuit was intended to replace a TL783C, unobtainable here at that time.
All Op Amp sucks, so use your favourite one.
Last edited:
As far as I know, unbypassed SRPP has a PSRR of about 6dB.
http://www.diyaudio.com/forums/tubes-valves/290993-ecc82-12au7-line-preamp-7.html#post4740824
http://www.diyaudio.com/forums/tubes-valves/290993-ecc82-12au7-line-preamp-7.html#post4740824
A totem pole, mu follower, or a beta follower (see Morgan Jones for that one) give lower distortion into most of the loads you see in audio when compared to the SRPP, and in my own testing, give much more favorable harmonics than SRPP when set up with the same tubes, and operating points. I'm not overly fond of SRPP with the better alternatives that are just as elegant and simple. People love SRPP either way, and just plain like using it. Nothing wrong there.
(Climbing down off my soapbox)
I really like lifting the supply for the heater for these circuits, it can often help with noise, and any free lunch method of noise reduction is fine by me. Less heater-to-cathode stress also gives me a warm and fuzzy too.
For those that aren't allergic to solid state, a mosfet ripple filter would be perfect for the supply voltage. Since the whole thing will not vary in current draw that much it would be a good place to use a stack of zeners or VR tubes as a gate reference, and a small heatsink will be more than adequate to allow it to eat some voltage. Run the reference voltage to the gate via a largish resistor, and run a cap from gate to ground for soft start. It will be stabilized, but open-loop so not truly regulated, but for a circuit like this it's still massively overkill, and will be dead silent.
Hell, even a pair of 0D3 VR tubes per channel on their own would probably work well, depending on total draw. Feed them with a nice current source, and it should be plenty quiet.
Last edited:
People love SRPP because it measures good and it sounds transparent, see some measurements here
http://www.diyaudio.com/forums/tubes-valves/74903-survey-aikido-distortion-18.html#post4721214
Nothing fancy, just an ECC82/12AU7 optimized SRPP
In attach a REAL mesurements of a SRPP of EL84 in triode connection ( for an article on Audio about this tube in different configuration of preamp and amp).
I like this configuration.
Walter
I like this configuration.
Walter
Attachments
If you think a triode connected EL84 rocks, try the Russian 6P15P triode connected. It'll give the 4P1L a run for it's money!
I agree that the ECC82 does nicely at higher voltages, but I always had issues with high dissipation to hit the distortion measurements I would see out of an equivalent 6SN7GTB in the same circuits, but up at such high voltages you find tubes that can't jive pretty quickly, even with rugged NOS stuff.
I agree that the ECC82 does nicely at higher voltages, but I always had issues with high dissipation to hit the distortion measurements I would see out of an equivalent 6SN7GTB in the same circuits, but up at such high voltages you find tubes that can't jive pretty quickly, even with rugged NOS stuff.
I know the 4P1L by Ale Moglia's blog, first place on the benchmark, not a clue about the 6P15P.
The valves obtainable here are from better times, a long time, so only a few Russian ones.
EL84 is just a curious case, not to uncork champagne.
I know only a few dual triodes that you can run at 650V in SRPP, ECC82 is one of them (Vkf=180V)
I was very distrustful about the ECC82, its curves are horfull, but in SRPP the sound is unbelievable.
As a TV repairman I am usually close to 25KV from CRT (it does hurt!), so someday I will make the damn 650V PSU, it will be a bit difficult to make it as a shunt regulator though.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.