Machsupport Forum
Mach Discussion => General Mach Discussion => Topic started by: fast89 on April 22, 2019, 01:23:56 PM
-
SO. i'm finally messing with an external motion controller. I've tried smoothstepper and didn't care for it. Now i have a uc400eth motion controller on the machine. The problem i have is software limits now cause the system to estop. Which of course de powers the servos and position is lost. Parallel ports never had this problem. Is this just the way it is? Because there really isn't alot of settings to be changed with the motion controller or am i missing something?
-
Send an email to CNC Drive describing the problem.
-
Hi,
you can turn soft limits off.
Have you homed (referenced) your machine, not just zeroed but referenced?. If your machine is not referenced then
soft limits make no sense and are counter productive and need to be switched off.
Craig
-
Machine always gets referenced but it is done outside of mach3 within the AC servo drives. I've used soft limits for more than a decade with parallel ports and it has always worked well.
-
SO. i'm finally messing with an external motion controller. I've tried smoothstepper and didn't care for it. Now i have a uc400eth motion controller on the machine. The problem i have is software limits now cause the system to estop. Which of course de powers the servos and position is lost. Parallel ports never had this problem. Is this just the way it is? Because there really isn't alot of settings to be changed with the motion controller or am i missing something?
I think you may need to either PULL UP or PULL DOWN all the unused input pins (with resisters) in the second port of the US400ETH.
-
I don't think this is a port/pin issue or even a board issue. It is the way the UC400ETH handles software limits. So i guess i'm saying i think this is a plugin issue.
-
Had a chance to do some testing. I generally end my programs with :
G0 G53 Z0 M5
G49
G0 G53 X-14. Y-6.
M30
If i take out the G49, it will run the program just fine. With the G49 in there, it rapids to Z0, reads the G49 line, then moves to Z.01 and trips the soft limits and estops. The buffer is around 100% on the UC400eth i/o screen when this occurs. If i put a little dwell after the G49 line (300ms), then it will run the program just fine.
-
Again, you need to send this info to CNC Drive. If it's a bug in their plugin, they'll fix it.
-
I don't think this is a port/pin issue or even a board issue. It is the way the UC400ETH handles software limits. So i guess i'm saying i think this is a plugin issue.
1. If your port 2 of UC400ETH is unconnected you need to pull up or pull down all the input pins (see attached).
It is TTL logic NECESSITY.
2. If you think it is a plugin issue, connect your PC to the internet (may require an Ethernet switch) and the plugin should automatically update to newer version.
People on CNCZONE already had false triggering of limit switches with UC400ETH even in 2017. So if CNCDRIVE had fixed the problem, the updated plugin should be already there for automatic update by now in 2019.
3. Maybe heed the moto "Sell it buy a VW" and get rid of the UC400ETH and buy Pokey 57CNC instead.
-
I don't think this is a port/pin issue or even a board issue. It is the way the UC400ETH handles software limits. So i guess i'm saying i think this is a plugin issue.
1. If your port 2 of UC400ETH is unconnected you need to pull up or pull down all the input pins (see attached).
It is TTL logic NECESSITY.
2. If you think it is a plugin issue, connect your PC to the internet (may require an Ethernet switch) and the plugin should automatically update to newer version.
People on CNCZONE already had false triggering of limit switches with UC400ETH even in 2017. So if CNCDRIVE had fixed the problem, the updated plugin should be already there for automatic update by now in 2019.
3. Maybe heed the moto "Sell it buy a VW" and get rid of the UC400ETH and buy Pokey 57CNC instead.
he is talking about SOFT LIMITS, not LIMIT SWITCHES
-
You cannot even [Cycle-start] in the beginning (unless you irresponsibly answer YES) when soft-limits are exceeded isn't it?
So it is most probable due to false-triggering of limit switches, I think.
That was what I read in CNCZONE in 2017. They increased the freq to 400K and it got x and y axis to work but still had trouble with the Z axis. At 200K or 100K, they always got limit switch errors and e-stops for x and y axis.
-
Hi,
Machine always gets referenced but it is done outside of mach3 within the AC servo drives.
If Mach is not referenced (homed) then soft limits are out the window. If Mach does not know exactly where
machine coordinates 0,0,0 are then how do you expect it to know where soft limits 500,500,200 (in machine coordinates)
is?
Craig
-
Hi,
Machine always gets referenced but it is done outside of mach3 within the AC servo drives.
If Mach is not referenced (homed) then soft limits are out the window. If Mach does not know exactly where
machine coordinates 0,0,0 are then how do you expect it to know where soft limits 500,500,200 (in machine coordinates)
is?
Craig
Craig,
Person who sold me the MC told me these:-
Router had No home switch(es) so:-
1. I must always power on and off at exactly the same point - to make that the machine x=0, y=0, z=0.
2. [Ref all zero] after powering sets the machine zero
3. Softlimits reference that Artificial machine zero.
4. Before shutdown, I must always execute
G90 (absolute units)
G53 (machine coordinates)
G28 X0 Y0 Z0 (go home)
to send the machine to the same "Artificial machine zero".
Then only I should power off the router.
When I power on the next day, machine will still ref the same "Artificial machine zero" which can be set to zero again [with ref all zero]
I found that too troublesome so I installed home and limit switches except for z--.
I could not have z-- limit switch because I always must move the spindle cylinder along it's mount to compensate for very long and very short bits.
So a block of metal is always used to set the bit height after manual tool change (moving the cylinder up and down). Then the lower soft limit will ALWAY prevent my bits from cutting the machine AL table surface.
-
Hi reuelt,
all that nonsense just to avoid having home switches......I know that's what I did for a while.....and had a few crashes
when I got things wrong.
Then I fitted Home switches (not limits, JUST Home switches) and life suddenly became much easier and soft limits were now
useful and accurate and have had one crash (operator induced) in the four years since.
Not quite sure how OP was doing it prior to the UC400. To my knowledge the ESS (my own controller) handles Referencing
in exactly the same manner as the parallel port used to and I can't imagine any reason for the UC400 to be any different.
Craig
-
The machine is referenced within the ac servo drives. It is much more accurate than letting mach3 do it. The drives have an input that activates the home feature. So the machine is in the exact same spot every time, within one encoder count. This is not a limit switch problem, it is not hitting or tripping limit switches. Softlimits are working the way they should. I tried an ESS on my tormach for a while and hated it. It never handled backlash properly among other things. I've posted on the cncdrive forum so we'll see what they say.
-
Hi,
that is called index homing and yes it is very accurate. Once the servos have homed do Mach's machine coordinates
get reset to zero? If they do then the soft limits you have programmed should be effective.
I tried an ESS on my tormach for a while and hated it. It never handled backlash properly among other things.
To my knowledge there is no 'proper' way to handle backlash, there is no software solution that's bullet proof.
Craig
-
1. If your port 2 of UC400ETH is unconnected you need to pull up or pull down all the input pins (see attached).
It is TTL logic NECESSITY.
This is NOT NECESSARY.
If you aren't using the port, then Mach3 does not even look at the pins.
People on CNCZONE already had false triggering of limit switches with UC400ETH even in 2017. So if CNCDRIVE had fixed the problem, the updated plugin should be already there for automatic update by now in 2019.
This is not a UC400ETH issue, but a breakout board or wiring issue.
-
It is UC400ETH 5V TTL electronics that is randomly being affected by noise because of FLOATING input pins.
LOOK UP ANY ELECTRONICS BOOK.
If however the port2 is connected to another BOB that has opto-isolation ICs for the input pins, then it won't be necessary.
-
In my case, port 2 is connected and used.