[sdiy] korg ms02 antilog amp and log amp

[Ian Fritz]

At 10:59 AM 3/28/06, René Schmitz wrote:

>>IIRC, the CA3083 is even less expensive than the Japanese pairs, and you 
>>can use one of the free transistors to make the diode for the hf tracking 
>>circuit.
>
>That is nice, of course. The price is way lower here too.
>I had thought about using the CA3083 for quite some time, and I would try 
>to parallel the transistors to maybe reduce the Rbe, and possibly getting 
>a poor mans version of common centroid layout. (But for that it would be 
>nice to have a picture of the metallisation mask for the CA3083) Do you 
>have made any tests how well the other transistors match to the matched 
>pair transistors?

Yes I have.  They are always worse than the matched pair.   I measured some 
of these when I was considering using a second pair for the linearization 
diodes in a DIY OTA.  You can check all three possible pairs and sometimes 
find a good one, but usually you will find between .1 and .3 mV 
mismatch.  For an expo this might not be a killer, but you would have to 
set up to do some careful measurements.  I ended up using a bridge circuit.

>The only thing that bothers me is the size of the package, its rather huge 
>compared to the japanese duals.

Yep, that's a drawback.

>Also coupling the PTC element to the chip is somewhat more cumbersome than 
>with the japanese pairs. (I use the KTY81, which comes in a TO92 case and 
>can be conveniently glued face to face on the 2SC types, whereas the leads 
>are to short to have it glued to the DIP size cases.)

Sure.  I use the big-old wirewounds, and they fit nicely along the top of 
the CA3083 DIP.  :-)  One might also use those tiny SMT Panasonic units.  I 
have a bunch, but haven't tried them in an actual circuit yet.

   Ian