Printable View
Anyone have issues w/ ESS being too sensitive? I get weird E-Stop errors at different times - but what is consistent is if I manually click to start "Flood" coolant while a program is running it ALWAYS throws a E-Stop error.
Shouldn't I be able to do this without the machine freaking out?
BTW - its a Pulsar Stepper that I'm using.
Thanks!
B
I'm no expert but that sounds like electromagnetic interference. The relay controlling the coolant pump might be a place to look for some shielding.
I have never had mine throw and Estop that was not intended and preformed manually. In fact no switch errors of any kind.
I did have a coolant pump go out, but John sent me a new one. I do start and stop both the spindle and the coolant at times without running Gcode, so I know mine does not fault like that.
Do you have any other electronics or machinery in close proximity?
I think maybe the only real difference between mine and yours is the position I installed the servo drive in.
Mine was not installed when I received it, so I had to install it later. These are older pictures. I do have the newer Arduino BOB installed in mine and the never reference above was after it was installed. I had a few issues with the original BOB.
Lee - I can start and stop the spindle and coolant manually no problem as long as G code isn't running (i.e. manual mode)
Yeah- sounds like you have narrowed it down. If this flood pump is on a contactor then you'll need a contactor (relay) snubber). There should be one on any/all contactors and relays in a CNC environment I've been told. Link below and of course google relay snubber.
Cornell-Dubilier - 104M06QC150 - EMI/RFI Shielding/Suppression - Signal Filters - Allied Electronics
You may want to check into Shielding of wire sets (power sets) or route away from control lines and PCB boards. Create a metal shield divider or surround can (faraday cage) to help.
Grounding- Do you have a clean earth ground for your rig? I used to see a lot of issues related to nothing more than grounding properly.
Just some thoughts. Best of luck, CG.
I have the same problem with my Torus pro. It often throws an estop if I start the washdown pump. And I get Estops for no reason while the machine just sits still or is running.Quote:
Originally Posted by brianbonedoc
Okay.
My Pulsar does estop when I use the wash down pump. I think that may be a safety. You don't want to be able to wash the machine down with it running. You should keep your hands clear when that is happening. :)
Sitting idle and estopping did occur with my Torus, but doesn't on the Pulsar.
I reversed my vfd 180 to get the inductive load wiring away from the Bob wiring. Most my problems went away. Also "zero crossing" solid state relays can reduce EMI significantly. I've never used a coolant pump though.
Thanks for the tips - sounds like it's just a function of the machines unfortunately
I put a bunch of Ferrite cores around different wiring such as the AC input, and the limit switch wiring.
I am still getting random E-Stops that are driving me nuts. The mill can be chugging along no problem, and then BAM, throws an E-Stop!!
The other thing that seems to be happening is that chips can trigger the limit switches, but this is less of an issue than those random e-stops.
What is the Arduino BOB Lee is talking about? Maybe that would solve the issues? I have the same BOB pictured above, Rev 1.
The best location for a ferrite bead (core) is on the ESTOP wire near the BOB. Better yet both wires leading to the ESTOP should go through the same ferrite bead to prevent a glitch on the common signal from making it appear ESTOP triggered. Ferrite beads on AC signals going to anything that draws decent current will be useless as the ferrite bead will saturate on every AC transition. Low amperage AC limit switch inputs might not saturate but will likely not be the issue anyway.
The practice of using a single ground point will clear up much of the noise on signals. Instead of one board / device grounded to another then jumping to another and another in series they all should go to one common location. The use of shielded cable between the ESTOP switch and BOB and grounded at BOB end only will dramatically reduce noise pickup. Using a metal backplane instead of plastic / fiberglass that is grounded behind the BOB will add additional shielding to the system.
Charlie.
Quote:
Originally Posted by brianbonedoc
Quote:
Originally Posted by cteufert
Charlie - thanks for the knowledge. I just put about 10 of them on a bunch of different wires including the limit switch wiring to the BOB. I will take the ones off the AC input as you suggested, and put em on the E-stop wires right at the input to the BOB.
Also - noticed yesterday the E-stop triggered during some heavy roughing. Not sure if theres a feedback loop to the servo driver that threw an error or what.
Again - thanks for the info and I will report back.
Brian
FWIW - Ferrite cores are completely useless in terms of preventing functional problems like you're seeing. They operate at extreme high frequencies - 10s to 100s of MHz - FAR beyond anything the logic can respond to. They are used to reduce EMI emissions only. If you suspect noise on an E-Stop or limit input, add a small capacitor between the signal and GND - 0.01uF or so. Ferrites will do no good at all.
Regards,
Ray L.
Ray - good idea! The problem w/ the random e-stops is they are not consistent and hard to tell if the intervention did anything.Quote:
Originally Posted by SCzEngrgGroup
I will try the capacitor as you suggested.
I had two issues with my estop and ss board. One was I had the estop in a metal duplex box mounted (replaced with a plastic one) on the wall and it was picking up interference in the shop from surrounding equip, and to remove the last little bit of problem ( i put a cap in the estop circuit for filtering).
Davy182 - so did us just put a cap across the 2 e-stop wires at the BOB, or in the e-stop box itself?
Also - what size capacitor?
Thanks!
The best place to put the caps is right at the BOB inputs.Quote:
Originally Posted by brianbonedoc
Regards,
Ray L.
Attachment 282694 Here is what ultimately worked for my situation with a c32 board from cnc4pc.
Ray - I'm not disputing your conclusion, but I was having a devil of a time with very sporadic false limit switch triggers for quite a while with an early Tormach mill until they sent me a ferrite core to install on the parallel port cable between the PC and the controller. In the 5 or 6 years since I've never had another false switch trigger. What did the core fix?Quote:
Originally Posted by SCzEngrgGroup
Mike
Davy - thanks!!
Is that .01uF as Ray suggested?
I had another 'external e-stop' message today while drilling some aluminum.
I'm going to try the capacitor as Davy suggested. Here is a pic of the ferrite cores which didn't seem to prevent the problem.Attachment 282740Attachment 282742
.01 sounds right
Try disconnect the estop cables from the bob completely and see if that solves your problem. If not you probably have problems with something else. For me it does not matter if I have the estop plugged in or not. Still triggers sporadically.Quote:
Originally Posted by brianbonedoc
A cap will work. But if you use Mach3 as the controller you simply set the debounce setting (adjustable) for the exact same effect. It will filter out short pulses.
I use Ethernet Smooth Stepper - so the debounce in Mach doesn't do anything. There is a noise filtering line in the ESS settings, but I haven't had success with that - but maybe I don't know how to set it properly too.Quote:
Originally Posted by hub
Ok, I personally would then use a schmitt trigger (have used many times for sensitive inputs, works great). The cap should also work as it won't discharge fast.Quote:
Originally Posted by brianbonedoc
Schmitt trigger - Wikipedia, the free encyclopedia
I had a similar issue on my ESS but it was with the PROBE signal for my TLO touch plate. Try setting the debounce in the ESS plugin to 50ms for the offending channel and it should go away.
Quote:
Originally Posted by brianbonedoc
Interesting. I had tried a couple ms in ESS but still had the issue. I will try 50msQuote:
Originally Posted by ChrisAttebery
brianbonedoc, did you finally get the random E-stop problem resolved? If so which fix or combination of fixes worked for you?Quote:
Originally Posted by brianbonedoc
CJ
Yeah I switched to pathpilot and a Mesa card and am happy.
Thanks for the input. I'm having similar random e-stop problems.
CJ
Sent from my XT1094 using Tapatalk
Cj - actually I forgot but I was getting swarf on the y axis limit switch that would trigger an estop - even w aluminum! I covered the limit switch w a plastic cover that keeps the swarf a little off the switch and that decreased the estops
Have you tried simply putting small capacitors (0.01-0.1uF is good) on the E-Stop and Home/Limit inputs? That will filter out any noise, with no ill effect on any other aspect of performance.
The inductive limit switches ARE prone to false triggers, even when milling aluminum. Novakon has, at some point in the last year or two, re-designed the limit switch mounts to at least mitigate this problem. On the earlier machines, like mine, those switches are a problem. I removed my X/Y inductive limits very early on due to the false triggers.
Regards,
Ray L.
Ray - good point. I think you suggested I do that as well. That filter cap worked great on my grizzly cnc lathe that was throwing e-stops left and right. I put a scope on the signal and it looked like a v-tach (for those medical folks) before the cap - an afterwards a nice square wave when the input was triggered.
Ray correct me if I am wrong but depending on the input you may have to put the cap across the input and GND VS the input and V+ (depending on if it's a high/low input).
Between input and ground is best, since the home/limit sensors are normally open inputs - they are normally HIGH, and go LOW when a limit is hit. You just need a cap on any one of them, since they're all wire-ORed on the BOB.Quote:
Originally Posted by brianbonedoc
Regards,
Ray L.
I had that problem on my Torus as well.
I just turned them off on it.
I also was plagued with random estops on my Torus pro. The problems mostly originated from the limit switches. I was also getting some from the power draw bar. As a temporary solution I disabled the limit switches in Mach 3 I only used them for homing. This got rid of most of the problem.
The final solution was to replace the VFD. It seems it was the source of the problem. I have not been running the mill a lot lately but I have not had an estop error in months. Going from memory but the original VFD was a Sunfar 550 and the replacement one is a Automation Direct GS2.