P
Paul Burridge
- Jan 1, 1970
- 0
Hang on! Let me guess: it's virtually impossible to get the
heat-response (Vf/If) curve of external diodes to match that of the
transistors they're mounted on?
Except the scheme doesn't work very well. It has a fundamental flaw.
Hang on! Let me guess: it's virtually impossible to get the
heat-response (Vf/If) curve of external diodes to match that
of the transistors they're mounted on?
I was retained by ICE (Integrated Circuit Engineering) around 1980
to analyse the LM195/LM117. At that time the circuits were virtually
identical.
I have a very faded copy that I will attempt to restore by scanning
enhancement. If I succeed I'll post it to my website.
Winfield said:Paul Burridge wrote...
Which is why we show a diode-connected transistor on page
94, immediately after the simple introduction, which that
amused Jim. Most commercial amps use a scheme similar to
figure 2.59. Of course they all suffer from the fact that
the power transistor's case temp doesn't match its junction
temperature, so the better ones generally compensate the
low-dissipation input transistor in Sziklai output pairs,
etc. BTW, we use the common shorthand of drawing a diode-
connected BJT as a diode. NOBODY just uses diodes. E.g.,
when you get access to a temperature-measuring diode in a
CPU chip, it's actually a diode-connected NPN transistor.
Jim Thompson wrote... [snip]I have a very faded copy that I will attempt to restore by scanning
enhancement. If I succeed I'll post it to my website.
I have old data sheets for these parts, but I imagine the design
schematic shown on them didn't change over the years. Does the
drawing in your old copy differ from modern .pdf files available
now at the manufacturer's web sites?
BTW, although I've admired the LM395 from the beginning, and kept
inventory, I usually use an LM317 or similar part for a protected
pass element. The LM317 and its cousins are available in serious
power-capability versions, whereas the LM395 is a wimp. In power
supply circuits the big difference between using a LM395 and LM317
as a pass-element is that the LM395's base input is 0.9V above the
output, whereas an LM317's adj input is 1.2V below the output.
Generally the LM317's more negative control voltage (by 2.1V) is
more convenient for driving from the output of an opamp, etc.
Finally, I used the LM395 in a design last year, and it worked
OK. An LM317 would have worked nearly as well, but in this case
I actually preferred the LM395's wimpy lower current limit. :>)
Thanks,
- Win
whill_at_picovolt-dot-com
I've never used an LM395... just LM317's and LM337's.
Which is precisely the motive for my original question. So you don't
know either why that hasn't prompted the power transistor manufacturers
to include diode connected sense transistors?
Paul said:Go for it, Stephan, you may well be on a winner!
Stefan said:Winfield said:Most commercial amps use a scheme similar to figure 2.59. Of
course they all suffer from the fact that the power transistor's
case temp doesn't match its junction temperature, so the better
ones generally compensate the low-dissipation input transistor
in Sziklai output pairs, etc. [...]
Which is precisely the motive for my original question. So you
don't know either why that hasn't prompted the power transistor
manufacturers to include diode connected sense transistors?
Go for it, Stephan, you may well be on a winner!
For an audio-amp manufacturer who cares, using the proper form of
Sziklai output stage, the input transistors are widely available
as pairs in one package. More to the point, they won't warm up
much, reducing the need for accurate tracking in the first place.
Thanks,
- Win
whill_at_picovolt-dot-com
For an audio-amp manufacturer who cares, using the proper form of
Sziklai output stage, the input transistors are widely available
as pairs in one package.