Using LU1014D in parallel for a power amp stage.

[myfirstamplifier]

Is there any circuits available to use LU1014D such as in a Pass PLH or similar output design that I can tinker with? I have a bunch of closely matched LU1014D and wonder if I can simply parallel a bunch of them to get 10w @ 8 ohm? I found this circuit that inspired me somewhat:

 

[andynor]

You could take inspiration from the Beast with a thousand JFETs. But you’ll need a few beast cells And you can probably easily avoid feedback alltogether. Worth looking into. Of course, the Beast being complimentary sorta rules out the Lu, but on that case the B1 circuit times many might do the trick

 

[Nelson Pass]

I say make a board that holds 10 of those circuits and find out. Maybe give the group an input buffer that can drive 1kohm...

 

[generg]

Very nice Idea!
I had to simulate it at once.
What did I do wrong? Too much distortion for 1W/4R and 2A bias. Bad bandwidth.

 

[tombo56]

What did I do wrong?

Maybe give the group an input buffer that can drive 1kohm...

You can't drive all those input capacitances through 10K. Check by using R21/23 at 10 times lower values.

 

[generg]

o.k.!

 

[Zen Mod]

even better - insert buffer ditto to gate of J3, so NFB loop is all around

that way you can leave it with present values, or go even higher

 

[generg]

But my friends the signal source in Spice has a very low Rin! Isn't this as good as a buffer?

:--))

 

[generg]

:--))

 

[Zen Mod]

But my friends the signal source in Spice has a very low Rin! Isn't this as good as a buffer?

:--))

problem is - whatever Rout of source is, you still have 10K in series

same penalty as with Schade mosfet - 1K in series with gate is price, resulting in loss of highs etc.

only when you arrange substantial (or entire) drive through source ( grounded gate mode) things are going back somewhat

so, if you want to tackle problem fully, insert buffer in NFB loop, ditto to LU gates

hehe, maybe you can even substitute NFB with Schade loop, in that case

 

[generg]

I understand half "Bahnhof"

:--))

Bandwidth is better with 2k2 and 520R.

 

[tombo56]

Back to the drawing board.
ATM, I don't use power amplifiers with less than 1 MHz power bandwidth.

 

[Quantensprung]

Like the amps from Spectral? I always wanted to hear one but until now never got the opportunity. A diy spectral would be nice 👍

 

[Zen Mod]

I understand half "Bahnhof"

:--))

Bandwidth is better with 2k2 and 520R.

I didn't simulate anything

 

[generg]

for myfirst amplifier...

:--))

 

[generg]

Big thanks Aleksandar, I am good In simulation trying to be an expert my whole life.....

:--))

 

[generg]

Funny not the expected gain, good bandwidth, does anyone see again a fault in my draft?

 

[Zen Mod]

add more LU cells

 

[myfirstamplifier]

Thank you Master Zen and Generg. This is extremely helpful to get me started.

 

[generg]

Done....
:--))

What why is the gain so slowly coming, even with 10 elements only 8dB gain