New RSA amp...

[shaizada]

OK billy its worth $5.

Which part is wrong.

paypal immediately.

Hey i might be surprized.

Cut/paste from official A-10 thread:

The A-10, Electrostatic Amp/Pre-amp has a fully balanced input to output circuitry, with two balanced inputs, XLRs, and three SE, RCAs

The input/gain stage incorporates two 12AX7 tubes that handle a balanced input signals via four section balanced DAC attenuator.

In the Pre-amp mode, the signal is passed to two 6SN7 tubes, via relays, which are used as pre-amp output stage ONLY.

If electrostatic phones has been selected, the input of the two 6SN7 is grounded to minimize noise and the signal is routed to four 5687 tubes,

via relays, which are used as drivers for the phones.

When single ended signal is used, it is routed, via relays, to a single 12AX7, center tube. This tube is used as phase splitter with a gain of two for both channels.

Then the signal is routed to the balanced attenuator which passes it to the two 12AX7 in the gain stage.

Because of this design the pre-amp XLR output, will always be in balanced even though the SE input source has been used.

Also the SE outputs will be available at all time.

The A-10 has a balanced and a SE tape loop. The selected source input will be present at these outputs with out pre-amplifications to the signals.

More to come….

END

Looks like KG owes Billy some money and coffee. Starbucks sucks though (having owned a few coffee shops and knowing a little about coffee beans; they over roast their beans! All dark).

Dr. KG, I still have mucho respect for your tube circuit knowledge, but mistakes can be made by the best of us

[grawk]

Cut/paste from official A-10 thread:

In the Pre-amp mode, the signal is passed to two 6SN7 tubes, via relays, which are used as pre-amp output stage ONLY.

If electrostatic phones has been selected, the input of the two 6SN7 is grounded to minimize noise and the signal is routed to four 5687 tubes,

via relays, which are used as drivers for the phones.

...

Also the SE outputs will be available at all time.

Those two statements contradict each other.

[mirumu]

Ah so it's really three different configurations with not all tubes active at any one time. That certainly explains why there's so many of them. Seems a bit wasteful if it was only ever to be used an electrostatic amp though.

[mirumu]

Those two statements contradict each other.

I read it as meaning the 6SN7s are only ever active for the pre-amp output, and the 5687s are only ever active for the electrostatic outputs, but the SE output doesn't use these and is always active. It's not stated explicitly, but it sounds like SE output comes off the circuit just after the pair of gain 12AX7s.

[spritzer]

This is what happens when you try to apply logic to the RSA crap. Here we were thinking that Ray might have just a hint of sense and used the high voltage version of the 6SN7 but he used a 300v max tube instead. We bitch about the 6CG7/6FQ7's in the Stax amps but at least they can handle 330v.

[kevin gilmore]

I was thinking that it was possible ray was using the 5687's as the output stage

to drive the electrostatics. But depending on which tube data sheet you read, its

good for either 330 or 350 volts max. And since there are 4 of them, the output

stage is srpp. Certainly not something i would want to buy. At least it is going

to sound very different from everything else out there.

Gotta go.

[justin]

An electrostatic amp with 12AX7 front-end (5751) and 5687 outputs has been shown on Tubecad before. KG, i would assume its paralleled sections of 5687 like the Stax 007t before assuming it's SRPP. I looked into using SRPP for the output stage of an electrostatic amp, but you have to do some tricks to set it up for a capacitive load and at that point it doesn't make sense to use it...lots of info on that on tubecad as well

[bhd812]

i love those!

I really like this place cause you peeps understand me either know you can't make any freaking sense from my words i type..that's coming from the heart!

Hey KG i usually don't go back on a deal involving money but instead of buying me a coffee can you PLEASEEEEEE Donate the $5 to the Kidney Cancer Association In Evanston (by you!)...if you want to do it then I will match your donation with $5 of my money!

it's up to you since i am changing my end..

Kidney Cancer Association - Donate Now

[kevin gilmore]

OK, so lets look at the 5687 datasheet

http://www.drtube.com/datasheets/5687-ge1959.pdf

Notice the 90 volt max cathode to filament, and the 300 volts max plate voltage.

Now lets look at all possible configurations of 4 x 5687 as electrostatic outputs.

1) Both sections of the tube in parallel with resistor plate load. Max Vcc 300 volts.

Max power 8 watts per tube. Same power as BHSE. No voltage violations.

And probably not ray's style. Capacitor output, or dual power supplies with

filament referenced to ground or V-.

2) Both sections of the tube in series, SRPP. Clearly ray's style from cRaptor and

B52. Capacitor output, or dual power supplies. In either case there is the cathode

to filament voltage violation, same as ES1, except only double the ratings instead

of 4x.

At least no one will accuse ray of swiping another's design as this is the only

electrostatic amp i know of with 5687's as output tubes. OOPS, justin proves me

wrong again...

With a couple more switches, ray can use the thing as a real B52 replacement

i.e. it could also drive regular headphones.

Open loop with absolutely no feedback. If you like the rest of ray's stuff you are

gonna love this thing.

[kevin gilmore]

Hey KG i usually don't go back on a deal involving money but instead of buying me a coffee can you PLEASEEEEEE Donate the $5 to the Kidney Cancer Association In Evanston (by you!)...if you want to do it then I will match your donation with $5 of my money!

it's up to you since i am changing my end..

done billy. proof in your PM. I doubled it. I expect you to do the same.

[kevin gilmore]

Well if that picture is what ray did, what is the possibility he got the

dual filament supplies thing right...

And running a 5687 on 600 volts... Double the specifications...

Thinking about it, justin is probably right, and ray is using the above design.

However since there is only one tube in the middle, that is a standard cathodyne

phase splitter. From ray's description.

[morphsci]

Those two statements contradict each other.

No they don't. He simply takes the input for the tubes and grounds that. The signal does not go through the 6sN7 tube circuit.

This is what happens when you try to apply logic to the RSA crap. Here we were thinking that Ray might have just a hint of sense and used the high voltage version of the 6SN7 but he used a 300v max tube instead. We bitch about the 6CG7/6FQ7's in the Stax amps but at least they can handle 330v.

I love the rampant speculation in this thread. What other headphone site does that remind me of.

[grawk]

The 6sn7s are for the single ended outputs. The 6sn7s are fed ground instead of input if you're using hte headphones. The single ended outputs are always active.

There's a contradiction there.

[en480c4]

I love the rampant speculation in this thread. What other headphone site does that remind me of.

:yawn:

[spritzer]

:yawn:

Couldn't have said it better myself.

[n_maher]

I wish there was a little less immediate dismissal until we get some interior shots. Do I think it'll be god's gift to stats, no, but I don't feel comfortable with the proclamations that it's 100% fail either.

[jinp6301]

I love the rampant speculation in this thread. What other headphone site does that remind me of.

x2 Until we see some internal shots, all we're doing is guessing

EDIT: Nate said it best.

and someone needs to make an all in one e-stat and dynamic amp.

[justin]

x2 Until we see some internal shots, all we're doing is guessing

EDIT: Nate said it best.

and someone needs to make an all in one e-stat and dynamic amp.

not hard, but to make a good one, you're just putting 2 amps and 2 power supplies in 1 big box. and it would be expensive and would never be taken seriously as a reference quality amp for either electrostats or dynamics

[morphsci]

:yawn:

I guess my expectations are too high.

Couldn't have said it better myself.

Not surprising.

[morphsci]

The 6sn7s are for the single ended outputs. The 6sn7s are fed ground instead of input if you're using hte headphones. The single ended outputs are always active.

There's a contradiction there.

Well I guess I see it as a poor explanation not a contradiction. But I see what you mean.

[mypasswordis]

and someone needs to make an all in one e-stat and dynamic amp.

Isn't the A10 close, with its preamp section?

I'm also waiting for internal pics before I make any final judgments but based on the capabilities of the tubes used for the electrostatic section it is obvious that it's not going to be the most powerful electrostatic headphone amp designed. Of course that title would go to the BHSE.

[Dusty Chalk]

This thread officially started becoming awesome when Billy chipped in.

lol, it really did.

I'm guessing you guys that like Billy maybe read Head-Fi or something? Because over here, he's only ever an asshat.

[grawk]

or we like the way he's an asshat

[Elephas]

Didn't some people hear a prototype A-10 already? I vaguely remember reading about some people going to Ray's place after a meet and using a Jade with an electrostatic amp.

[nikongod]

Isn't the A10 close, with its preamp section?

the output impedance will be pretty high (about 150 ohms at best, could easily be higher/worse if ray did the "right" things incorectly although he has experience from a similar output stage on the B52... why not use 5687 here too? ) and unless they are electrolytics, the output caps are probably too small.

it may work for 2000 ohm headphones dont waste your time for anything else. increasing the size of the caps could make it do good things to beyers, which like higher output impedances.

im curious if ray thought far enough ahead to use the preamp buffers (6sn7) to isolate the 12ax7 from the miller capacitances of the output tubes... bandwidth fail.