Pausing in the middle of motion

[squarewave]

Has anyone else ever had the problem with motion stopping in the middle of a G01 or G02/G03 line? My Mill does it occasionally, and my Wire EDM is doing it nearly constantly. If I hit ESC and select Continue on Path, it will resume. It makes unattended burning out of the question.

It seems like the slower the commanded speed (the wire machine is running less than 0.1 IPM), the more often it occurs.

The Wire and the Mill are being run from completely different computers, so I don't believe it is a hardware problem. I have a Galil 17xx in one and a 18xx in the other.

Any ideas?

[Carlo Rechi]

I had a similar problem thats almost the same when I first started. At very low feedrate the servo would cog. I was told that servos run better at faster rpm and my friend put a inline gear box on his motor. Now that problem went away. Your problem seems as if the feedrate is so slow that any resistance or drag slows down to a stop. A little push helps.

Carlo

[HuFlungDung]

I wonder if a higher resolution encoder would help? You have to consider that to maintain a low velocity, you'd have to maintain at least one encoder count per second (just a guess). Maybe that threshold is way too low. More encoder counts might help.

[Billyboy]

This may be related but I think Carlo could of just increased the P value and decrease the I value in the servo tuning settings. This stopped the motor jerk at very low speeds for us and may help run servos better at low rpm and slow feedrates speeds too.

Billy

[squarewave]

Thanks everyone - as soon as Carlo responded, I realized what I should try - which is just what BillyBoy said. The servos and encoders are extremely high resolution and are OEM, but the drives are not. I believe it may be a tuning issue and I will give increasing the P a shot.

[squarewave]

OK, I don't think servo tuning was the problem. I increased P from 125 to 500, and I and D as well. There is no change in the pausing. I still need to do some servo tuning, but that is not the current problem.

I did some more troubleshooting. I checked the voltage out to the motors and it drops off as the motion stops. The Motor Command signals at the interface board also drop off to near zero when the motion stops. Just to further check, I leaned heavily on the table and the Motor Command signals jumped up to keep the table in its current location - this is while in the middle of o program that had just paused. I hit ESC and Continue on Path and it starts off and runs again.

So far, it only seems to happen while burning is in progress. I seem to be able to dry run OK. Could this be some type of interference issue? Wire EDM generates quite a bit of electrical noise when it runs. I have everything shielded and have used all of the manufacturer's shielding as well.

Any other ideas?

[Al_The_Man]

If you switch the position screen to 'Distance to go' What is shown?
Al.

[camsoft]

We have to say that we are impressed. Some of you are getting pretty sharp.

All of the suggestions made are valid and do come into play at varying degrees.

In essence what appears to be happening is that the slight resistance of the wire against the metal while burning at such a low feedrate simply make the machine run out of steam so to say. There's not enough velocity to counter the resistance. Voltage amounts sent to the amps are determined by gain settings which apply to a direct correlation between servo tuning and amp pot adjustments. So one solution is to get Galil's WSDK kit for tuning, If you haven't already done so.

In drymode there's no resistance.

In a perfect world you could afford gear reducers as already mentioned since servos do have the least torque at low rpm's. This is the opposite for steppers were there torque is best at low rpm's and drops significantly at higher cutting feedrate rates. Servos need a spinning action to maintain smooth motion. The higher the better. Applications such as Wire EDM that are highly accurate and slow burning need their servo's to maintain a higher rpm for the best cut quality.

Tech Support
CamSoft Corp.
(951) 674-8100
support@camsoftcorp.com
www.cnccontrols.com

[squarewave]

I will also check the Distance to Go - that is a good thought. I have run with diagnostics on, and there are no clues there. It just shows the "GO" statement and nothing after the motion stops.

I do plan on doing some servo tuning this weekend using the WSDK software.

I'm still not sure I am following the explanation. While the machine is running the part, burning at about .1 IPM, it just stops. Not at a program step, but anywhere in the middle of the programmed motion. At this point, the motion command voltage is near zero, wheras while it was running it was fluctuating around .3 volts. I see this as the control telling the machine to stop. Is this not correct?

Now while I am still at this point with the machine not moving, but in the middle of my program, I lean on the table IN THE DIRECTION of the cut. Not only does the motion does not start again, but the controller motion voltage signal increases to RESIST the motion.

It seems to me that if it were a servo tuning issue, nudging the table in the direction of the cut would make cutting resume (as someone earlier suggested) instead of this motion being resisted to keep the current location.

[Al_The_Man]

I would think if the distance to go is registering that there is move left it should be putting out a command to the drive, and usually if the following error gets too great then it should error out. I can see it where the move is practically at the end and the command is too low for the amplifier gain to overcome the last amount due to inertia, but not in the middle of a move?
What amplifiers do you have? if they have a gain pot, Most have to be adjusted until the motor starts to oscillate and then back off 1/2 turn, this before any Galil tuning is done.
Al.

[camsoft]

squarewave,

You're on the right track and basically the previous suggestions that all the good people gave, including servo tuning are still valid. On the other hand, we do not believe it is being told to stop on it's own. You would have to go out of your way to program such logic and would certainly be aware if this were the case. Instead it still seems as if it just ran out of steam, not enough force, rpm or velocity to counter the resistance. The gain settings have the greatest influence on how hard it trys to move. The value I represents how hard it trys to prevent drift. P represents gain or power force.

One word of caution. We suggested the WSDK. Depending on your CamSoft software version it would have to match the same version of drivers on the WSDK. Galil frequently updates these drivers to keep up with Windows operating system changes and released service packs. If the CamSoft and WSDK are not of the same vintage one or other would have to be updated.

Tech Support
CamSoft Corp.
(951) 674-8100
support@camsoftcorp.com
www.cnccontrols.com

[Al_The_Man]

Originally Posted by squarewave I did some more troubleshooting. I checked the voltage out to the motors and it drops off as the motion stops. The Motor Command signals at the interface board also drop off to near zero when the motion stops. Just to further check, I leaned heavily on the table and the Motor Command signals jumped up to keep the table in its current location - this is while in the middle of o program that had just paused. I hit ESC and Continue on Path and it starts off and runs again.

I re-read this and it almost look like a program stop or feed hold situation if the controller is trying to hold the motor stationary.
Al.