Moskido - Hybrid Aikido/MOSFET amplifier

[tubee]

Originally posted by beau2317
I figure that the PS impedance is dominated by the last caps. Down side is it's wasteful, but class A amps are wasteful anyway.

Yes, my hybrid's fet section stays allways powered, even more wasteful....

[beau2317]

Somebody asked for details on components (eg power, voltage ratings). So here are some notes to go with the schematics on page 3 of this post. (post #25)

Main Amp Schematic:

-- All resistors, except the 0.2 ohm parts : 0.5W or more.
Note that the cathode resistors and the grid-stopper and gate-stopper resistors should be "high quality"

-- 0.2R resistors : 1W or more.

-- 0.1uF capacitors : 300V at least
-- Other capacitors, at least 65V (or whatever power rail you use for mosfets)

Power Supply

-- Capacitors need to be at least rated at the rail voltage, depending on where they are used.
-- 470ohm resistors : 2W at least. I used the white ceramic box shaped wirewounds.

-- transformers : High voltage, Hammond from Partsconnexion. Low voltage 500VA 50V RS Components

-- Rectifiers, FRED, 600V, 22A, Partsconnexion

Note : John has posted some more Moskido variations on his blog at www.tubecad.com

Please : if anyone actually makes a Moskido, please post the results and pics. Thanks. Bob.

[marsupialx]

I'm working on mine. I am going to have a Moskido (SE) and a First Watt (Nelson Pass) F4 (driven by the C2 output) in the same chassis. I'll be able to use it as an extra burner when my stove is full.
I have the Buz900 mosfets and the F4 board, acquiring parts as I can.
Here is my version of John's logo idea and your name (moskido).
I hope it shows up ok. Thanks, y'all. - Jeffrey Hill (marsupialx)

[beau2317]

Originally posted by marsupialx

Here is my version of John's logo idea and your name (moskido).

Cool, thanks. I just love JB's idea of the V8 with 8 rectifiers..

[Valverine]

Hi Bob (beau2317),

Do you think it would be physically possible (using John's board) to make a 2-stage SE Moskido by placing MOSFETS instead of the second tube stage? Granted, PS connections should be somehow fiddled around because of the voltage differences between the driver and the output and adding extra parts posted in the prev. schematics. I don't have the board (yet) so it is tough to know whether it is possible.

I believe I can drive the output from a single beefy driver, also because having Aikido in the preamp incarnation before Moskido will swing good voltage.

Thanks!

[beau2317]

Can you draw what you mean exactly. I am not clear on what you are planning.

btw, this is a logo that JB made and sent to me..

[Valverine]

Hi Bob,

This is what I have in mind – all the (important) details omitted, just to show a sketchy thought:

[beau2317]

Right so you want to basically omit the aikido-style cathode follower and drive the mosfets directly from the first stage of the aikido.

Firstly, voltage gain does not come into it. The stage you are omitting has pretty much unity gain so the gain would be the same.

Second, the idea of the cathode follower stage is to provide a low-impedance drive for the mosfets. And secondly to inject PSU noise to do the aikido noise canceling trick.

If you leave it out then I suspect that the mosfets would not be driven so nicely (they have quite high capacitance) and would roll off the highs.

And the design as you have drawn it simply will not work. In the SE design the bottom mosfet sets the overall bias, with the voltage divider (100k and R) setting the DC point of the gate. So you can't replace that with the cap/R/R as you have drawn it. Biasing the mosfets the same way that the tubes are biased will not work when you need 2A going thru them.

And you would still need the blocking capacitor between the tubes and mosfets.

Even if you fixed that, I doubt that you can just fit the mosfets to the board. They have to be mounted to a heatsink. A BIG heatsink.

Given that you are using the aikido PCB I would just go with the original plan as it makes much more sense.

Also unless you have a real reason (like you have very efficient speakers) I would go with the PP circuit. Because it's class A it does not suffer crossover distortion (at all). The more I think about it, the more I suspect that nearly all the reason that people love SE designs so much is that SE designs are by definition, class A.

So the bottom line is this : even if you could get this working, it would likely not sound as good. And I can't see what you gain apart from the cost of an extra tube and a few resistors.

Cheers
bob

[Valverine]

Hi Bob,

Thanks for your reply! I guess it clears my head (for the time being...)

I will go with SE version, the speakers are efficient. The MOSFETs are on their way; I plan to order the boards some time next week - need to decide whether to to 9-pin or octal (I remember your prev. comment on the sound, but choices for double triodes are more plentiful in the 9-pin format...)

I was also thinking about possibility of a piggyback board to be able to use single triodes – I guess I first need to have boards in my hands to see whether it is possible

Thanks again!

[beau2317]

In my experience there is quite a difference between the 9 and 8 pin tubes (the octals being a lot better).

I built an aikido with high spec 9-pin tubes like 6H30 and ECC99 and it still was not up to the octal standard. You might find the 6DJ8 is good sounding, but you still need a high voltage gain tube to do the voltage gain.

My suggestion is to go the octal route and use these special russian tubes called 1579 (search ebay). They sound better than my expensive RCA red base 5691 ! And at $8.50 a pair are a complete bargain.

In the second position use a 6SN7 or a 12SN7 (a 12 v heater version of the 6SN7). This is less critical anyway...

If you really want to go the 9-pin route then look for maybe 6072 in first position and 6DJ8 or ECC99 or 6H30 in second position.

But don't do it....go for the octals!!

I was thinking...the mosfets are half way there. They get hot. Now if only they glowed...shame you can't get BUZ900 in an octal package