LM4780 Paraclone

[jackinnj]

A problem with Darlingtons in general is that the bases of the drivers already have an internal connection to the emitter, so you are forced to have a split resistor between the bases of the pair (with a tap at the output).

In some brands (such as ST), it has a few 100 R at the most.

Not sure about the value of the internal resistor in the Sankens, but for most OnSemi darlingtons, it is in the 1 to 10 k range, so you can overide it with a 100 to 330 R resistor directly between the bases.

looks like 70R as the internal resistor in the Sankens

 

[capslock]

Well, might be the culprit, worth a try. Once I had tried the untapped resistor in my discrete darlingtons, I never went back. The difference was very obvious, jus looking at the scope.


I was just looking at your brigeclone article. Have you tried an opamp for phase inversion and then let both halves of the 4780 run in the same mode?

This might give you even lower distortion, as both amps will generate exactly the same spectrum, and an opamp will probably more faithfully invert the signal than a 4780 ever can under load.

 

[jackinnj]

This might give you even lower distortion, as both amps will generate exactly the same spectrum, and an opamp will probably more faithfully invert the signal than a 4780 ever can under load.

While I considered it, I didn't try it out. Didn't want to have to deal with the issue of powering and regulating a device on the board.

I have plans/gerbers (they are somehwere in my Ultiboard files) for a configureable parallel and bridged 'clones -- incorporating a quad opamp on the board and some molex female headers such that you could configure for low pass, high pass, all pass, etc. if there is interest i could have some made up.