Machsupport Forum
Mach Discussion => General Mach Discussion => Topic started by: KE8MIR on March 18, 2023, 07:49:02 PM
-
what would make a machine loose steps
using mach 3 with cnc4pc c62 break out board and teknic cpm-sdsk-2310S-RLN motors
set up machine and all was good for about 3 months.
then one morning the first cutout the ending cut did not line up with where it started
first thought motor was going bad swapped out for new motor ran motor setup per teknic
still off swapped new controll wires still no good
the longer the g-code file the more it is off
this is only happening on the X axis Y axis is good
did the axis calibration per mach
anybody have any ideas what to check next
thanks in advance
KE8MIR
-
You might try to go into motor tuning and reduce the acceleration/deceleration parameters but I don't think that is your problem. Check for a mechanical bind which can be caused by lack of lubrication, bearing failure or contamination. Barring that, I am not sure what your issue could be but I don't believe it is a Mach problem.
Ed
-
Is there a way tell if the motor is pulling full load?
this is on a plasma cutter it should have on load on the motor other than the x axis rail
this machine is only 3 months old
-
Unfortunately I don't know of any way to monitor the motor load with Mach or any other stepper driver/control set up. But that doesn't mean one doesn't exist, I am just not aware if there is. AC and DC servo setups often, but not always, have the ability to monitor servo motor loads. I would take the X axis motor off and try to turn the lead screw/ball screw by hand to feel for any drag, roughness or bind along the entire travel window. Is the lead screw and axis guides or linear ways protected from the environment? If not, regular cleaning and lubrication is in order. With that said, being 3 months old does not mean that you don't have an issue that might still be under warranty.
Ed
-
Hi,
Clearpath servos can monitor servo load but relies on PWM of the HLFB output of the servo. Enquire of Teknics......the trick then is to have Mach interpret that
signal.
Other servos have separate channel for analog output and its just a matter of hooking a voltmeter to that output to monitor servo loads, but not so Clearpath.
Craig
-
Ah, I didn't realize his motors are servos. My question then is why doesn't the system shut down for a following error? Where is the encoder signal going? I have an integrated stepper servo set up and the encoders report to the stepper drivers. When there is a following error the offending axis drive shuts down and sends a fault signal to my control board setup. Mach does not enter into the picture at all and has no concept as to why the fault occurred; it only knows that the C94/UC300ETH control setup is not happy.
However, this does not answer why the axis is loosing steps.
Ed
-
Hi,
Clearpath servos have one digital output, and its programmable. The most common use is that it be a combined 'Servo Fault' indicator that would combine various overload conditions
and Following Error. Has OP bothered to program the HLFB? Has he bothered to wire it back to the breakout board? Has he botherd to program the motion controller to
respond to a HLFB signalled fault?
Clearpath encoders are built-in and go direct to the control electronics in the servo.....they do not ever get presented to the outside world, including Mach4.
However, this does not answer why the axis is loosing steps.
There are a number of possibilities including electrical noise. Noise could potentially be in effect signalling extra step are required, so Mach issues 1000 Steps but electrical noise adds an additional 10.
Another possibility is that the servos are being overloaded or more likely being demanded of greater accelerations and/or velocities than its capability. That would be commonly detected as a
Following Error, but then maybe OP has not programmed it so.
Craig
-
It just occurred to me that the system may not be loosing steps but rather the positioning error is happening as the result of a loose coupling on the ball screw. Is the motor direct coupled to the lead screw or is there a pulley involved that might be slipping? If the system was working for three months without issues, what changed?
All good points Craig. I would think that electrical noise interference is an equal opportunity problem that would affect all axis' but with that said, are the encoder wires shielded and grounded appropriately?
Ed
-
wow that is a lot of info we ran a spare cable from the bob to the motor
there was no change in operations still loosing steps
this machine has a rack and pinion drive system on the x axis
and has four bearings that ride on a angle iron and are not protected from the machine
i am thinking that a bearing going bad
it is unfortunate that the builder of the machine went bankrupt the day after christmas
i am going to pull the bearings off tomorrow and see if the local supply house can track down
replacements will report back
thanks for all the feedback
-
Hi,
I don't think that is the right move. I think just throwing parts at the machine without trying to find a cause is a good way to waste
time and money.
If the machine is loosing steps then it should fault out 'Following Error'.
Do you have the HLFB programmed to signal Following Error? Do you have your motion control board ready to accept and act on a Following Error?
If the machine is loosing steps Mach should know about it.
Craig
-
His breakout board uses Rj45 connections to the servo drives so he is definitely wired correctly. If a servo drive produces a fault, would it not shut down regardless if Mach is made aware? I'm just wondering out loud here. Even if I had not wired the fault signal output from the servo drives to the breakout board, the servo drive shuts down when the following error is exceeded. My servos will not run without an encoder feedback but is it possible that there is a parameter in the HLFB that turns encoder feedback off and on?
He could still disconnect the motor and push the axis back and forth by hand to pick up on any noise, roughness or drag just to eliminate that as an issue. In my opinion, V roller bearings on V tracks and rack and pinion drives are not on my list of good ideas. The up side is that both are usually adjustable to compensate for wear and backlash. The down side is that, you need to do that occasionally. Also, a little Vactra 2 goes a long way toward making things work smoother longer.
Ed
-
Hi,
If a servo drive produces a fault, would it not shut down regardless if Mach is made aware?
Not necessarily, and overload fault I would expect to stop the servo every time but a Following Error fault, who knows?. It would stop if that whats its programmed to do.
OP does not seem inclined to answer questions about how the servos are programmed or how the HLFB is connected to Mach4 if its connected at all. Not much that I can do.
Craig
-
Hi,
how about this for a test. Deliberately stall the servo and then command a movement. As it cannot is should produce an overload error (due to peak current) or
a Following Error, in either case Mach should see it. If Mach does not see it, then either the HLFB is not programmed, of it it is, then its not correctly wired back to the breakout
board or the motion control is not set up to monitor it.
Craig
-
I just went through the Clear Path user manual and a following error, Tracking Error Limit as they call it, does shut the drive down when activated. There is no capability to disregard the encoder input or, oddly enough, the ability to change the tracking error window (whatever it is). So, if the system really is losing steps, it SHOULD shut down hard. Given that, I think it safe to say that the HLFB is not seeing a following error; the motor is moving exactly as it is commanded as far as the HLFB is concerned.
Ed
-
Hi,
then try the experiment and deliberately stall the servo and see whether it faults out.
Craig
-
not sure that the c62 board can be programed for the HLFB
-
not even sure what the HLFB is
-
The HLFB, as I understand it, is the stepper driver. I don't think there is any programming that can be done with the C62 bob as it is only a communication and signal processing board. What Craig is wanting you to do is to force a fault on the driver by deliberately overloading the axis motor to see if the fault shows up in Mach. I am guessing it will but I can understand your trepidation at deliberately causing your motor to overload. Don't worry about that as the system is designed to protect its self. Another trick might be to swap drives to see if the problem follows the driver. If so, then you have a faulty drive (HLFB). Just swap the X and Y Rj45 outputs on the C62, swap the motor and encoder terminal plugs and program a simple circular motion and see if the lost steps goes to the Y axis. If it does, then you have an issue with the HLFB.
Ed
-
Hi,
The HLFB, as I understand it, is the stepper driver.
It is not a stepper driver, its a servo. High Level Feed Back is a signal that is programmed into the servo, usually to signal a servo fault. Has OP programmed the servo?
What Craig is wanting you to do is to force a fault on the driver by deliberately overloading the axis motor to see if the fault shows up in Mach. I am guessing it will but I can understand your trepidation at deliberately causing your motor to overload. Don't worry about that as the system is designed to protect its self.
Correct you can stall them all day long without damage.
Craig
-
as i understand it the c62 board has 6 RJ45 output jacks if the X axis output is faulty
could you switch to a spare output jack
-
Hi,
how about just answering the question....have you programmed the HLFB? If so what is it programmed for?
Start at the beginning, then worry about the RJ45 cabling. Just changing things without any sort of plan is going to waste a lot of time and money,
start at the beginning and work your way through it....and it all starts with the HLFB.
Craig
-
I don't know just trying to help out a friend he bought this machine
was not consulted when he bought the machine
I have not did anything to this machine except to try a new cable on the X axis
there is no documents with this machine
I am just trying to help out a person in a bind
-
Hi,
well then you'll have to download the Clearpath manual, its freely available, and install the set-up and tuning software on a PC, and
at the very least interrogate the firmware to establish how its been programmed. It may well be that the HLFB output has never been programmed
and that would certainly explain part of your issue. You can also interrogate the servo with the PC and find out things like Following Error.
Your friend has no doubt paid a premium to get a machine with servos.....now is the time to dig into that capability.
Craig