[sdiy] Linearized 2164 VCA Warning

[Neil Johnson]

Hi David,

> Ah!  Sorry, I misspoke (mistyped?).  I'm not actually that stupid  
> (at least,
> most of the time).  I checked my email archives, and it was in fact an
> OPA2132 that I heated up.  I had a couple specimens lying around so I
> replaced the TL072 with it based on Ian Fritz's suggestion.  He was as
> surprised as I was that it heated up, as neither of us could see  
> any reason
> for it to do so based on my schematic, which (at least for that  
> part) is
> really quite ordinary.  Of course, the offset specs of the OPA2132  
> are not
> that much more spectacular than a B-grade TL07X.

<humour>
OPA2132:

http://www.rock-grotto.co.uk/opamp.htm

"Smooth, highly analytical, warm, uncoloured, laid back yet all  
detail is present"
</humour>

Its a high-speed op-amp, so I hope your feedback loop was nicely  
compensated, otherwise it could have been oscillating at a few MHz,  
causing the aforementioned heating.

> (Does this qualify as an "Emily Latella moment"?  If so, then  
> "Nevermind!")

Not sure... possibly a "Caffe Latte" moment?

> My misadventure with the OP-07 was something entirely different.   
> When I was
> having such problems with my 2164 Expo VCO, I was seriously  
> thinking about
> using an ultra-low offset opamp to convert the 2164 expo current to a
> voltage and then send this voltage into the saw integrator.  I did an
> experiment with the OP-07, which has very low offset indeed, but  
> the fact is
> that there is no offset low enough for this particular application.
> Ultimately, I found another solution.

Well, the OP07 will give you very low offset voltage, granted.  But  
the bias currents of the bipolar input stages are huge compared to a  
FET input op-amp (for example, Ib(typ): TLC272: 0.6pA, OP07: 1.2nA) -  
it may be that in that particular application the bias currents  
dominated error budget, resulting in poorer performance as you found.

> Hey, they both start with OP!  :B

Well, you're definitely an OPtimist!

Cheers,
Neil
--
http://www.njohnson.co.uk