[sdiy] Temperature Compensated Exponential Converter Using SSM2164

[Michael Zacherl.]

On Aug 30, 2009, at 2:27 AM, Ian Fritz wrote:

> At 06:15 PM 8/29/2009, Michael Zacherl. wrote:
>
>> On Aug 29, 2009, at 11:41 PM, Ian Fritz wrote:
>>
>>> At 02:51 PM 8/29/2009, Magnus Danielson wrote:
>>>> Ian Fritz wrote:
>>>>> At 11:37 AM 8/29/2009, David G. Dixon wrote:
>>>>>> 2164 is advertised to give more than five orders of
>>>>>> magnitude of log conformance, and we only need about three for a
>>>>>> VCO, so I'm
>>>>>> not worried.
>>>>> Conformance at what level?  Do they say?
>>>>
>>>> The SSM2164 has log control input and claim good conformance for
>>>> 100 dB, which cranks out as 5 orders of magnitude. If you look at
>>>> the design you will see that CV hits the control voltage input
>>>> while the current input is wired to -15V.
>>>
>>> The question is how *accurate* is the conformance over the 5
>>> decades.  1%?  5%? ...
>>
>> that's the point: I wouldn't mind if the scaling is 0.9 or 1.1 V/Oct.
>> or whatever or how's the log conformance in general, only if the VCOs
>> in use behaved exactly the same as much as possible. IOW connect them
>> to the same CV and how much they do get apart over the range?
>
> To be clearer, it is magnitude of the deviation from a log function  
> that I am asking about.  It has nothing to do with .9 vs 1.1, since  
> that is set by the v/o trimmer.

yes! I tried to point out that a minimum (or ideally no) deviation is  
more desireable (and probably harder to achieve) from my POV.