G540 Fault

[matttay]

Hi, sometimes in the middle of machining the Mach3 will stop due to an external e-stop.

I have input 1 and 2 configured as X and Y limits respectively. Those work, and it correctly reports in mach3 that a limit has been reached when tripped.

On the 540, Pin 10 is hard wired to pin 12. So I'd not think e-stop should be possible.

Is there anything else that can make the G540 signal e-stop? Overcurrent? Too hot? Anything?

Thanks

[mif73]

Hi, I had this fun with mach3 as well, it is more likely to be noise causing a false trigger of the e stop line into the pc, rather than an e stop signal from the drive unit itself.

One option is to put a capacitor between the common and e stop pin going into the pc (about 0.1uf I think), the other is to put a dwell in the Mach3 config screen (I think its called interval debounce?), try setting a figure higher than the default of 0, I have a figure of 3000 to get rid of false external e stops with a cap across the e stop input pins too.
I hope that helps you, good luck.

Cheers, Andrew.

[Nick1983]

I have had the same problem with Mach as well. As far as my experience goes it was the noise (debounce interval). Mine was set at like 7000 from the manufactor, I had to actually cut it in half to something like 3500.

[matttay]

Does the G540 ever decide to assert estop on its own? Several of the times it has faulted it was under very heavy load as I was trying to determine cutting rates and depth of cuts.

So it seems that G540 definitely can assert estop. Overcurrent? Overtemp? Is it possible?

[dmauch]

I have built about 35 systems that use the G540. I test every system for a couple hours and have found a few things that could be the cause fo the problem.
1. DB9 cables. Db9 cables are not all created equally. I have found some chinese cables to get very warm and cause a fault.
2. I have run into an occasional chinese stepper motor that had intermittent shorts. Very rare.

Dan
dmauch@camtronics-cnc.com
low cost cases for G540's and other gecko products
www.camtronics-cnc.com

I now only use cable that don't get warm and while I have various suppliers I try to only buy from the one that has the two ferrite beads near the connectors.

Originally Posted by matttay Hi, sometimes in the middle of machining the Mach3 will stop due to an external e-stop. I have input 1 and 2 configured as X and Y limits respectively. Those work, and it correctly reports in mach3 that a limit has been reached when tripped. On the 540, Pin 10 is hard wired to pin 12. So I'd not think e-stop should be possible. Is there anything else that can make the G540 signal e-stop? Overcurrent? Too hot? Anything? Thanks

[Dan Craig]

I had "limit switches" (quotation marks used because no actual pin was showing as active) tripping on my mach 3 / G540 setup and I knew it was the router causing the interference because it only happened when I had the router on...turned out to be a bad connection to one of the brushes.

Can you narrow the fault down a little more on your machine?

[matttay]

But mach3 can tell if it was limit switch or estop, right?

If estop is tied low, how in the world can estop still occur? I will confirm tonight that the connection is solid.

However, my #1 question right now is whether or not the G540 will assert estop for overcurrent, overtemp, etc. What safety mechanisms does the '540 have?

If the answer is "none", then I'll focus on the mechanical connection of estop. If the answer is "overcurrent and/or overtemp" then I'll address those separately.

[Dan Craig]

I haven't read of any over current or over temp protections. The info I've always seen says don't exceed Xvdc or your burn it up, keep it cooler than warmer and you won't overheat it and burn it up...so I'd take those to mean there are no safety mechanisms for over current or temp.

Like I said - I was getting limit switches (acting like e-stops) which were phantoms and it was noise.

[matttay]

I'm a bit loathe to accept the noise theory, because the estop signal is tied to ground. Assuming it is correctly tied to ground, that signal cannot be corrupted by noise because it's too low of an impedance.

Is it possible that a momentary charge pump failure could cause the G540 to fault, and then a true fault condition would be propagated back to Mach3, which would then indeed turn off the CP and allow the fault to persist?

<sigh>I'm probably going to have to stick a scope on this...

One other point, I'm using a SmoothStepper USB board. It hasn't been a problem so far, but the scope will start answering some of these questions.

[Dan Craig]

good luck! report back with the findings so we can all learn from your experience.