Time constants in compressors

[ulysses]

Every time I've called them, I think I've talked to the same guy. Very helpful, and very willing to work with the little guys. I think they're little guys too, as far as semiconductor manufacturers go. I bought 100 pieces of the TO-92 LSK170, and used some in an OEM module alongside some Toshibas without incident. The Toshibas aren't any more expensive or difficult to get yet, but they might will be some day, so the good 2nd source is... good, I guess. I've also dicked around with some samples they sent me of the surface mount LSK170, and also the various forms of the LSK389 which is finally available. Kind of expensive, but if you consider the time it takes to match singles (time is a production cost), I think it's worthwhile. The original 2SK389 is getting hard to find, and expensive lately (since they haven't been made in like a decade and a half or something).

 

[bcarso]

Justin, do you have any data on the real offset voltage distribution of the SK389s? I suspect they are much better than the listed spec.

 

[ulysses]

I've only used about five 389s so far, in prototypes. They seemed to be about as close as the 170 pairs I hand matched, maybe better. But it's hard to tell because the mismatch in the input pair is not the most significant source of offset in my particular circuit. The only way for me to be more sure would be to install a 389, measure it, remove it, install it backwards, and measure it again. The surface-mount packages do allow this (the original through-hole 2SK389 doesn't really); but I'm still squeamish enough about soldering SMT chips that I haven't been willing to try it. Maybe once I get some more appropriate tools and more practice, I'll give it a try.

 

[bcarso]

Reason I ask is I would like to use them in a d.c.-coupled woofer EQ. One can always put in a trim adjust for offset voltages, but when much trim is required the temp drift is usually not so good.

What I'm probably going to do instead is use one half of a dual as a temp sensor and heater, and bias the other in-circuit device close to its zero-tempco operating point, which for these low threshold devices is close to zero Vgs.