Servos will not move

[Bdrangdo]

Hi, I am trying to fit a Fagor 8055m control to a milling machine. The control had lost its battery and lost everything, I sort of wrote a PLC program by cribbing it from the OEM manual and kind of worked out the parameters. I am using brushless DC servos. When the control is switched on the servos will lock up and moving them by hand with power off will throw up a following error so clearly the servos are being monitored and controlled OK. As soon as I touch a jog button an alarm shows a "hard axis over limit" I believe the error code is CNC 200(X) depending on the axis. It seems as if the control thinks the servos are locked. They are set up as torque control, out of curiosity I altered one to velocity control but it made no difference. There are no "M" codes. I have a floppy disk with a PLC programme which is in Spanish and other file lists one of which is the PLC parameters. There are no sub routines I wonder if my problem is because the are missing. How do you program in the PLC parameters ?, are they necessary?. I have tried different gain settings to no avail. The original Spanish PLC program is for a machining center and is mostly redundant anyway and almost impossible to decipher and if I load it up throws up all manner of errors. Is there anyone out there who can help. Thanks in advance. Dave

Similar Threads:

• finnpower strange move part in nest and tools remain not move with all geometry and
• Need Help!- OSP 200M linear move and rotary move on same line. How is feedrate calculated
• move to saved location or move a fixed distance
• Problem- Control a largescale enormous 1Kilowatt servos to move a waterjet.
• Mach3plasma can't move my servo motor during Mach3Mill can move it ?

[Jim Dawson]

Is it possible that the software thinks the travel limit switches are tripped?

[Bdrangdo]

If only it were that simple, tripping the limits gives a PLC error that I wrote into the program, this is a CNC error code, I suspect it may have something to do with the fact that the M codes are missing but that is guesswork. Thanks anyway.

[Bdrangdo]

Turned out in the end that you were correct. I have always understood that limit switches should be wired normally closed giving a positive 24v input to the control and had written the PLC to give the relevant error messages if this signal goes low. Turned out that the signal should be normally low instead of high. This does not seem to be documented in any way and to make it more confusing the control does not error out until there is an attempt to move an axis. Ah well we live and learn. Now I have another problem. Cutting a long story short all three servo amps are brushless DC the X and Y are brushless motors, and the Z is a brushed motor. All three were configured for torque control but I am finding that tuning the PID filter impossible. Out of curiosity I changed the "Z" axis over to velocity control and it works perfectly. It is possible to change the brushless motors over to velocity control and I may end up going down that route although it means rewiring everything. The question is are these controls capable of working with torque control. There is only mention of velocity control in the documentation. Thanks Dave

[Jim Dawson]

Happy to hear you solved at least one problem. I wish I could help with the other problem. It depends on if the drive will handle velocity control, if the drives are the all the same, then I think it should be no problem.

[Bdrangdo]

Yeah the drive is no problem, they are designed to work with either velocity or torque control. There is no tacho on brushless motors but the drives will accept encoder signals to derive a speed signal. I am not in love with the idea of connecting the 5 volts from the control to the 5 volts from the servo drive, maybe it isn't needed. I have visions of creating an earth loop and of course there is extra wiring which is why I wanted to go with torque control. I tried to use a Fagor 8025 with torque control and it just would not work correctly and that is how these drives seem behave. They are slow to react and appear to need more gain than the control can give to get a faster reaction. I suspect that these controls just aren't fast enough to react with torque control and maintain a reasonable following error. or maybe I am missing something. I have waded through the documentation but there is no mention of torque control. Dave

[Jim Dawson]

What drives and controller are you using? I have mine wired with the drive and controller grounds tied together, thus far no problems. I also done this on a number of customer machines over the years and have had no problems.

[Bdrangdo]

The motors are Parker SM they are a bit quirky in that the encoder will not output signals unless the motor is powered up, in other words there has to be 5 volts from the controller plus 5 volts from the feed to the hall sensors. Now to complicate matters to run these motor in velocity mode the encoder has to go back to the amplifier which are Aerotech B20 to provide speed feedback. The analogue output from the Fagor control is single ended with one side of the output grounded. The input to the amplifiers is differential, the problem I am running into is that the amplifier inputs are effectively connected together on one side but the input according to the documentation is referenced to ground level as is the differential input for the encoder which is taking its 5 volt from the motor which in turn is coming from the control and is skewing the demand signal. I hope that sort of makes some kind of sense. It works fine with a brushed motor which an independent tacho output. I guess I am trying to achieve the impossible and will probably revert to brushed motors that I can use velocity mode with. I do have some Parker OEM670 drives that are designed to work with these motors but they are only torque control. As far as I can tell Fagor only stopped making the 8055i controller a few years ago, it seems odd that it will not work correctly with torque control which I understand are easier to tune. I have used the Parker motors and the OEM670 drives with Mach3 using a CSMIO motion controller and they worked fine with that but from memory the analogue output is differential and unreferenced so there are no problems. I can get the drives to work correctly as individuals if the other two are completely disconnected. When all three are connected there is crosstalk between the demand signal causing all three drives to react. There is provision in the amplifiers to disconnect the differential analogue inputs reference to ground which in theory at least would cure the problem but sods law prevails because of the strange motor electronics. Pulling the jumper which removes the ground reference also removes the ground connection to the hall sensors which in turn powers down the encoder output. It is a bit like chasing rainbows. Thanks for reading all of this, I hope it made some kind of sense. Dave

[Jim Dawson]

I have a few of those motors and drives on the shelf, but I haven't played with them yet. Picked them up at an auction, going to be installed on my surface grinder when I get to that project.

Without having the hardware on the bench I'm afraid I can't be of much help. If you were using a Galil controller I could probably help, that what I use on all of my machines.

[Bdrangdo]

Since I wrote the last post I studied the doc's for the drive and in half a sentence it specifically says that when using a single ended demand signal the ground has to be connected to a specific terminal. I will check the connections tomorrow and see if I have them right or not. It would account for my problems if they are the wrong way round. I have used a Galil cards before and they worked great, always seemed odd that such a professional control did not seem to have any specific software for it other than Mach3. I recall having huge difficulty getting hold of what seems to be a non standard connector. They are a bit rare in the UK where I live.

[Jim Dawson]

I didn't like the way Mach3 worked with the Galil cards, so that's what prompted me to write my own CNC software. Rather than Mach3 letting the Galil card do the heavy lifting, Mach3 takes over completely and makes the Galil card a really expensive breakout board. Pretty much a waste of a very intelligent card.

[Bdrangdo]

I eventually changed over the motors to SEM DC, using the tacho's for feedback which are isolated from each other. I left the original drives in and set them up for velocity control and it works really well, no big deal with the PID settings, I even tried using the DRO scale for feedback and I was pleasantly surprised that the motors stayed quite even with the dead band parameter set to around 0.01 mm and I know there is a degree of backlash in the screw. I can only conclude that this particular control just does not work with torque control unless I have missed something. I memory serves me the Galil board had the ability to accept feedback signals from both the linear scale and the rotary encoder which I could never see the point of, surely one would end up fighting the other, or again am I missing something. Dave