Welcome to LinuxCNC.org

My computer does not have a serial port, so I bought a cheap PCI serial card and then a RS232 - RS485 converter. It was working fine, except sometimes the spindle stops for no reason. Or spins the wrong direction. It is commanded correctly - but the VFD shows it as being commanded to stop I do not really know how to troubleshoot modbus stuff so I lived with it and just assumed it was a combination of the cheap PCI card, adapter, and me being a novice at this. Now I have a job where I need the machine to run "lights out" and cannot have the spindle randomly stopping. I thought the mesa system would work, but looks like I did not do enough research. I am thinking of buying a better quality RS485 PCI card to eliminate the adapter to see if that helps. I also have a live tooling motor that I would like to control the same way. (Not sure if you can load the GS2 driver twice though...) (My controller is actually a GS3, but I think it works the same as far as the communication is concerned).

If that does not work, I will probably try using PWM.

brianm wrote:That does sound less than convenient.I would say there was a fair chance that the Mesa kit could be made to work.
What hardware do you have? Perhaps we can think of ways to link it all together.

I have 2 Mesa Cards

first is the 5i20 with a 7i33 daughter card. (the daughter card runs the steppers and handles the encoder feedback from the spindle) the rest are gpio

Now I have a 5i25 and a 7i74 daughter card.

The VFD is a GS3-23PO

brianm wrote:That seems an unusual choice for a stepper system, in some ways. I assume you are using the 7i33 for analogue voltage generation and encoder feedback.
The natural fit would be the 5i25 / 7i76 or 7i78 (the 7i48 has no GPIO thought) but it makes absolutely no sense to buy _more_ hardware when you already have a hardware surplus and a functional machine.
Looking at the GS2 component it would be a natural fit to the Mesa UART component, but I have more than enough irons in the fire, so it wouldn't be me doing it. I could advise if someone else fancies picking it up.

I went with the 7i33 because I had planed to change to servos. I actually have the motors and amplifiers on order and plan to change it next week.

brianm wrote:OK. The best plan might be to keep the 5i25 around for a future machine then. that 5i25/7i44 combo can connect a huge amount of IO. (8 x 48 channels) Perhaps there is some home-automation project in your future.

Next time maybe ask here, or ask Mesa, before you buy hardware?

It defiantly will get used in the future.

These things have a way of not working for one project but being perfect for another.... Next time, I will be posting if i am the least bit unsure.

Thanks for the help

where are you located ?
their are a few differences between the GS2 and GS3 , mostly to do with the pid values and a few extra registers , i'll take a closer look , but unfortunatly i dont have a GS3 to hand to test
can you post your modbus register list and your hal file , i'll take a look . and see if i can see anything

Just comparing the GS2 modbus communications to the GS3 and the communication protocol is the same and most other things look similar. I would be surprised if the GS2 component doesn't work with the GS3. If it needs a tweak that should be easy to do.

John

yes i'd agree John , they look just about identical , apart from the pid registers are not used in the GS2 , and the added encoder of course , i'm wondering if somehow it's trying to access say one of the pid registers and because of an error returned it's crashing out ... hence the erratic movement or stop .

it could of course be a simple matter of grounding or earth loop tripping the modbus , or bad rs232 etc , if the supply is not clean ,
( are the serial cables sheilded all the way etc ) i must admit i hate using convertors or multiples of

so question is , is it totaly happening at random , or is their a repeatable sequence triggering this event