[sdiy] DsPIC timer issue when used as DCO!

[karl dalen]

Tim, Eric, Dave, etc!

> I wasn't aware that there was an issue with dsPIC timers
> used to drive 
> the P08/Mopho/(Tetra?) DCOs. Do you have any pointers to
> this elsewhere 
> online?

This are in fact old news and has been discussed in several places!
I thought the issue had been resolved by now? People describe it as
poping/click sound between high and low notes.
Reminds me of the OSCar rubber band oscillators! :) 

anywho, heres one spot its discussed!

http://www.vintagesynth.com/forum/viewtopic.php?f=1&t=49602
http://www.vintagesynth.com/forum/viewtopic.php?f=1&t=49602&start=60#p528540

This picture suppose to be an actuall measurement from low freq triangle
wave and an instant jump to higher frequency, it do look a bit odd!?
Kind of  counting, counting!  oh! a new value! reset! deeeelaaaay!
And then load new buffer count? 

http://www.mediafire.com/imageview.php?quickkey=yijnnyyy2o5&thumb=6

The old 8253 data sheet says something like this:
After writing a Control Word and initial count, the Counter will be
loaded on the next CLK pulse. This allows the Counter to be synchronized
by software also."

Which suggest that an instant frequency change are done! But i recall
the 8254 had this in hardware, automatically from one of the other modes?
I have never noticed this kind click in machines like JX3P or JX8p
ect which could relate to the use of mode 3 in the timer!? 

>Does the NCO reset its counter after the terminal count is changed?  Say >the counter is set for a terminal count of N, and then is changed to >M<N.  If the counter has already passed M when the update occurs, then it >would have to wrap around to 0.
>Sounds pretty obvious and broken, and not something that would get >released to production, but who knows...

I get the impression that the dsPIC timers only loads new value when
it reaches last count , and not when the new value are written into the
count buffer? This would cause a delay but can one hear it?
 
> I haven't used the dsPIC timers for DCO control before, but
> it would be 
> good to know what to watch out for. My own experience with
> 
> microcontrollers in general (and Microchip parts in
> particular) is that > there are lots of undocumented 'features' and most folks > with any  > experience have a mental list of where the dragons lurk. > Eric

Indeed, i had to discover myself some pretty bizarre
undocumented timer *features* in the MSP430 some years ago! :(
And im a pretty lousy assembler programmer to!! But the latest
28 pin dsPICs are so tempting!

Besides that the P08 are far from free of problems, gain balance, distortion issues in the PA397 chip.its interesting to note that
DSI apparently had vast problems with the early runs of Evolvers
according to some people Dave and Doug manually selected chips to
resolve a serious RFI issue, clk leak into the PA397 or something!
But hey!! The chip is cheap so what! :)





      __________________________________________________________
Ta semester! - sök efter resor hos Kelkoo.
Jämför pris på flygbiljetter och hotellrum här:
http://www.kelkoo.se/c-169901-resor-biljetter.html?partnerId=96914052