matched transistors

[René Schmitz]

Hi all,

I second this statement. At least as far as the expo is the intended
application. Matching of cheap duals or arrays is much better than spec'd.
(Besides I still believe that you don't need a tightly matched pair in
expos. The fact that you can build precise and stable VCOs with cheap
parts, suggests that its a myth.) 

However there are good reasons towards LM394, MAT0x etc. when you are using
these in the audio path. And in expos it eliminates the need for adjusting
Rbe compensation, which can save time when you do "mass production".

Bye,
 René

At 19:21 14.04.00 -0600, Ian Fritz wrote:
>I would propose some softening of this statement. If you just go by the spec
>sheets, then yes, the inexpensive arrays seem pretty bad. However, as I have
>mentioned previously, modern parts often far exceed specs. For example I
>recently bought a half dozen CA 3083's. All but one have offset voltages too
>small to measure reproducibly -- < 0.2 mV. Betas are well over 200. Series
>resistance is only 1.3 Ohm (which can be exactly compensated with a diode
>and a trimpot, anyway). Are you sure you need a lot better than this? Have
>you compared the other sources of error and drift with those stemming from
>using this "cheap" array?  I'm using one of these in the VCO I'm currently
>developing, and I don't see much (if any) difference in performance going to
>an expensive pair.
>
>  Ian
>
>----- Original Message -----
>From: "Haible Juergen" <Juergen.Haible at nbgm.siemens.de>
>>
>> Depends on the application.
>> When you're talking about CA3046 and the like, these are fine if you only
>> need
>> equal temperature for both transistors. In some applications you also need
>> low offset voltage, low offset voltage drift, high beta, low noise. Then
>the
>> cheap
>> array will not do.
>
>
>

transconductance | uzs159 at uni-bonn.de
isfutilepreparet | http://www.uni-bonn.de/~uzs159
obeassimilated.. | http://members.xoom.com/Rene_Schmitz