MechMate CNC Router Forum

MechMate CNC Router Forum (http://www.mechmate.com/forums/index.php)
-   Troubleshooting (http://www.mechmate.com/forums/forumdisplay.php?f=83)
-   -   Loosing Steps - underpowered ? (http://www.mechmate.com/forums/showthread.php?t=2486)

Claudiu Wed 20 January 2010 11:57

Loosing Steps - underpowered ?
 
I did cut some stuff for testing and learning. What I remarked after proper zeroing Machine Cords and proper zeroing starting point is that after the cut is done, lets say a full sheet of different shapes, when the spindle goes back to the starting point I see a difference of 1,4 on the table and on the screen ?
Why is that so, shouldn`t it go back to 0,0,0???
Next is that when the gantry is mooving fast over x direction I hear a crrrrccrrrr sound, gantry is not mooving and when it stops I realize that during all the sound that I heard the motors were active but did not turn. I checked rails, they had no obstruction. Pinions are well fitted on motor, screws are tight.
It`s strange that both motors are doing this simultaniously, my gantry remains ok, it`s not pulled on one side more or less...!
The only thing that I forgot about was the current resistor on the geckos, was still set to 1,5 Ah. Now I changed the resistors to 4 Ah and hope that this will change the behaviour of the twoo motors.
Can it be that underpowering or limiting current to the motors result in keeping them in a state of brrrrr but not turning?

Gerald D Wed 20 January 2010 12:03

"I see a difference of 1,4 on the table and on the screen ?"

If the screen always corresponds to the position on the table then you have NOT lost steps.

"Next is that when the gantry is mooving fast over x direction I hear a crrrrccrrrr sound, gantry is not mooving and when it stops I realize that during all the sound that I heard the motors were active but did not turn. I checked rails, they had no obstruction. Pinions are well fitted on motor, screws are tight.
It`s strange that both motors are doing this simultaniously, my gantry remains ok, it`s not pulled on one side more or less...!"

. . . .now THAT is the classic lost steps due to too low current or too low voltage supply to the motor. (Or motors that are too small or incorrectly geared)

Watch the motor temperatures after changing the resistors.

Claudiu Wed 20 January 2010 12:18

I use the geared OM 7:1. I will check temps when cutting next time.
Thanks Gerald

Claudiu Tue 16 February 2010 00:22

Hi all,
I`m back with some new info on my loosing steps problem.
After changing voltage limitation from the geckoes to 4Ah, loosing steps got better. Temperature of the motors, when doing some surfacing along x axis raises to about 50 degrees celsius, but I was doing surfacing with some 18 mm tool so they had to run a lot.
But it`s not gone...
I`ve tried to do some other exclusions to find out where the problem could be.
1.st I`ve gone into WinXp and cut down more of it`s services wcich I will never use on the Mach comp. I stopped touching the computer during cutting something.
2.nd found out that I had to redo some earthing on my exhaust tube because I once in a while get a shot when touching table with my pants.( It`s a tiny shot like 1/10 of the power you sense when touching an electric lighter). It seems i still get some static from that evac system.
3.rd I`ve checked also the rails for mechanical defects, but they are all ok. Also loosing step does not appear allways in the same spot.

What is going on with the machine:
It is only one motor which causes the lagging ( lost step), the one oposite to 0,0,0 on the x axis. Mostly it appears when going fast from x-max to the 0 direction, and only into that direction. This is also the motor with the biggest backlash. I found out that the motor axis holding the gear is not 100% vertical on the rail, I have a dif of about 3 degrees. ( could this provoke jumping one tooth?)

When lost step appears it`s allways something like 1,5 to 2 mm, so it`s not only one but a whole bunch of steps...

I will check now my screening again inside the control box to be sure there is no emi inside near the bob where the output to the gecko starts.

bradm Tue 16 February 2010 05:50

Claus-

I suspect the motor/pinion are binding to the rack somehow, increasing friction to the point that the motor stalls briefly. Since the other X motor keeps going, the gantry twists out of square, which somehow frees the binding enough to continue.

I'd try either tightening or loosening the slot and/or pivot bolt for that motor mount plate to see if it changes the behavior. From what you describe, sounds like it's a bit tight as a guess.

Tooth jumping should be a dramatic event that you would notice, so I think you're dealing with something else.

Gerald D Tue 16 February 2010 06:15

Claus, something else you can try is to swop motors, or geckodrives, with your y-axis and see if the problem moves to a different place.

gixi Tue 16 February 2010 06:19

Claudiu you should check if the motor pinion is not to tight to the rails. I had this problem when put back a motor on X axis and I forced the motor pinion to stay extra tight to the rail (it was the A motor). The only tension should be aplied to that pinion is the tension given by the spring.

sailfl Tue 16 February 2010 08:14

Claus,

Did you grease your racks? After you do what Gerald suggested.

Claudiu Wed 17 February 2010 07:57

2 Attachment(s)
Hello and thank you all for the feedback,

What I´ve done so far and some results:
I checked rails for mechanical defects, they seem ok.
I rearranged motor plate so pinion is vertical on the rail.
I recentered pinion to slide evenly along the rail.
I greased rails and pinion with an anti seize lubricant.(thanks Nils)
I loosened the spring, and rechecked skrews holding motor plate in place.(thanks Brad) They´re easily moveable now. (thanks Marius)
I set back Mach frequency to 25kHz.
I tuned motors again from geckoes.
I reduced Motor velocity to 5000 mm/min.
I changed acceleration Settings to 50.

Allways checking MM behaviour between each step did not bring any success until I changed:

Step pulse to 5
and Dir pulse to 5

I still have to read more about these settings but it seems that symptoms disappeared.

After that I started raising velocity again but got a stall in the motors above 9000 mm/min. Stall means I heard them brrrr but no moovement anymore, simultaniously on both x motors. I guess thats because of the frequency setting and motors came to their max rotation value.!?

I set now velocity to 8000 and accel. to 70. I will do some more testing also under working condition.
Attachment 8593
Gerald, I thought about swapping motors, do you think it is enough to swap only the back side of the geared motor and leave the gearbox mounted on the machine or should I change the complete units?

Am I demanding too much speed? It is tehnically by the limit to go with 8000 mm/min?

I also post here some CV Settings. Maybe it helps.
Attachment 8594

lumberjack_jeff Wed 17 February 2010 08:51

I run EMC, but I have my acceleration set to something like 10inches / min / min.

The speed limit on my machine is comparable to yours and is a function of corner speed. Does the Mariss formula suggest that you should be getting higher speeds?

Gerald D Wed 17 February 2010 08:57

Quote:
Originally Posted by Claudiu View Post
Gerald, I thought about swapping motors, do you think it is enough to swap only the back side of the geared motor and leave the gearbox mounted on the machine or should I change the complete units?
It is not a good idea to split open the OM gearboxes - I had some problems there once and the motor was jammed afterwards - no movement at all. After a few attempts, it suddenly was running again and I am still not sure what I did wrong. :o

Claudiu Wed 17 February 2010 09:07

ok, you convinced me. I change whole unit when nothing changes in behaviour.
Gerald, what do you think of the speed settings?

Gerald D Wed 17 February 2010 09:27

Claus, I am out of touch with the speed settings on the geared motors. (I have not "tuned" a MM for a few years). But 8000mm/min [320 inch/min] sounds a bit low.

bradm Wed 17 February 2010 10:18

I have my OM 7.2s at 450 in/min, largely due to the limits of the PC I'm using. I went out and pushed the PC beyond it's reliable limit, and got 500 in/min. I recall running rapids at 600 in/min on a different PC in the past. There are no symptoms of motor problems at any of those speeds. So I would expect you to be able to get to 12000 to 15000 mm/min with those motors.

Gerald D Wed 17 February 2010 10:39

Have you looked at Typical settings for Mach motor tuning ? Looks like you are not far off the mark for speed, but that your Acceleration setting can be increased a lot. This will have a big effect on CV settings. (Your accel. is 70 while everyone else is around 500)

Richards Wed 17 February 2010 13:48

What is your power supply's voltage? Remember that Current mostly impacts torque and Voltage mostly impacts speed. You stated that your motors were running about 50-C. That seems a little low to me. When I run my PK296B2A-SGxx motors with a 35VDC power supply, they temperature stabilize at about 65-C. (I try to keep motors below 80-C.)

The OM motors are derated because of the gearboxes. The identical PK296-03AA motor is rated at 4.5A when wired half-coil, but the geared motors are rated at 3A. Drawing less current will allow the motors to run cooler, so that may be the reason your motors run at 50-C.

Under test loads, I've run my motors at 900 RPM, which would be about 8-ips (or about 12,000 mm/minute) assuming a 1.25" diameter spur gear.

bradm Wed 17 February 2010 14:10

Claus' build thread shows a 29 VAC toroidal, so about 40VDC, presumably. Claus can you confirm that is the measured amount?

Also, can you confirm the measured value of your current set resistors?

Richards Wed 17 February 2010 17:54

The current set resistors for a PK296A2A-SG7.2 motor and a G203v stepper driver should be 35k. I use the closest standard resistor, which is 33k, on the motors at my test bench.

The formula is: 47 * Amps / (7 - Amps) = Current Set Resistor X 1,000. Example: 47 * 3A / (7 - 3A) = 35.25 and 35.25 X 1,000 = 35k. A 1/4 watt resistor works fine.

Claudiu Thu 18 February 2010 02:34

Quote:
Originally Posted by bradm View Post
Claus' build thread shows a 29 VAC toroidal, so about 40VDC, presumably. Claus can you confirm that is the measured amount?

Also, can you confirm the measured value of your current set resistors?
The sec value of the torroidal is 29 V measured, after the rectifier from PMDx it´s at 39 V. All measured.
The resistors on the geckoes have a measured value of 56 kOhm.
My motors are half coil. Would it be a good idea to swap coils, maybe there´s something wrong with one of them?

Richards Thu 18 February 2010 04:55

I looked at your 'build' section and noticed in the photo from the first page of the posts that you have JP1 on the PMDX-122 set to positions 1-2 (5V) instead of to positions 2-3 (ground). That would cause problems for the Geckodrive G203v. The G203v uses a common Ground signal and requires that the step and direction signals be Active High. (The G201 and G202 are just the opposite. They require +5 volts on the common and Active Low signals.) If JP1 is set to +5V, that might also explain why you get better results when you set Step and Direction to 5 in Mach3.

You may have changed that setting. I only scanned your build for photos of the PMDX-122.

Because I sometimes use G202 stepper drivers on my test bench and sometimes use G203vs or the G540, I have to constantly change the JP1 setting and the Mach3 pin settings. When I forget to change JP1, I get missed steps and rough running.

Claudiu Thu 18 February 2010 06:28

1 Attachment(s)
Mike, I´m working on swapping motors like Gerald proposed. I´m gonna change z with x2. x2 is the one making trouble.
I checked BOB.

Attachment 8652

The jump 1 was in the right place but I also checked my ports and pins settings and found out that Step and dir low active was checked !!! Can this be a source of problems?
I unchecked them now and will test after swapping motors. Thanks

sailfl Thu 18 February 2010 06:34

Claus,

I was accussed of changing too many things. When you are testing you need to change one thing and then test it. If you change two things like swapping the motors and unchecking dir low active, you don't know which one fixed the problem. Change one thing at a time, test. If it is still not working, put what you changed back and try the other thing. It some times will take multiple things changed but you don't know what your changes are doing if you change multiple things at the same time.

It takes time to fiqure out what the problem is and it takes consistant testing to do it.

Gerald D Thu 18 February 2010 06:42

Claus, you have told us that both your X motors give a similar performance now - there is no need to swop motors for more troubleshooting. Work with Mike Richards on your voltages, currents and settings.

Claudiu Thu 18 February 2010 07:54

Nils, you are perfectly right. In my list of the things I have done there was allways a checking and jogging to see if the simptoms dissappear.
Unfortunately it seems not to matter if I go fast or slow velocity, nor acceleration, nor direction.
You can hear only something like a clonk and thats all, gantry is out of square by some mm. There is no rule when it happens.

Gerald, the simptoms were remarkable on one motor. X2, the one far away from 0,0,0. After I set step and dir pulse to 5 it only seemed to be better ( in my hopes) but still I have those small blackouts. (not me, ...the motor:))

Richards Thu 18 February 2010 13:53

Claus, having the signals checked 'active low' in Mach3 will cause roughness. (I've had to scratch my head a few times when I've changed between the G202, which uses active low signals, and the G203 or G540 which uses active high signals. When I had the polarity upside down, the motors ran noisy and rough.)

Verify that all signals are properly set. Having one X-axis motor active low and the other active high will not only cause roughness, but it will cause your gantry to 'rack'.

From a technical point of view, the problem with defining the active level has to do with when the signal is seen by the software. Normally, an Active High signal is seen when the signal changes from Low to High and an Active Low signal is usually seen when the signal goes from High to Low. Because there can be a large amount of elapsed time between signals, having a signal 'trigger' on the wrong edge can mess things up.

When I have problems that are hard to trace, I work with one axis as a time. Write a G-code file that only exercises one axis at a time. Try various speeds. When that axis works properly, modify the G-code to exercise another axis. If the X-axis is faulty, you may have to drop the motors away from the rails. That will tell you immediately whether the problem is mechanical or something to do with Mach3 and the steppers. But it's harder to trace problems when two motors are involved. In fact, when I've had problems with a slaved motor, I dropped the motors so that they could run freely and then I exercised one motor as the X-axis motor, without a slave, and then I exercised the other motor as the A-axis motor, without a slave. (In that case each motor seemed to work perfectly by itself, but the two motors wouldn't work together. If I remember correctly, it turned out to be different acceleration settings for the two motors in Mach3.)

Step by step. It takes time, but when you follow a systematic approach you can cross off the list everything that works properly until you find the problem.


All times are GMT -6. The time now is 01:08.

Powered by vBulletin® Version 3.8.3
Copyright ©2000 - 2024, Jelsoft Enterprises Ltd.