Triac PC (1994) - random operations and spindle stops

[el$syd]

Hi,
- My Triac PC using Eurostep RS232 card (Red LED) and VR Milling 2.31 will occasionally do some random movement, usually after milling for some time. It gets bored and decides that it will try a diagonal cut for about 50mm. This has ruined a couple of work pieces already
- Sometimes also, the spindle will just stop mid job - as though an M5 command had been issued. Restarting the program cures that - but in a long program this is very time consuming.
- I read elsewhere on the support BB, that a potential cure is to upgrade to a USB Eurostep card and VR Milling 5.2+ are these the only cures for these 2 problems?
- Is this an RS232 handshaking issue? If so, at which end is it most likely to be? If it is a the Black Box end, could I drive the RS 232 there via a USB (from computer) to RS232 (at black box)?

Any help on this gratefully appreciated

[Martin]

What version is the Eprom on the Motion Control card?

You need to be on Version 2.1 or the Beta version.
It's fitted to the card on the left in the rack. Just behind the red 7 segment LED

[el$syd]

Thanks For your reply.
I upgraded a few months ago to MINT EPROM version 2.1.

[Denford Admin]

Version 2.1 of the EPROM had better error detection on the RS232 comms, and got rid of this kind of random happenings...

So, I can only guess that the RS232 cable has got serious problems with noise, such that invalid data is still getting through, but the checksum occasionally matches the invalid data.
Try making up the cable here: viewtopic.php?t=544 with decent quality, twisted, shielded pairs.
And going direct from PC to the Baldor card

We did have some machines which didn't have handshaking (pins 4 and 5) connected on the internal cable (from cabinet to Baldor) and this caused issues.

[el$syd]

Thanks for your answer.
I believe that the cable is an original Denford, and I will look to see whether the handshaking has been implemented. And also, I will make up a heavy duty cable, with lots of earthing to remove any interfering noise...
Do you think twisted pairs with one of the pairs, for each of the signals, grounded would help?
If that is the problem it should be curable!
The error only happens very rarely - however it has ruined a couple of pieces, and has left me wondering whether I can do larger pieces that would take several hours to complete?

[Denford Admin]

Being rs232 twisted pairs will not provide the benefit of noise cancelling (twisted pairs help with noise when you have a signal and a NOT signal scenario as with encoders, eg: A and not A).
I mentioned it because twisted pair cable is usually much better quality than standard multi-core, and incorporates individual pair shielding foil as well as overall shield and earth drain wires.

[el$syd]

Thanks - OK, understood. A differential A+, A- is for twisted pairs and the drivers/receivers are not differential - so there would not be a benefit in putting A+ as signal and A- as ground for a twisted pair.
I will get an appropriate cable made up, using the shield as ground.