PLAN B analog blog

Replying to my own message here, please note that in the initial 22
units shipped, regardless of the range of the SMP04 spec we did not
see a droop faster than 5mv per second. 



--- In PLAN_B_analog_blog@yahoogroups.com, "(i think you can figure
that out)" <peter@...> wrote:
>
> I have received a couple of letters from owners of Model 23's
> regarding a rise in held voltages which begins after a few seconds of
> the voltage being sampled.  Please hear me out on this:
> 
> This anomaly is known as voltage droop and is a characteristic of the
> Analog Devices SMP04.  I've uploaded the SMP04 sped sheet to the misc
> folder in the photos section of this yahoo group so you can see what
> I'm talking about.
> 
> When I set off to design the Model 23, I attempted to address the main
> issues that ASR's are famous for f'ing up :  somewhat long refresh
time between taps (mening how long it took the device to grab and hold
the voltage form the previous tap), gross variations in thevoltages
once fanned form tap to tap, and fluctuations caused by the
> falling edges of the input trigger.
> 
> The SMP04, in conjunction with the microproc. driving it address these
> issues.  However, in order to get the accuracy of the inherent to the
> SMP04 (1mv accuracy form tap to tap) , there is another spec which may
> make this module problematic for people wanting to use it on slow
> passages, such as processing of slow keyboard lines.
> 
> The SMP04 has a spec called voltage droop, speced form between 2mv.sec
> to 25mv/sec which will cause the held voltages to rise at the same
> pace after a few seconds.
> 
> 
> Please note:  the 1mv spec refers to the accuracy (tracking if you
> will) of the four sample and holds.
> 
> In most applications where an ASR will be used, I dare say standard
> ASR applications, this will not be an issue because the sampled
> voltage would be updated before the rise begins.  However, if you
> attempt to use it for keyboard passages or other slow applications, it
> begin before you update.
> 
> - P
>