Help with improved Wilson current mirror

[nelsonvandal]

I haven't got enough transistors to perfectly match all four transistors in the improved Wilson CM. Does it matter if one of the pairs have higher Hfe? Should the transistor pair with higher Hfe be at the "bottom" or the "top"?

[vynuhl.addict]

As the CM is degenerated with the 68R resistors HFE matching has some room for mismatch, as for top or bottom its up to you as this would come down to personal preference. Discreet AD797 i see?, the second post with the DB output stage gave it away .


Colin

[eyoung]

HOLA Y'all,

I would like to see the rest of the schematic, the front end looks interesting...


Regards, Elwood

[jam]

nelson,

That is almost the same circuit I have used. The lower transistors are probably more critical but I would replace them with similar types when you can.
How do you like the sound?

Jam

[ostripper]

Wow, 4t on bottom and a 3T on top.. . I'm about to build
something similar with 'frugalamp 2' and have simmed the hell out of it with a 3 and 4T on the bottom CM. With just a 3t
(final choice) on the bottom CM I couldn't get much unbalanced
operation with even grossly mismatched Re's. I then simmed
with different devices (for the hell of it) and by adjusting [Rj]
brought it back in balance again.
BtW like your input.. might try fet input or OPS in future..
OS

[nelsonvandal]

Originally posted by vynuhl.addict
Discreet AD797 i see?, the second post with the DB output stage gave it away .
Colin

Actually it's a FET input discrete AD829 clone, but I'm also going to build a FET input AD797 like in the picture below. I'm just about to order some parts. With the "magic" transistor, it simulates almost no distortion as compared to "AD829". OTOH it simulates 100 dB OLG while "AD829" simulates 64 dB.

Originally posted by jam
nelson,

That is almost the same circuit I have used. The lower transistors are probably more critical but I would replace them with similar types when you can.
How do you like the sound?

Jam

You're right, the "lower" pair is more critical. I guess it doesn't matter if there's some divergence in Hfe between the upper and the lower pair, right?

As for the sound it fares very well compared to the original AD829 and AD797, but I'm probably biased. The choise of transistors is very critical when it comes to sound quality. With BC550/560 it's worse than the monolithic amps, but with those Sanyo's it sounds perfect to my ears - warm and full yet with all the details and authority. With Toshiba SA970/SC2240 the quality is the same, but I find it to bright. I've just found out that it sounds better with ONSemi's BD137/138 vs Sanyo SC3953/SA1538 as output transistors. I had such high hopes for the latter.

[nelsonvandal]

Originally posted by eyoung
HOLA Y'all,

I would like to see the rest of the schematic, the front end looks interesting...


Regards, Elwood

This is the FET input AD829-clone.

[ostripper]

What a lovely preamp. Is it class A?? At least you show schema,
unlike J. Curl ... will sim. Thanks..

[nelsonvandal]

Originally posted by ostripper
What a lovely preamp. Is it class A?? At least you show schema,
unlike J. Curl ... will sim. Thanks..

Yes, it's class A driving most headphones at my listening levels. The output transistors are biased to 20 - 25 mA.

[ostripper]

Have you built it , or only simmed..??

Will it run on +-12v (+- 5 is kind of low)?