[glbreil]

I am using a Centroid M400 control with a digitizer. I tried it several times and got the error message above. The error occured at differnt places in the run and of course canceled the job each time. I finally decided that it was going to fast and over working the thing. I slowed it down and ran a complete part taking about 9 hours to digitize it with no problems.

I rotated the part 180 degrees and started over, it ran not very long and gave the error again.

Looking at the book, the settings seem to be correct for this parameter.

PID115 and time .5000 - I am not sure but I thinks this means no motion for a 1/2 second and you get the error. I realy don't think I am losing motion.

What could be happening. Never had any kind of error before and have had the digitizer since I got the mill and had never had a need to use it until now.

Any suggestions would be helpful.

Thanks Gary

[AjaxCNC]

Gary,

Full Power Without Motion is usually due to one of three things. 1.) Limit switch is tripped or noisy. 2.) DC servo power to the drive was removed. Here check the e-stop and DC power supply circuit. 3.) The servo drive is damaged. Probably not if intermittant though.

Most likely the problem is a limit switch. When digitizing the machine tends to shake a little and if you have loose or noisy switches they can trip just long enough to flag this error.

[glbreil]

What exactly do you mean when you say a noisy switch. I notice my centroid manual said the same thing. I guess I rule out the limit switch because the Z axis is not aperating anywhere close to the limit.

I have started it again this time at a pretty slow speed and it has run for 9 hours without any problems. I guess that goes back to to the thought that the machine shaking could be the problem, since it is quite a bit calmer at the slower speed.

Is there a maximum or proper speed for digitizing. The machine will run a whole faster than I am currently pushing it.

I will check the limits.

Thanks again Gary

[Michael M]

Gary, I presume he means a switch that has the guts rattling around enough that under a heavy shaking it momentarily changes state.

You might try swapping the Z limit switch with one from a different axis (if they are all the same) and see if the problem moves to the other axis. If you can access the switch you might try setting up some program loop that keeps the Z axis moving slowly and then tap on the switch or try to put some vibration into the machine somehow to see if you can trigger the fault.

Making sure the screws holding the switch in place aren't loose might be another good thing to check, as would checking the wires going to the switch. Maybe the switch is OK and the wires at the far end have loose terminal screws that make for a random loss of contact.

Don't you just hate intermittent problems.

cheers,
Michael

[AjaxCNC]

Michael is correct. Knee mills are notorious for loose and shaky limit switches on the quill. Due to the confined space a heavy duty switch like the X & Y is not possible.

As for digitizing speed, keep in mind that accuracy is reduced at higher speeds. We normally recommend 30 ipm fast probing and 3 ipm slow probing rate. The fast rate is used for the initial approach while the slow rate is used to actually record the point. Parameters 14 & 15 control these speeds.

[glbreil]

Still haven't figured out what is happening. I checked the switch and can not cause it to make the error.

The mill is not a knee mill however, it is a MillPort Rhino bed mill and the limit mount is very substantial.

If you think of any thing else I can check please let me know.

Thanks also for the info on digitizing speed.

Thanks Gary

[Michael M]

Gary, when I did the conversion on my 325 Tree I had to go to a switch at the end of each axis (the Tree used a single switch with a trip at each end of the movement, but the Centroid wanted two switches on each axis so that it would know which end was tripped, presumably so that it could then slow jog off the switch in the proper direction).

I'd suppose your mill also has two switches per axis. Are all six switches the same? If so my next step would be to swap the Z switches out with the other two axes and put one on the X and one on the Y. If it is a flaky switch you should then see the error move to a different axis.

I can see that if the switches are oriented differently that might have some impact on how they fail (if the switches are at fault). But then that might also cure your problem.

At least if you move both switches and it still errors on Z, you'll know it isn't the switches.

I'll also mention that recently my machine had a software upgrade and a part replacement (part replacement due to my bonehead move, not the part's fault!). Right afterwards I was getting a "loss of motion" type of error, and it looks like that was because I'd not yet run autotune to get all the new parts/software comfortable with my machine. Maybe you could copy down the settings in your control (PID screen?) and then run autotune and see if that makes any difference in either the settings or the error. Perhaps there is some stickiness that has crept into the Z axis over time and now it is just at the point of being marginal. If you do that, I think it also might be worth while to run autotune with the Z at different points of the travel, and record the settings with each run.

My Tree has 6" in the quill, but your MillPort has a much heavier Z axis to move and a lot more travel so it may be that it is binding in just one spot that you happened to hit while digitizing.

I don't have a good understanding of how a lot of this stuff works, so I've had to try and come up with methods that let me generate enough data to try and spot an anomaly that I can then take to someone who does understand it and say "does this give a hint as to what is going on?"

cheers,
Michael

[MarcL]

What model M400 do you have?

Is it an original DC-drive M400 (gray console, 10" LCD, sealed keyboard flat in console face)?

A new-style DC-drive M400S (black console, 15" LCD, conventional or sealed keyboard on sloping shelf; DC brush servo motors)?

Or a new-style AC-drive M400 (same console as above, but AC brushless servo motors)?

If it is either DC-drive system, you can use DIP switches on the servo amplifier to defeat the limit switches. If you do this after you power up and home (being careful what you touch, of course) then you can run the machine and see if the error has gone away. If the error still happens with the limit defeater DIP switches thrown, then you know the problem is not in your limit switches or their wiring.

To defeat limit switches, toggle the DIP switches up away from the board. To enable limit switches (normal operation) toggle the DIP switches down towards the board. WARNING: if you try to home with the switches defeated, it will run through the limit/home switch and bottom out the axis.

If it is an AC-drive system, then a faulty limit switch cannot cause "full power w/o motion". A faulty switch on an AC-drive system can only cause "Z+ limit (#6) tripped", or words to that effect.

What values did you have in Parameter 15 (slow probing rate) when you got the error persistently, and what did you have when it ran without errors? Or, if you were using the feedrate override to slow it down, approximately what were the override settings?

I wouldn't expect "full power w/o motion" due to a too-high digitizing speed, but I wouldn't rule it out. When you are digitizing, and the probe touches the surface, the control stops the axis without the benefit of any deceleration ramp. This works fine at 2 to 5 in/min, but is a pretty harsh at higher speeds.

There is an option in Parameter 10 to enable a ramped deceleration in digitizing and probing. This is mainly intended for AC-drive systems, but could be used on a DC system as well. Set Parameter 10 to 4.0 (assuming it was zero before) to enable ramped deceleration in digitizing and probing.


Summary: If you have a DC-drive system and the error goes away with limits defeated, then take a hard look at your switches and wiring connections. If the errors continue with limits defeated, but go away at lower speeds, then either learn to live with lower speeds or enable ramped deceleration.

[glbreil]

My controller is the first one you mentioned with the 10" screen ans flat key board.

I have not had an opportunity let to try it with the dip switches in bypass, but that is a great tip and I will try it as soon as I can get back in the shop.

Idid however check the parameters 14 and 15. They were set at 40 & 10 and I hadn't realized I could change them. I was turning down the feedrate overide and had it set at 43% to get it to run without any errors.

I guess that means I was actually operating at about 17 and 4.5 hich is still the upper end of what was recommended by AjaxCNC on the slow probing rate. I will reset it to 30 and 3 like he recommended and see what that does at 100% feed rate.

So am I to understand as the machine learns the proximity of the probe to the work piece it runs at high speed until it gets close and then slows to the slow probe rate before it touched?

If I do added the parameter 10 value it will continue to slow to a stop as it touches? Is that correct?

Thanks again for the help. I will let you know what I learn.

Gary

[MarcL]

So am I to understand as the machine learns the proximity of the probe to the work piece it runs at high speed until it gets close and then slows to the slow probe rate before it touched?

If I do added the parameter 10 value it will continue to slow to a stop as it touches? Is that correct?

Not exactly. In the bulk of the digitizing moves (moving Z down to measure a point) it moves steadily at the slow rate (Parameter 15) until it touches. Then it stops immediately (if P10 = 0) or begins a smooth deceleration to stop (P10 = 4).

The adaptive part has to do with how far it backs up in Z before moving over to the next point. As it learns the shape of the local surface (the trend of recent points) then it adjusts the Z pullback amount to most efficiently get to the next point.

[glbreil]

In my recent post about the error message "Z Axis Full Power Without Motion" I thought the problem was related to my trying to use my digitzer, however I have discovered that it is not.

When I tuned my machine on today it booted up to the first screen and prompted me to press Cycle Start to Home the machine.

When I pressed it the Z axis took off in an upward motion at full speed. Then it stopped suddenly and the screen said "Stall" Program cancelled.

I pressed it again and the same thing happen except it went all the way to the limit and the screen said "Z+ limit (#6) tripped" program cancelled.

I hit the -Z jog button and it took off down at full speed, went about 6 or 8 inched stopped and said "limit cleared" then it said "Z Axis Full Power Without Motion"

Anytime you hit a Z jog button it takes off in either the up or down motion and does this 6 or 8 inches at a time until it trips the limit in the diection it is headed. Each move is 6 or 8 inches with with each touch of the jog button and it goes full speed.

The X and Y axis both seem to work fine and there is also no number on the DRO becuase it has never completed it home cycle.

I can work the limit switches manually and they trip and reset fine. I can not make them do anything by shaking or moving wires or anythig of the sort.

Any clues? I am rather sure that my intermittant problem I had during digitizing has decided to become permanent.

Any Ideas would be appreciated.

Thanks Gary

[MarcL]

99% likely you are losing encoder feedback from the Z servo motor to the PC console.

Could be a failure in the motor itself; could be a loose connection or damaged cable between the motor and the console.

My guess is that it is the encoder on the motor itself. Presuming your control is newer than 1996 or so, the CPU board would detect a bad cable or connection by the loss of the encoder differential channels and report "bad encoder connection". More likely the encoder's line driver is working, but its optical reader isn't.

If your motors have military-style connectors and you have enough cable, you could swap cables with another axis (e.g. put the X motor cable on Z, and the Z motor cable on X). Then you could see if the problem stays with the motor, or stays with the cable/drive/console/software axis.

Obviously, you want to be pretty cautious and keep a hand on the big red button any time you are experimenting with a machine that is inclined to run away....

Incidentally, you can see the raw encoder counts, even when you have not homed and don't have numbers on the DRO, by viewing the PID Configuration screen: F1/Setup, F3/Config, "137", F4/PID. Encoder counts are displayed in the "Abs. Pos." column in the lower center of the screen.

[glbreil]

I have learned a few things here this evening. First the I do not have enough cable to swap the wires at the motor, but it does have connectors that would allow it.

I was able to get the machine to home even though the Z axis encoder quit reading at anout 12" instead of the usual 17". Once it homed and my DRO got some numbers on it I was able to see what was happening.

When you Jog the Z axis the encoder reading did not change so the motor took off wild at full speed. It has 1/2 second to prove motion and when that doesn't happen even though it is really moving it stops and gives the Full Power No Motion alarm.

I did swap the motor and encoder wires in the control panel and the problem stays with the Z axis. I have to assume since I can find no wire damage that it is the encoder. Oh and when you diconnect the encoder wire it does give you an error message that lets you know the connection is lost.

So the question is do I have to get a whole new motor? Is it repairable? I will of course look for bad connections in the motor and hope that is what I find.

If I have to get a motor or even parts do I go back to Centroid or will that come from somewhere else? Do all the motors have to be a like or will any motor of the same power and such work?

As you can tell I am new at this and this is the first problem I have had. I guess that is good luck.

Thanks for your help! Gary

[glbreil]

I got the motor off and don't see any visible problems. Here is the motor information.

SEM Made in England
Type: MTS30M4-48 NO: T18913
C. Stall Current: 6.8
Max. 3000 RPM 140 V 37 AMP
IP 54 IC 400
S/N E97-62154

The Encoder has the Following information.

SEM Made in England
7557HV 747-1-47811
PPR: 2000 V+ 5-24
NO: 740148


Thanks Again any information would be appreciated.

Gary

[MarcL]

With the motor off the machine, you could plug it into one of the other cables easily enough, just to confirm the problem is in the motor and not the external Z cable.

Cheapest solution is to just replace the encoder. You need to match shaft diameter (this one will be hollow shaft with flex mount) and output type (5V differential line driver). Ideally you also want to match cycles per revolution (2000) and connector type (9-pin black Waldom C-Grid?).

The shaft diameter could be 3/8", 1/4" or 6mm. I am thinking the "-1-" in the part number says it is 1/4".

Presuming it is one of those three sizes, is on a flex mount, and has the flat black connector on it, I have replacements in stock, for $160. Give me a call at 360-738-4829 if you want one. Centroid would have them too, perhaps for a few dollars less.

If you ever do need to replace a motor, any of several models from SEM, Baldor, Glentech or others would serve. You just need suitable voltage rating (preferably 140V); comparable current ratings (e.g. 6.8A continuous, 37A peak); and a suitable encoder (lines and output type). The connector pinout seems to be pretty standard, but may not be universal.

- Marc Leonard

[glbreil]

I got my hands on an encoder. I actually tried to call you Marc and got an answering machine and wanted to get it shipped so I would get it before the weekend.

I got it from Centroid and it is of course different than the one I had.

This one is a Danaher and the wiring all seems to be the same but the mounting is somewhat different. The old on had a spring mount and locked on the shaft with two set screws that tightened to two flats on the motor shaft.

The new one has a collar with ttwo set screws that fits over a hollow shaft that fits over the motor shaft which will all work fine, but there is no definate alignment as with the other one there was only two possibilties. Theis one has infinte possibilties. the directions talk about aligning with an Osciliscope which is not avaiable in my shop. Is this all necessary or can I just mount it and go?

Thanks again Gary

[glbreil]

Well I went a head and gave the encoder a try and it seems to work alright. One thing I do notice without have the motor mounted yet is that it seems to be kind fighting itself.

When you move the motor say .001 it will sit and sound like it is having trouble holding that position, like it is trying to rock back and forth but it is not moving then it will quit after a moment as sit quietly. You can move it another and it might just move and sit quiet and then it mightmake that noise for a while and sometimes it is sitting quiet and will start making the noise without being moved.

I think I have heard that even before the encoder problem and didn't really know what it was.

I think I might hear the Y axis doing that also once in awhile, but I don't think the Xaxis is doing it at all.

Sound like a problem?

Thanks Gary

[MarcL]

I am pretty sure that alignment using an oscilloscope is only necessary for commutation encoders on AC servo motors: not applicable here.

When you move the motor say .001 it will sit and sound like it is having trouble holding that position, like it is trying to rock back and forth but it is not moving then it will quit after a moment as sit quietly. ... I think I might hear the Y axis doing that also once in awhile, but I don't think the Xaxis is doing it at all.

Not a problem. If you want to try to quiet it down, then rotating the encoder on the shaft (spring mount encoders) or rotating the encoder body around the hub (Danaher encoders) very slightly in either direction might make a difference.

Also, you can sometimes quiet an axis down by reducing Kd on your PID configuration. If you have Kd = 15 or 20, try cutting it back to 8 or 10.

[MarcL]

Regarding the Danaher encoders, be very careful inserting the wrench to tighten the collar, and don't forget the step of unclamping the disk. These encoders are easy to damage.

[glbreil]

Marc thanks for all your help. You seem to be very knowledgeable about this stuff. Are you a parts dealer or repair business or just someone who knows a lot about this stuff?

It does seem to be working good. If the noise is not a problem then I won't worry about it. I did try yesterday rotating the encoder a little and could tell no difference. I am pretty sure it did thia prior to the encoder problem, but with the motor laying out on the ladder you can here it good.

I will install the motor this morning and I have a gunstock waiting to be modified.

Thanks Again Gary