[sdiy] VCO Accuracy

[Tom Wiltshire]

On 30 Dec 2007, at 22:45, John Mahoney wrote:

> At 05:15 PM 12/30/2007, Tom Adam wrote:
>
>> A noob question about VCO's...
>> How accurate does one calibrate a midi 2 CV unit (+/-0.001V)?
>> How accurate does one calibrate a VCO? (+/-1Hz)
>> I understand that the accuracy is limited by design, but what is  
>> the "allowed" deviation before our human ear detects the error?
>>
>> I'm about to calibrate a MOTM 300 & 310 and an oakley VCO.  
>> Although I've tuned them after building I'm looking to get all 3  
>> of them in "perfect" tune.  Is there an easy way to accomplish this?
>>
>> Thanks,
>> ToAd
>
> No such thing as a perfectly in tune VCO, though some are better  
> than others. The best you can do is optimize one for a given  
> frequency range.
>
> For "perfect" tuning you need to go digital.

Just to make you feel better about it, there's no such thing as a  
perfectly in tune digital oscillator either - it's just that the  
performance that you can expect improves. You're still dependant on  
the temperature-sensitive accuracy of a quartz crystal, and usually  
on the limited accuracy of some fixed length counter.

With respect to the Korg Polysix, in my experience the manual's +/- 5  
cent deviation is really a best case. Perhaps they managed when the  
instruments were brand new leaving the factory, but if you get it  
that close now (20 years later) it's a real balancing act!

T.