SV811-3 SE : Lots of questions...

[Paracelsus]

I’m trying to build myself my first tube amp. So far I’ve decided on a SE with SV811-3, with 2 stages, and as little (none?) feedback as possible.

However, there are a number of things that aren’t completely clear to me, and I would appreciate your views on a couple of subjects:

How critical is the PSU transformer? What concerns should I have if I have it made in a local electronics store?

What is the real advantage of having separate power supplies over one larger common power supply?

It seems that using fixed bias (besides giving a little more power) produces a different sounding amp. Is this true? What are the differences?

I’m considering the SV 811-3 for power tubes, but it appears that since it is a transmission triode, I have to take into account the grid current. How much current are we talking about and in which conditions can I expect it?

The Svetlana datasheet for this tube recommends -165V grid bias + 30Vpp signal, but the curves only go to -160V. Isn’t that a bit odd? How much bias can a tube handle?

I initially thought about the 6n1p for the gain/driver stage, but it seems that it doesn’t have enough juice. What are good alternatives?

I read somewhere that the gain stage would benefit from using pentodes instead of triodes. Is this true? Why?

I thought about using an active load configuration on the gain stage. Would it be an advantage?

Is there any disadvantage to dc coupling?


Thanks for the time, and sorry about all these questions...
-Miguel

[Fuling]

Mains transformer: One important thing is that it´s properly insulated. 811-3 is a high voltage tube and you don´t want any insulation problems in the transformers (also applys to the OPT´s)

Separate PSU´s: Better channel separation. Not absolytely necessary, many good amps uses a single supply.

Grid current: You can push 811-3 into class A2 (where it draws grid current), but it´s definitely not necessary. Don´t mix it up with 811A which was made for A2. (Ok, it was made fro transmitter usage)

Bias: Those Svetlana curves are a bit odd, yes, since they only show half of what a designer needs to know. Can´t give you much advice here.

Driver stage: Can´t point you towards any specific tube, but I would definitely consider IT or LC-coupling. Getting this kind of voltage swing from a RC-loaded driver stage might not be easy.
A friend uses 6SN7 (both halves in // and a 15k plate resistor)
but he doesn´t have much margins.

Active loading might be a good alternative to plate chokes and IT´s. You will have plenty of voltage to play with if you use the same HV supply as the power tube.

DC-coupling: Yes. You´ll need one hell of a power supply!

[Eli Duttman]

Here's a URL for a mu follower that's intended to drive a SV572-3 (http://www.svetlana.com/docs/TechBul...noteNO.27.html). The SV572-3 shares filament and grid construction with the SV811-3. The difference between the 2 types is in the anode. The SV572-3 has a graphite, instead of a metal anode, which allows higher plate dissipation. If the mu follower can drive a SV572-3 it can drive a SV811-3.

Cathode bias for such a large negative grid voltage seems impractical, due to I^2R heating losses in the cathode resistor. IMO, Dennis Boyle's combination bias is a better choice. Connect the wiper of the hum balance pot. to ground via an unbypassed 100 Ohm resistor. Build a bias supply from a "120" VAC winding and a bridge rectifier of 4 UF4007 diodes. A 100 muF. 250 WVDC 'lytic is all the filtering needed. Measure the voltage drop across the 100 Ohm cathode resistor as you adjust the bias pot. (10 KOhms). Use Ohm's Law to calculate the voltage drop needed for the idle current you want.

[Eli Duttman]

Sorry, I was wrong about using a 10 KOhm bias pot. At the high bias voltage needed, a 100 KOhm pot is in order to hold the current in the voltage divider under 2 mA. Increase the bias supply filter cap. to 470 muF.

The bias adjustment pot. will do double duty. Idle current will be controlled and the part will function as the SV811-3's grid leak resistor.

[DougL]

I read somewhere that the gain stage would benefit from using pentodes instead of triodes. Is this true? Why?

Pentodes swing closer to the supply rail than a triode. This can be an advantage if you require large drive voltages.

The down side is usually higher distortion than comprable triode.
It requires careful balance of goals. High voltage swing at low distortion is a non-trival design challange any way you cut it.

Cheers;

Doug

[SHiFTY]

I would go for a transformer that is well over rated for what you need, perhaps 50% over your VA requirements. Better regulation and cooler running.

Fixed bias i don't know much about, cathode (autobias) is what I always go for, easy, safe and sounds good.

Seperate PSUs, well they can be pretty good. If possible, get a second, lower B+ winding on your transformer, avoids the need for yet another box.

You can expect grid current under overdrive conditions. With a Transmitting triode, you could use a cathode follower to drive it, they can provide a bit more grunt.

Direct coupling: no advantage in the power stage. Possibly useful in the preamp stage, although unsure. The main advantage is to get rid of a coupling capacitor, which IMHO isn't too much of a worry.

If I were you, I would use 1/2 a 6SN7 into a 6L6 or 6V6 cathode follower into the transmitting tube. Something like that would have a lot of grunt... Just a suggestion.

[AKSA]

If you are really serious about using the 811-3, then I suggest you pay Bob Danielak's webpage a visit here:

Bob Danielak's Website

This is, without doubt, one of the most experienced and least prejudiced design centres for this and many other legendary tubes. Bob is a very nice guy, will help you if your questions are considered and polite, and is a very creative engineer.

If I were driving a 811-3 (I have two, but am loathe to use them as I've not been able to get replacements, so am now concentrating on a 6C33C-B SET) I would use the 6BX7. These are high voltage dual triodes, with mu of 10 and rp of 1K3; highly linear and very suited to distortion cancellation driving the 811.

Cheers,

Hugh

[Paracelsus]

AKSA

I have two, but am loathe to use them as I've not been able to get replacements

What do you mean?
I thought that the sv811-3 was still being produced and was easily (and cheaply) available. That's one of the reasons I'm considering it... am I wrong?

[AKSA]

Hi Para,

I've been looking on and off for about six months; I have two beautiful Svetlanas, lovely tubes, but if you can find SV811-3s for me I'll be a happy man, buy two and move ahead!

I'd rather build an 811 SET than a 6C33C SET; more expensive, but better sound by far I'd reckon........

Cheers,

Hugh

[fdegrove]

Hi,

You could try this source:

SV811-3

Cheers,