Machsupport Forum
Mach Discussion => General Mach Discussion => Topic started by: Tarak on June 03, 2017, 05:36:01 PM
-
Hi All,
I have this issue that happens once in a blue moon:
I believe I have a leadscrew with a tight spot in it. And occasionally when the machine is ramping up speed at this exact point it will tripout my Z axis servo drive (TEK10), but it doesn't stop Mach from running, so my other axis keeps machining, is there a basic setting I haven't enabled to stop the system when a servo drive trips?
Its easy to re-position once it does trip, I just re-home the offending axis and it's good to go again.
BTW, I'm running three ID33003 servo motors with three CNC TekniX TEK10 drives
Thanks,
TaraK
-
Hi,
not familiar with your particular drive but drives I have all have a 'fault' output which can be monitored by Mach and
Estop on fault. Have a close look at your manual to see if indeed your drive is equipped with such an output.
Craig
-
Hi Craig,
Thanks for that, I had a look, and you're right the drive has an output that for when the drive is enabled that goes LOGIC LOW when the drive disables.
I also had a look at the INPUT config in Mach3, but I cannot figure out which INPUT signal I need to connect to.
Could you advise which INPUT signal I should connect the drives to?
TaraK
-
Hi,
I would've thought that if your Z axis servo faults you want to immediately go to Estop to protect the tool and/or work piece.
I don't know how your Estop is wired but if its like mine logic low represents fault/emergency stop. Thus could you hook the output
pin to your Estop circuit. Depending on the current that could be delivered by your existing Estop circuit you may have to protect
or otherwise buffer your pin. Does the manual tell you how much current the pin can sink?
Do you have many inputs to spare? I know that you're talking about a fault on the Z axis but ideally you would monitor both X and Y
as well. I would combine those inputs to one 'servo fault' signal and interface that one signal with your Estop. The reason I'd adopt
that strategy is so that a servo drive which is yet to be enabled at power-up does not prevent you from enabling Mach. At power up
you need to enable Mach but if a disabled servo prevents you and you can't enable the servo until Mach is enabled you can't do anything!
Craig
-
Hi Craig,
The datasheet doesn't mention anything about the voltage from the RUN OUTPUT (Pin 8 on the IDC10).
I've attached the datasheet for the drive.
I thinkI have the Estop settings as Port 1 Pin 10 with ACTIVE LOW enabled.
TaraK
-
Hi,
I can't see any spec for it either. I do see that the upper board logic supply needs be capable of 0.1A. Clearly the outputs are logic level,
I wouldn't consider asking an output to sink more than 5mA, the risk of blowing it up is too high.
The Estop terminal on my BoB is logic high by virtue of 5V being applied through the series of normally closed contacts of my limit switches
to the Estop terminal with a 5k resistor from the same terminal to earth. Thus with all contacts closed 5V is applied directly to the pin, when
one or more contacts are open the 5V continuity is lost and the 5k resistor pulls the pin logic low and Estops Mach. If I were to hook the output
pin of a servo drive such as your TK 10 to the Estop in addition to the limit switch circuit it would fail. The 5V of the limit switch circuit would
attempt to hold servo drive pin high even when the pin was trying to go logic low, it would try to sink a lot of current and fail and damage the drive.
You'll need to introduce some resistance in the 5V supply to your limit switch string. If you put 5K in there with the limits all closed logic high would
only be 2.5V by virtue of the voltage divider formed by the 5k resistance to earth at the Estop pin, such a low logic high could render your Estop
noise sensitive. If you reduce the series resistance to 1k then the current to be sunk by the servo pin would be 5mA and yet have a logic high of
4V, acceptable. I would in addition put a diode in series with the servo pin so that if in fact its voltage goes higher, 12V say, it cannot blow your
Estop pin of you BoB if as in my case your BoB is a 5V device.
Craig
-
Hi,
hopefully this makes a bit more sense.
Craig
-
Hi,
after talking about putting in a diode to be on the safe side I bloody well forgot to put it in there!
Craig
-
I always wonder what would happen if a motor fails during a run.
There are 2 scenarios here: one is if the motor fails and the other is if the motor driver fails and as far as I know there is nowhere on Mach3 to notify what to do if any of these fail.
The same goes with the spindle: if the spindle stops for any mechanical / electrical reason Mach3 will keep run until the end of the program.
What Craig is mentioned here is very interesting
-
I'd connect the VFD fault input using simple isolating relay if your drive has a solid state output instead of a dry contact relay. Then you know your E-stop won't cause you intermittent grief. On my machine I have the limits switches all in series with the mechanicly held E-stop buttons, the Mach 3 E-stop output through a relay and a reset button that closes across all of the E-stop circuit momentarily so Mach 3 can be reset and allow you to jog away from a limit or out of a part while holding the reset button in. The E-stop circuit holds in a power contactor which removes all power from everything on the machine, spindle drives and axis drives.
-
Very good thinking geryhlucas, I will have to think how to implement this idea into my setup
Thanks
-
Hi All,
garylucas has a great idea. The circuit as I've drawn it may well give Tarak problems when he tries to enable Mach at startup.
A Reset button like garylucas mentions would solve it.
I had in mind a software setup where the drive/drives were not monitored for faults until either Mach was enabled or was actually
machining, a complicated way of doing what garylucas suggests.
Craig
-
Thanks Craig/GaryhLucas, you've certainly given a bit to think about, I'll certainly look into it.
TaraK
-
Simple is good if it can be done. A hardware solution like this is a lot more bulletproof. No software problem can override the limits.
-
Hi All,
I had problems with my spindle stalling. Its a 24000 rpm VFD driven unit and I was trying to use it to drive 6mm endmills cutting
steel. The problem is that high speed spindles have so little torque. Consequently you could only take the lightest of cuts and if
you over did it or your Gcode called for a plunge or similar temporary overload the spindle would stall. With the tool stalled but still
engaged in the material the XY movement would snap the tool no problems, my wee mill has near 500kg of thrust, 6mm endmills
don't stand a chance.
For this reason I monitor the VFD for excessive difference between commanded speed and actual speed to Estop in the event of a stall.
It works mostly, the VFD can take some milliseconds to signal a stall and so an Estop may occur too late to save the tool.
I have since made a servo driven spindle with 6.5Nm rated torque and solved the whole problem of stalls. Now I have enuf torque
to 'twist' smallish (less than 8mm) endmills off like a carrot!
Craig
-
I probably should've also mentioned I am also using a smoothstepper with the MB02 breakout board (Homann Designs).
I've attached the user manual for my breakout board, just in case you were interested.
For the most part, it's been a bullet proof system for quite a few years now, it's just this issue of tripping one drive, but continuing to machine, that rears it's ugly head every once in a while.
TaraK
-
Hi Tarak,
I use MB02's as well, I'm guessing you must be an Aussie whereas I'm Kiwi, either way Homan Designs is 'local'. In fact that's how I came
to suggest putting a diode in the circuit, I've blown the input buffer for just one pin on one of my MB02s. MB02s DO NOT accommodate
being reverse biased....nor are you doing yourself any favours by demanding that IOs either deliver or sink a great deal of current.
Depending on the actual IO MB02s are good for 24-32mA source and about the same sink. I personally like to keep it to 5mA or less. Having said that
I don't like to overload IO neither do I like it to loaf. In the example above I used a 5K resistor as pull-down, ie 1mA. I could have used a much
higher resistance, say 100k but the pull-down current would only be 50uA and therefore be subject to noise.
As garyhlucas has said the robustness of relays and the moderate voltages and currents used in such designs can prove to be a much better
performer in noisy environments compared to micro-power low voltage signals.
Craig
-
Hi Craig, yup sure am an Aussie.
I've had a good amount of success with the MB02, I only found that it was effected by noise a fair bit, until I set the debounce to 5000, then voila, all good!
TaraK