[235r]

G'day all, I've been trying to avoid this but, I'm at a loss. It all began with a move form Sydney to Perth. Everything including valve radios made the trip safe but....not the CNC. The first time I hooked it up it threw the house circuit breaker. Eventually I realised it was the AC fan on the control box, so I bought a new fan only to get the same result. I disconnected the AC fan & everything was good, for awhile. Then it started to stop mid run for no reason. Now, the machine is only one of those cheap Chinese pieces of junk from eBay but, I've had it for 10 years & it's evolved. A homebrew control box based on a Gecko G540, even though I purchased a 48V psu, I decided to stay with a 24V system. A USS USB smoothstepper was added. I've replaced the motors & all the cables & added a ground system (not in that order). This all worked fine for a couple of years before leaving Sydney. There's no noise issue as the machine has moments of clarity & works as expected. It turned out the old laptop running on Win7 has a HDD problem & kept crashing. At first I thought that was being caused by the CNC but, no. I've changed laptops & it now run on Win10, so it's a clean install with all the recommended drivers. I recently upgraded from VCarve pro 9 to 9.5 & the CNC runs on Mach3 (v3.043.062). I thought it was a ground cable problem so I checked them all & cleaned the contacts with a positive result, again for awhile. I've pulled the control box apart & rebuilt it. I've done everything I can think of, I've checked previous posts but, the problem continues. It seems to be a software issue rather than a hardware one as I have been able to complete some small ABS signs. But, a new setup next job & it goes 10 seconds & stops. If it doesn't stop on the same line, it stops a few lines later. I've run it for 10 minutes before it stopped & out of curiosity run it again just to have it stop in the same place. I've gone back to VCarve & recalculated the job only to have it stop in more or less the same spot. The only thing I haven't done is reinstall VCarve but, that's next. Any suggestions are much appreciated. Thanks, Rob

[Sterob]

What VFD are you using?
Are all your stepper signal and control cables screened and earthed to the same point?
What Spindle do you have?
I it earthed and the cable screened and earthed?

Some don't like using Laptops for CNC machines ( me included ) Do you have a reasonable spec PC to test with?
USB can be susceptable to EMI. screened USB cables are available, I think ( no experience , I use ESS)

Nothing is overheating?
I have one fan blowing in and another sucking out to ensure good airflow in my case.

[235r]

What VFD are you using?
Are all your stepper signal and control cables screened and earthed to the same point?
What Spindle do you have?
I it earthed and the cable screened and earthed?

Some don't like using Laptops for CNC machines ( me included ) Do you have a reasonable spec PC to test with?
USB can be susceptable to EMI. screened USB cables are available, I think ( no experience , I use ESS)

Nothing is overheating?
I have one fan blowing in and another sucking out to ensure good airflow in my case.

Thanks for the reply. I've been using the same 1.5 kW inverter, a HY (Huan Yang) for 10 years. No to the screened control cables but yes they all grounded to a star point & they were fine before the move, nothing has been added to change the situation. The spindle is only a cheap one Chansheng 800W but I've been using it for again a couple of years & yes it's grounded. The laptop is an i5 2.4 with 4 Gig running on 10 so, yeah. Screened ground cables are homebrew but they date back to 2016. In the back of my mind I can't help thinking, what was the start of all this. It all started with the AC fan on the control box tripping the breaker & after I changed the fan I still got the same result, what caused a setup that did work to stop working?. I built this control box back in 2016, I've done my research & built a level appropriate box. I run a small set of fans mounted on heat sinks attached to the G540 & until recently an AC fan for after shut down to keep airflow threw the box (the AC fan receives power directly from the AC switch on the rear) & no the issue occurs at anytime, hot or cold. Thanks Attachment 429894

[Sterob]

Maybe check all your terminations for tightness. Something may have come loose on the way over.
The fan on the Power supply working? Sometimes they jam and the power supply overheats and shutsdown.

[235r]

Maybe check all your terminations for tightness. Something may have come loose on the way over.
The fan on the Power supply working? Sometimes they jam and the power supply overheats and shutsdown.

Good onya Sterob for trying, nobody else is lol. As a matter of fact, I'm surprised the usual suspects aren't having a go...guess it's over their heads? Nah mate, all tight. One of the first things I did was rebuild the control box after the move to make sure that wasn't the problem & there's no shutdown, it's only Mach that gets tripped & stops the machine. I've done minor changes like slow the cut down & all of a sudden it works fine & finishes the job (?). Setup a new job at the same speed & it's back to square one (???). I'll keep looking, thanks again

[Sterob]

No problem.....just trying to help out.
Can you run a job with the Spindle turned off? ( air cutting ) See if that makes any difference.
You don't get any error codes?

[ericks]

How does Mach trip? Does it throw an estop?

[235r]

Thanks again Sterob, spindle on or spindle off, same result. Water pump on, water pump off same result, & I don't use a vacuum.

[235r]

ericks thanks for the reply, yep false estop. I say false because it's come good at times & completed short runs at least (8 mins) that wouldn't work minutes before, then you set it up again & it's back to square one (?). I've done the same run twice & both times it stopped after 10 minutes in the same place & more or less the same line. It's confusing as hell, I think I'll start replacing cables next & consider maybe replacing the breakout boards after that. Thanks

[ericks]

Well then it can be either electrical noise, loose connection or high resistance across the contact of the e/stop switch. If you switching 5v through an e/stop contact then any bit of high resistance across the contact can cause you issues. Personally i dont switch 5v on external devices like e/stops etc. Best to switch higher voltages like 12V or even better 24V.

[ericks]

Before you start replacing cables...disable the e/stop in mach and run it to see if the problem is still there. Just be aware that for that time you won't have a functioning e/stop. If the problem goes away then we can advise you further

[Sterob]

Before you start replacing cables...disable the e/stop in mach and run it to see if the problem is still there. Just be aware that for that time you won't have a functioning e/stop. If the problem goes away then we can advise you further

Agreed....Short out the E/stop at the Break Out Board and try again. Good pick up Erick.

Also agree about the 5V thing, but you don't always get a choice.....

[235r]

Thanks again ericks & sterob for the input. Ok, at this point I'm calling it problem solved. I'm not sure if it's the correct way to do it though? Steps were: Open Mach - Config - Ports & Pins - input signals - scroll down to Estop & uncheck "Enable".
Only one problem...it won't let me disable it? So, a quick search to make sure that I was doing the correct procedure & tried again, with the same result. I guess because it's a custom config file for the USB SS I can't change it? So, I looked for another way to do it & just guessing I tried - Plugin Control - USB Smoothstepper v17e Config - Noise Filtering section (right of page) - Estop - change value 0.00 to 1.43 (because I didn't know what else to enter? & that's what it mentions in "Noise filtering"). Now, this doesn't disable the Estop in Mach & as a matter of fact I'm not exactly sure what's it's doing? but, it did work. I guess it makes Mach ignore small variations in voltage uless they last longer than 1.43 ms? I don't know, maybe somebody else will? Anyway, I was able to complete 3 short test runs (with all devices on), the longest of the 3 runs was 7 minutes & not one false Estop with the external Estop still functioning as normal. I've now completed a couple of small jobs, around 20 minutes each, without any false Estops. Thanks guys

[ericks]

Glad you got it sorted, well done!

[Sterob]

Thats great news. Maybe a google on that particular on the USS Plugin setting might give you the info.
I don't have the USS, I use the ESS, but even so I have not seen that setting. I should go and check it out.

As long as you still have the External E/Stop that is good.


Erick gets the Chocolate Frog!

[RCaffin]

I suspect you had a noisy eStop line. This can be fixed.

Can I assume that the trigger for the eStop is a mechanical switch? A common problem here is that the contacts may be rated for 240 VAC 10 A, which they handle easily, but when you used them for 5 VDC 2 mA, they can have problems with oxide on the surface of the contacts. Same applies for 12 VDC 5 mA. The solution is to run the switch at a higher current - say 50 mA at least.

Once upon a time we could get two sorts of micro-switches. There was the type rated for 240 VAC, for industrial control of power, but there was also another type with gold cross-bar contacts. These were very definitely for low level signals - logic signals. Using them for power destroyed them. Can one still buy them? I am not sure, but they do solve this problem.

Edit: ah yes: Honeywell, Omron &others; RS Components, Newark, others maybe? But beware of anything which is also rated for AC voltages - 110, 240,or amperes of current. They do NOT have the gold contacts.

Cheers
Roger

[Sterob]

I suspect you had a noisy eStop line. This can be fixed.

Can I assume that the trigger for the eStop is a mechanical switch? A common problem here is that the contacts may be rated for 240 VAC 10 A, which they handle easily, but when you used them for 5 VDC 2 mA, they can have problems with oxide on the surface of the contacts. Same applies for 12 VDC 5 mA. The solution is to run the switch at a higher current - say 50 mA at least.

Once upon a time we could get two sorts of micro-switches. There was the type rated for 240 VAC, for industrial control of power, but there was also another type with gold cross-bar contacts. These were very definitely for low level signals - logic signals. Using them for power destroyed them. Can one still buy them? I am not sure, but they do solve this problem.

Edit: ah yes: Honeywell, Omron &others; RS Components, Newark, others maybe? But beware of anything which is also rated for AC voltages - 110, 240,or amperes of current. They do NOT have the gold contacts.

Cheers
Roger

Good pickup Roger. I had not thought of that, because my machine uses 24VDC for the I/O, but they are valid comments.

[RCaffin]

'Some assembly may be required ...'

You would need to adapt the microswitch to the mechanical part of the common (twist to release) eStop button.
But that might be the fun part!

Cheers
Roger
PS: yes, I have had the same problem myself!

[mactec54]

Thanks again ericks & sterob for the input. Ok, at this point I'm calling it problem solved. I'm not sure if it's the correct way to do it though? Steps were: Open Mach - Config - Ports & Pins - input signals - scroll down to Estop & uncheck "Enable".
Only one problem...it won't let me disable it? So, a quick search to make sure that I was doing the correct procedure & tried again, with the same result. I guess because it's a custom config file for the USB SS I can't change it? So, I looked for another way to do it & just guessing I tried - Plugin Control - USB Smoothstepper v17e Config - Noise Filtering section (right of page) - Estop - change value 0.00 to 1.43 (because I didn't know what else to enter? & that's what it mentions in "Noise filtering"). Now, this doesn't disable the Estop in Mach & as a matter of fact I'm not exactly sure what's it's doing? but, it did work. I guess it makes Mach ignore small variations in voltage uless they last longer than 1.43 ms? I don't know, maybe somebody else will? Anyway, I was able to complete 3 short test runs (with all devices on), the longest of the 3 runs was 7 minutes & not one false Estop with the external Estop still functioning as normal. I've now completed a couple of small jobs, around 20 minutes each, without any false Estops. Thanks guys

Most likely when you reinstalled Mah3, you lost any debounce settings you may have had, this is where you can set the Debounce to help filter out any small noise problems

You find the Debounce in the General Logic Configuration page right hand top corner

[mactec54]

'Some assembly may be required ...'

You would need to adapt the microswitch to the mechanical part of the common (twist to release) eStop button.
But that might be the fun part!

Cheers
Roger
PS: yes, I have had the same problem myself!

I doubt that his EStop circuit is 5v they are mostly all 24v on these machines unless he changed it