Vic's comms errors

It has always seemed unwise to me that the RS232 comms are converted to TTL levels to be sent around the ring of PICs. TTL only has guaranteed noise immunity of 0.4V. CMOS is a bit better, but you have heavy currents sharing the same ground connection from the stepper motors and broad-spectrum noise from the DC gear motors. I always remember an EMC expert telling me that motors with commutators emit noise from DC to daylight when we had a problem with a PIC and a motor!

I have experience of driving both types of motor over serial links and generally find that using RS485 differential comms sorts out the problem of ground bounce. The drivers are a bit pricey but you only need one serial port per device because it is bus rather than a ring. You could probably dispense with the sync line because you can broadcast a message to all nodes and have it received simultaneously.

Vic's current problem may be nothing to do with this of course but it is something to bear in mind when designing motor drives.

Well I think that the whole point of the comms board is to get the circuit up to rs232 levels so that there won't be the problem you just described.

---

Wealth without money..
Coding without effort..
Ok, the second one's a lie..

As I understand it the comms board takes the RS232 levels from the PC down to TTL levels to send around the ring of PICs. As the PIC boards are distributed around the machine their ground leads must drop some voltage due to the motor currents flowing down the same path. This voltage gets added or subracted from the comms levels between the boards.

Sorry, Mea Culpa. I'm not using the Max 232 chip, I'm using a simpler circuit and just chopping off the negative swing and inverting the signal logic so between 0 and the somewhere near the Pic VDD of +12. I was wrong I just got confused. That's my story and I'm sticking with it.