[sdiy] Prophet 5 scale problem

[Adam Inglis]

Thanks Michael
I had a look at the output of the buffers for the SH circuits after the 4051. I’m looking for stable voltages here, yes? They all seem OK. 
As I understand it there is no exact right voltage for reference here, as the CPU calculates a bias that will be different for each of the 10 VCOs.

I should have been a bit more precise - the problem exists only at oscillator frequencies below high C - about 520 Hz. Wherever I pitch the oscilaltors across the keys, the problem follows the frequency. It is irrespective of the key CV.

2 things occur to me:

1. The P5 uses the upper 7 bits of the DAC for the voltage controllable voice functions, only the individual VCO pitch CVs use all 14 bits.

2. If the tuning mux (4051) or it’s comparator were faulty, the CPU may read the frequencies incorrectly during computune, thus generating incorrect biases, hence throwing out the scale accuracy.

?

AI


> On 13 Oct 2016, at 10:18 AM, Michael E Caloroso <mec.forumreader at gmail.com> wrote:
> 
> You do not mention if you probed the key CV path fanning out from the
> 4051 with a 'scope (DMM isn't enough).  That should tell you a lot.
> 
> MC
> 
> On 10/12/16, Adam Inglis <21pointy at tpg.com.au> wrote:
>> I have a P5 here that seems to be working OK except that the lower ranges of
>> the 2 VCOs for each of the 5 voices are randomly off scale/out of tune
>> across the keys.
>> 
>> It seems to correspond roughly to a key CV lower than 3 volts with the VCO
>> pitch pots at their mid-range (as evidenced by the CV out jack). Above this
>> both VCOs track well in tune and scale. If I pitch both VCOs to their
>> highest range, the top 4 octaves of the keys are in tune, just the lowest is
>> out. Doing the inverse, most of the keys are out of scale except the top
>> octave.
>> It does appear to be all 5 voices affected, so I’m having trouble narrowing
>> it down to one chip, say e.g. a CV demuxer 4051, which are a common problem.
>> 
>> All 14 bits of the DAC are pulsing.
>> 
>> Further tests are obviously needed, but any clues greatly appreciated.
>> 
>> AI
>> _______________________________________________
>> Synth-diy mailing list
>> Synth-diy at dropmix.xs4all.nl
>> http://dropmix.xs4all.nl/mailman/listinfo/synth-diy
>> 
>