For my modular preamp im working on a grandious version of a diamond buffer with cascoded current mirrors- 4 transistors per mirror, and only with to92 transistors. In trying to tweak the most out of the circuit by implementing a thermal tracking, i had an idea thats probably more doable then looking for an unobtanium to92 heatsinks.
What if, a flat headed to92s were used (such as zetex brand ones) and and they lined up single row front to back and then bam, tied all up with shrink wrap? Im not good enough to do the layout in my head but i imagine the traces will be suboptimally long. But still this simple solution of thermal tracking is hard to ignore.
What do you guys think?
What if, a flat headed to92s were used (such as zetex brand ones) and and they lined up single row front to back and then bam, tied all up with shrink wrap? Im not good enough to do the layout in my head but i imagine the traces will be suboptimally long. But still this simple solution of thermal tracking is hard to ignore.
What do you guys think?
P.s. i meant flat headed as in the back side of the transistor is flat, instead of the usual curve.
The idea you are talking about ---when it comes to TWO active devices at an input being in surface contact is an old idea that works .
I have done it as a means of obtaining a balanced thermally stable output which means a more stable offset adjustment. Of course its just an old version of "dual pairs " which usually cost more , but when you come to more than two the thermal transference isn't equal for obvious engineering reasons . In that case its better to terminate them onto a heatsink for a better distribution of a reduction in thermal temperature once an equilibrium is reached .
I have done it as a means of obtaining a balanced thermally stable output which means a more stable offset adjustment. Of course its just an old version of "dual pairs " which usually cost more , but when you come to more than two the thermal transference isn't equal for obvious engineering reasons . In that case its better to terminate them onto a heatsink for a better distribution of a reduction in thermal temperature once an equilibrium is reached .
unless you're really stuck on using TO-92 parts, consider something else. There's a member here who was (and maybe still is) selling 2SA1349 and 2SC3381 monolithic duals. Also, you could consider duals in surface mount package.
There even used to be quads in DIP packages but the specs of the devices were not necessarily as high perf (like quad 2N3904 or 2N3906).
And I "think" I've seen Ayre products that used TO-126 devices inline with a single screw clamping them together. Tough for layout, but probably pretty good for thermal performance.
🙂
Good luck
There even used to be quads in DIP packages but the specs of the devices were not necessarily as high perf (like quad 2N3904 or 2N3906).
And I "think" I've seen Ayre products that used TO-126 devices inline with a single screw clamping them together. Tough for layout, but probably pretty good for thermal performance.
🙂
Good luck