[sdiy] CV offset question.

[Dave Kendall]

Thanks...

> Change the resistor connecting the wiper of the pot to the op amp to a higher
> value.  What's used there now?

It's a 100K going to the neg. input of the opamp, with the CV input also
connected via a 100k to the same point.

Will have a go at the weekend. The "Sudden Long Hours Day Job Monster" has
me in it's claws 'til then...

Cheers,

Dave 



on 21/11/05 17:52, Roy J. Tellason at rtellason at blazenet.net wrote:

> On Sunday 20 November 2005 07:11 am, Dave Kendall wrote:
>> Hi All.
>> 
>> Hope this is an easy one.
>> In getting the pitch wheel CV from my trusty PAIAmidi2CV to produce -5, 0,
>> +5 instead of the 0, +5, +10 that it does, I've tried a few of the simple
>> offset and scaling schemos around, but all have the same problem. The
>> offset adjustment is a pot with wiper connected via a resistor to an opamp
>> input, and the pot ends are connected to ±15V. This makes setting the
>> desired offset almost impossible, as just breathing on the pot seems to
>> make it drift. Using 22-turn trimmers doesn't help, as the range covered is
>> still too great, and there is drift with temperature changes.
> 

> 
>> First idea was to find a way of providing smaller ± voltages to the pot
>> ends. The smaller the voltages, the easier it should be to set the pot. As
>> one common setting would be ±2 semitones, the total range would only need
>> to be -.08V to +.08V, so the negative offset would only need to be -.08, so
>> the voltages sent to the pot ends could be very small.
>> Would using ± voltage regulators work?
>> I haven't found any regulators that produce such small voltages yet (but
>> I'm still a newbie really......)
>> Has anyone cracked this problem before?
>> Any ideas would be welcome.
> 
> No need for regulators,  just put a resistor at each end of the pot,  if the
> first suggestion doesn't do it for you.  Scaling them is simple,  if you use
> two resistors equal in value to the pot,  your range will be a third of what
> it was.  But try the above first...