Importance of matching in expo convertors?!

[Ben Nise]

To create an exponential function, you only really need one transistor. I
think
the matched pair only comes into play if you want the exponential converter
to be referenced
to something (i.e, 0V in = some set current).

There are other mismatch mechanisms that affect effect BJT's besides Is,
namely
Beta mismatch and tempco mismatch.  If you are using two different devices
that were manufactured
from different processes, you may also have a problem with the exponential
conversion fudge
factor, NF.  (NF is somewhere around unity, but I'm not sure how much this
varies).

IC = Is*exp(VBE/NF*VT) 


Regards,

Ben Nise
bnise at sonici.com

-----Original Message-----
From: Rene Schmitz [mailto:uzs159 at uni-bonn.de]
Sent: Monday, August 30, 1999 12:45 PM
To: synth-diy at mailhost.bpa.nl
Subject: Importance of matching in expo convertors?!


>From the Ebers Moll equation follows: (With no matching assumed)
(with Ic >> Is) dVbe= kT/q * ln(Is1/Is2 * Ic2/Ic1) where Is are the reverse
leakage currents. Is depends on material, doping, geometry and temperature
(something else?). I don't know the exact dependancy, but I *think* that
for any given pair of transistors Is1/Is2 is a constant, if both
transistors are on the same temperature. One gets the usual form of this
equation when both transistors are matched, then Is1=Is2. Now, wouldn't
this mean that we needn't use matched pairs in expo convertors?!

Any opinions? of experts in semiconductor physics?

Bye
 René





error:asciiartisnotfully  \ uzs159 at uni-bonn.de
supportedwithyourcurrent   \ http://www.uni-bonn.de/~uzs159
displaydriversettings...    \ http://members.xoom.com/Rene_Schmitz