First Watt F7 review

[Michael Rothacher]

 

[wdecho]

He has spent some time using and testing this Genesic GA10JT12 judging by the last article published. Any chance it will use this transistor? Could be that the last article is a hint on what is to come. It is fun speculating and guessing at least.

 

[Bigun]

LATFET - lower Gm, lower input capacitance which is more linear than vertical fets, negative temp coefficient at 150mA - 200mA, different biasing voltages than vertical fets, more expensive

Vertical FET - higher Gm, higher input capacitance which is more non-linear, positive temp coefficient until high currents (10A or so), cheaper, better availability

I was thinking that source degen (not emitter! oops) of vertical would reduce gm to same as lateral, with similar linearity (thanks to degen) and improve thermal stability so as to narrow the differences between the two devices considerably ??

 

[drjbf1]

He has spent some time using and testing this Genesic GA10JT12 judging by the last article published. Any chance it will use this transistor? Could be that the last article is a hint on what is to come. It is fun speculating and guessing at least.

I am sure he tests heaps of parts for his designs. I could be wrong, but I interpreted that article as implying reasons why he would NOT use such a part. It goes against all the principles of elegant simplicity and minimalism that is so important to FW and Zen type amplifiers. Too much complexity to get it to work well. Unlike the 2SK77, for instance, about which NP said at BAF 2015 "they give rise to designs that say 'just put me in there with a power supply, and I'll take care of everything else'." BJTs, or whatever type of part that is, don't seem to do that very easily.

 

[WBS]

I am sure he tests heaps of parts for his designs. I could be wrong, but I interpreted that article as implying reasons why he would NOT use such a part...

I agree.

 

[2 picoDumbs]

I agree.

I agree also.
However he only purchased those parts approximately 2 months ago, so even if he loved those parts I'm not sure he would ship something out the door that fast. I know Nelson is superman but that would be a bloody super impressive effort, to go from purchasing parts, to testing, to circuit design, to designing pcbs, to listening tests, to refining the circuit, to building and shipping the product for review all in 2 months.

 

[wdecho]

At least I sparked some more speculation. Let's keep Nelson laughing with some more guesses. The quote:

in an idle moment, I tried a variation of the circuit using a new concept and the result was striking, both in terms of measurement and sound. It’s a really different animal.

 

[Itsmee]

The F7 is very different, more similar to F5, but with some refinements and new ideas.

A bit more speculation, F5(ish) type input (possibly high current); here's the twist, the summed multiple outputs form the second half of a current mirror (multiplier), only a few extra active devices needed but still only a two stage amplifier.

 

[Srajan Ebaen]

The window to speculate has sadly closed

Well, not quite. Some mysteries remain

But the owner's manual PDF for the F7 is now available in our preview. Dissect away

 

[Zen Mod]

flabbergast !

 

[woody]

What preview ?

 

[2 picoDumbs]

This confirms it. Nelson got his time machine out spied on my eSCAPEd Goat and added positive feedback to cover his tracks.
We'll see if he has used lateral mosfets (like me) or he has gone the more fearless route with some other device.
Hahahaha

Edit: Now I know the real meaning of Star Trek Education.

 

[lhquam]

What preview ?

Here: http://www.6moons.com/audioreviews2/firstwatt2/F7.pdf

 

[MASantos]

So when does the turbo version come out?

 

[lhquam]

So when does the turbo version come out?

My guess is that a Turbo version would require cascoding to keep the JFET drain voltages within limits and, unless beefier single output MOSFETs were used, would require either extremely careful matching of the MOSFETs or source degeneration.

 

[lhquam]

Here is a schematic this simulates well, but I am not yet prepared to show the parameter values. Much more experimentation with the feedback resistor values is required.

 

[MASantos]

Does anyone have any idea about the "trick" mentioned to reduce RF noise on the psu?

 

[Zen Mod]

whatever it is , it's certainly pulled from old hat

pick your own poison ....

 

[2 picoDumbs]

Srajan
I want to see a picture of your Christmas present

 

[lhquam]

It would be nice to know which lateral fets Papa decided to use. My guess is that 2SJ162 and 2SK1058 would work, since they have the desired temperature coefficients for operation without source resistors.