New Design-HEXFET Poweramp. Sonic benefits of this approach.

[Mooly]

Where to begin ? Well I wanted to see just what was and wasn't possible with HEXFET's. I have sometimes been a bit less than complimentary about them as audio devices, but thought it about time I gave them a whirl.
Some of you will be familiar with my MOSFET amp on this forum which uses Laterals FET's and that circuit formed the starting point for this little adventure.
Why stick with an existing design -- because it's tried/tested and above all else, it sounds great.
Right on to the HEXFET's -- prone to oscillation, prone to thermal runaway, not to mention the unreasonable high gate threshold voltage-- no they have no place in linear audio. 10 minutes then to show that they really are that bad. Er well not quite -- I was impressed ( actually very impressed ), and one thing led to another such as seeing if they could be taken out of the feedback loop altogether.
So here it is, very much work in progress, but with a few actual shots to show what really happens.
What I really want to know ? Anyone --- how does taking the outputs out of the loop affect sonics.

[Mooly]

This is at 20Khz with no load and then with 5 ohms loading. For a single pair of outputs I was amazed -- that loss amounts to around 0.67 db. Very good indeed. This is at 16 volts peak/peak.

[Mooly]

Squarewave response at 20Khz again no load then 5 ohms resistive ( Well a 50 watt wirewound resistor ) This is showing 20 volts peak/peak.

[Mooly]

Now this is the killer. 20Khz into 2.2 mfd. The shot on the left is with the amp set up as before, the one on the right is with the 18K feedback resistor connected conventionally.
So the big question -- is this an approach worth pursuing ?
I had to readjust the 'scope sensitivity to fit it on, the amplitude didn't suddenly alter
No output inductor either.

[nikwal]

Yeah ,that's what i've been saying for years , hexfets are alright if you drive them hard :-P
many years ago I used a irfp hexfet buffer with a 6n1p as a preamp built as a mu follower(also wth a hexfet). Thought that was pretty ok, might have been even better with just a little feedback..
Hmm I use lower gate resistance than the ones in the schematic hmm in fact i think I even tried with like 12ohms and 3ohm and so on, dont remember which one that was left there, I might even have replaced it with a wire hahaha hmm .. oh well

[destroyer X]

I am glad to know.... maybe i will try this one.

regards,

Carlos

[gain]

Mooley, good to see ya! congrats on your new design!!! very sexy square wave pattern ... nice job!!! thank you for all your hard work and contributions to diy!!!

[gain]

oops just read post#4. hummm, little ringing in that ^2 wave. i'm not worried, you'll figure it out and come through with another amazing design ... keep at it and thanks again!

[eyoung]

hola you all...

I bought a bunch of IRF640/IRF9640 to play with (they are cheap). I was looking to incorporate them into an output section ala Mr. Pass or one of Mr. Rollins(I do not have any small signal jfet's, but I have a bushel of small signal bjt's found in haflers)
Let us build somthing...


Regards, Elwood

[lineup]

Originally posted by Mooly

Right on to the HEXFET's --
-------
-- I was impressed ( actually very impressed ),
-------
So here it is, very much work in progress

Mooly.
I actually posted a similar idea over 1 year ago.
Using HEXFET for Voltage 'VAS' amplification.
With some circuit schematics to illustrate...

Regards, Lineup


Solid State >Experimenting ... Volt Stage Hybridically
17th June 2007
Experimenting ... Volt Stage Hybridically

Originally posted by lineup
.
Experimenting with Voltage Amplification Stage.
There are many ideas about input & ouput transistors.
They are important too.
But to have a good and not to un-linear VAS stage .. is a real blessing.
At least when every other stages are trimmed to last decimal.

As you can see from attachment
this experimental setup use IRF TO-220 HEXFET for voltage amplification.
At a comparatively High Current.
My first version here, which shows good performance
uses JFET 2SK170 for input and TO-3 MJ15024 / MJ15025 for output.
.

Finally, this is according to my ideas.

How I have got the strength of diffrent transistor technologies:

1. JFETs are good for input. One quality is High Input impedance and low bias current.
2. MOSFETs are are good voltage amplifiers, they are fast!
3. BIG BIPOLAR working as Voltage Followers have low losses.
Even if now are very good Output FETs
that are just as good and probably will be the future ahead.

Audio Regards
lineup