HEXFET amp

[silverchris]

HI I havve a design for a HEXFET amp that doesn't work. I was wondering if anyone can point me in the right direction for getting it to work? I aperciate all comments even if it is telling me I am an idiot lol. It was designed by a friend very quickly.

[richie00boy]

Shoot your friend. Is he a first year EE student by any chance?

If it didn't have major faults that prevented it from even working, it would sound utterly dreadful due to operating in class-b with a very wide transistion region.

[silverchris]

wanna point out the faults in it? I can put them together but not design or find faults. also I would like to know how to make it class A or AB

[richie00boy]

1. Only unity gain operation
2. Not possible to add a bias generator (so it will never be anything other than class-b)
3. Inadequate drive from first stage
4. No feedback from output stage which is horrendously non-linear
5. Upper and lower parts of output stage operate in different modes thus have totally different output impedance for positive and negative swings

There's probably some more if I spend more than a minute.

Sorry but it's pointless trying to do anything with this as a power amp. The best you could do would be stop trying to make the output push-pull and convert one of the output devices to a current source. That would gain you class-a operation, but performance would still be very poor especially if run off anything other than a powerful stable supply.

If you want to use n-channel MOSFETs there are circuits around that do the job much better.

[silverchris]

I see. Can a HEXFET be used as a N channel MOSFET( Excuse my newbieness)? Pr os ot by chance the same thing. I already have 12 IRFZ46N in my parts bin. Thanks for the info

[richie00boy]

A HEXFET is a MOSFET. You have n-channel ones (which are by far the most common) but you can also get p-channel ones.

Unfortunately the ones you have are quite low voltage, limiting you to just a handful of watts.

[Tim__x]

Change the 1meg to something around 220k or 100k. Add another resistor of equal value from the base of the PNP to the positive supply. And change the collector resistor to a 1k pot (this becomes the bias pot).

It'll work fine. Don't expect spectacular performance, but it'll work much better than richie would have you believe.
I'd reccomend class A if you're going to use it open-loop.

Look up the JLH amplifier for more insight into this topology.

[Tim__x]

P.S. thermal stability isn't that great, so use hefty heatsinking, and possibly a source resistor on the lower mosfet.

[lineup]

Originally posted by silverchris
HI I havve a design for a HEXFET amp that doesn't work. I was wondering if anyone can point me in the right direction for getting it to work? I aperciate all comments even if it is telling me I am an idiot lol. It was designed by a friend very quickly.

Another guy, which I gave some advice in this topic,
My simple class A approach, has made a working amplifier!
Not only working, but GOOD WORKING!

Is also jerluwoo FIRST TRYOUT. And almost the same idea as yours.
Even if you use HEXFET and jerluwoo use 2N3055 ordinary big power transistors.
http://www.diyaudio.com/forums/showt...22#post1017622

When I say it is almost same, then I mean the output part.
Because jerluwoo, as you can see from schematic
has some input transistor and another one providing Voltage Gain.


As said, your output stage, will give more current to loudspeaker
but the voltage gain will be less than 1 ( less than Unity Gain )
and a power amplifier is most often needed to give
BOTH current gain AND some voltage gain.


I attach the output stage of jerluwoo simple but good WORKING amplifier.
It is the two 2N3055 (NPN) he uses instead of your both N-Channel HEXFET.


lineup

[ilimzn]

I would suggest searching for Nelson Pass' 'PLH' amplifier, either here on the forum, or at passdiy.com...