[sdiy] Matched transistors - critical parameters?

[Paul Perry]

No reason why they shouldn't match, I agree.. and I would hope that the two 
halves of a LM13700 were reasonably alike - but going by the data sheet, the 
transconductance specs do indeed vary by 2 to 1 (well, min 6700umho, max 
13000umho.)
Matching of the transconductance tracking is 0.3dB - certainly good enough 
for my ears if you are making a stereo VCA, but you can imagine some SDIY 
applications where there could be problems.
----- Original Message ----- 
From: "David G Dixon"
> Well, in the interests of full disclosure, I typically put current mirrors
> on my filter OTAs.  Of course, I haven't used an OTA in years, ever since 
> I
> discovered the 2164 quad VCA, and I've never looked back.  No expo
> converter, no currents or current mirrors -- just a single buffered 
> control
> voltage sent to all VC pins, and four VCAs gain-matched to within a few
> hundredths of a dB.
>
> But, looking at the schematic of the LM13700, I see no reason why the Iabc
> pin current draw of different OTAs shouldn't be very similar.
>
>> And maybe this is why multiple OTA filters sound so different
>> from each other.. my workshop is in storage, but maybe
>> someone can take a few measurements. I think that
>> transconductance is not guaranteed to be better than 50% in agreement.
>>
>> ----- Original Message -----
>> From: "David G Dixon"
>> > In my experience, with multistage filters, if you are
>> sending the same
>> > current to the Iabc inputs of OTAs, you don't even need to
>> mirror the
>> > current -- just connect all of Iabc's directly to the
>> current source.
>> > Each
>> > one should suck the same amount of current.
>
paul perry melbourne australia