DSPIC33CH128MP202 share ADC pin between master and slave

AlastairM · Joined: 28 Apr 2008 Posts: 32

Is it possible to share (internally) an ADC pin in the slave and master code?
e.g. pin1
master:AN1/RA1
slave:S1AN15/S1RA1
Or does the CPRA fuse setting mean that the pin is physically 'master' or ''slave' and just not available for the other core?
I wanted the master to monitor the waveform frequency in the background while the slave samples
I have a spare pin for an external connection but I'd prefer not mod the pcb...
Thanks
Alastair

temtronic · Posted: Tue Jan 21, 2025 7:18 am

yessh , I have enough 'fun' with one PIC in a package and you have TWO !!
Honestly, I'd look at the 'drawings' in the datasheet, only what, 5-6-700 pages to scan.......
What you ask seems 'possible', actually a good 'feature'. I know some PICs can 'interconnect' internal peripherals(her that here) so it 'might' be possible.

My first thought was of course a jumper.....that's what a 'dinosaur' like me uses !

edit. sigh, I downloaded the datasheet..834 pages !!!! I tried to scan,to see if the ADC s could be linked but yeesh, 'it's complicated' !!! sorry....
I want to got back to 16C84s !!!

AlastairM · Joined: 28 Apr 2008 Posts: 32

I found that bit of the datasheet quite confusing...
I'll try a few configurations on the actual hardware and come back after I've finished 'hacking'

Ttelmah · Joined: 11 Mar 2010 Posts: 19649

I suspect you probably can. Obviously only on a pin that is accessible both
to the master ADC and a slave ADC. However I'd also suspect that the
accuracy may well be degraded slightly.if you do this..There will be some
noise introduced when the sample and hold connects and disconnects, so
for a low accuracy application it might be OK. It is noticeable that the
examples from MicroChip, where they want something controlled by a
voltage always work from one core, and send the result to the second
core.

AlastairM · Joined: 28 Apr 2008 Posts: 32

You're correct about the signal integrity - but I'll live with it for now.
It worked OK on the hardware!
I now have the slave sampling continuously and the slave sample rate being changed by the master using the mailbox interrupt when the master sees a change in the waveform.
I think the #FUSES CPRA & CPRB are just relevant for IO pins to stop the cores fighting - the compiler doesn't complain when used for ADC. Interesting...

AlastairM · Joined: 28 Apr 2008 Posts: 32

dsPIC33CH128MP202 2-core 200MHz lots of goodies
28-pin SOP - ridiculous Razz

temtronic · Posted: Tue Jan 21, 2025 9:15 am

NUTs..200 MHz. need to be careful about PCB layout and cleanliness at that speed

My remote energy control system ran at 24 BAUD. yes 24 bits per second.

sigh....I miss the good old days....

enjoy you're making PROGRESS !!