Power follower class A Mosfet DC coupled.

[Mr. Steve]

Here is an amp I threw together over the weekend. Very simple power follower with a lm350 used as a constant current source set between 1-2.5 amps depending on the mosfet used. DC coupling is adjusted by 100k pot. Pre amp required to get full output. Choke or chip regulated dual supply required. I use a choke regulated supply and ripple is below audible levels. Zener is to protect mosfet. I used IRF510 5A 100v and it sounds excelent, very clean with low distortion. Larger mosfets can be used but you cant beat the 510's 150pf input capacitance. I'm going to build a larger version with some 230's and p240's at higer voltages with a different regulator. I would be happy to hear any suggestions and someone please build one and tell me what you think, It doesnt take long. Schematic Included.

[boholm]

Originally posted by Mr. Steve: Schematic Included.

Where ?

[Mr. Steve]

schematic repost

[jeapel]

hi

I just read your post oct 2004 ?? but i build similar
project with good result.

www.loginnovation.com/jeapel

if you jump from irf 510 (150pf) to irfp240 1500 or 2000pf
you need a low zout driver

good side:

low distortion without feedback
good damping factor

bad side:

very neutral sound so a tube driver give some good harmonics

hoping you read this may 2005 ??

[sajti]

I use this tube/mos balanced follower:

sajti

[triode_al]

Originally posted by Mr. Steve
schematic repost

I have tested this schematic with a 1,8 Amp currentsource using a LM323 volt regulator and a 2,7 ohm Dale 25 watt resistor.
It does cook a meal at 25 volt.

This was a lot better sounding than a IRF250 source with a current feedback with a BC546 Wilson transistor. That tended to be very harsh sounding). Even with a 1000 ohms gate stopper.

You need a lower bottom rail of course.

The problem I have with a pair of IRFP250/IRF250 running at 1 amp and 25 volt is that it sucks details so the level sort of is uneven - an unwanted physiological effect.

[jeapel]

hi sajti very interesting design and if possible
i would like to see photo

-first your mosfet output:

1-your ccs bias ~3 amp is very similar to me but
without my small .01uF on gate my ccs is
unstable with high frequency
2-irfp 140 => cin=1700pf =>ron=.077ohm
irf540 => cin=1700pf =>ron=.077ohm
=> irf540 ~= irfp140
3-you are lucky to get big heatsink like that here
in quebec it s very difficult to get big heatsink
4-i like the auto bias with integretor tl072

-tube input:

1-ok long tail with ccs on cathode but which tube and
what ma. bias in each tube
2-why c3 and c4 return to Out and not ground
3-my infos is to drive mosfet with low zout tube to
integrate tube and mosfet together
-Pavel Macura follower said zout max.= 1000ohm

-http://www.zenn.com.sg/Why%20hybrid.htm

i am very intriging about your high frequency
and your high zout drive did you test with oscillo
bye :-)

[jeapel]

http://www.zenn.com.sg/Why%20hybrid.htm

[jeapel]

hi triode_al

"running at 1 amp and 25 volt is that it sucks details so the level sort of is uneven - an unwan"

I run 26 volt 1.55 amp with irfp140 ccs and the sound
is very good i dont understand your point maybe more
explications :-)

[triode_al]

Originally posted by jeapel
hi triode_al

I run 26 volt 1.55 amp with irfp140 ccs and the sound
is very good i dont understand your point maybe more
explications :-)

Jeapel,
The difference is the CCS which is (IMHO) probably better than an N/P pair - maybe I did something wrong, but there is an effect that the music is louder than one moght expect. As if the FET pair stores the signal and then --- jumps. I know this sound crazy.
But on my scope I see no problems, a 4 ohm// 0,5 muF is completely solid rising edge; and the 8 ohm/4 ohm changes chows no drop in level.
Even frequency goes to 200kHz +.
Still this singular effect of what I call 'physiological'.
So I have set aside this N/P pair with currently SJ135+SK50 one channel and IRF150/SK50 other channel aside.

My plans: I will restore my earlier CCS loaded IRF150 at 1,8 Amp. To see how that is again now.

I don't expect to understand the strange behaviour of the N/P pair. Could it be that one transitor blocks the other and blocks a movement of the voltage until some microseconds? Any idea someone? So here we have a secret of the SE style of Steve's and your design!?