Steppers out of sync and skipped steps?

[Things]

Hi, I've finished the X axis on my DIY laser cutter (I keep calling it Y in the video!)

http://www.youtube.com/watch?v=lepq-4LaV4g

I explain it in the video, sorry about the video being so dark, didn't realize until after I had filmed it.

Basically, I have 2 belts to move the X axis, with 2 steppers on each belt (4 steppers in total)

2 of these steppers are wired onto channel A on my controller board, and slaved.

Now for some reason, my steppers occasionally seem to loose steps, and the right side of the X axis seems to loose significantly more, and eventually they get so out of sync the carriage comes off the track!

The steppers have plenty of torque, and don't make any kind of noises to suggest they are skipping steps, they are set to 1/8 microstepping and 200 steps per in Mach3.

I have tried increasing the dir and step pulse durations with no success. One theory I have is that the computer is occasionally provided with a background task, which causes the parallel port to loose priority and miss steps, while Mach3 is still counting.

Anyone else have this problem?

Cheers,
Dan

[Tweakie.CNC]

Hi Dan,

You are not alone, many peoples have had similar problems.
Knowing just where to start looking is difficult because everybody will have their own ideas but I would suggest that electrical noise could be the cause. Shielded cable for your steppers (grounded at the controller end only) and a grounded metal box for the controller may be a good starting point. Check that you have not created any 'ground loops' in your wiring and that all interconnected electrical equipment is grounded (preferably to one single point).

Just my thoughts so far.

Tweakie.

[stirling]

Another idea for the mix.

Basically, I have 2 belts to move the X axis, with 2 steppers on each belt (4 steppers in total)

2 of these steppers are wired onto channel A on my controller board, and slaved.

and presumably the other two are wired onto channel X on the controller board?

In both cases, driving two steppers from one channel CAN work but it doesn't allways turn out that way. IMHO, the likelyhood of this working is further exacerbated by the way you have them physically coupled, i.e. via a belt.

Ian

[Things]

The 2 steppers on each belt are wired together, with 1 wired in reverse obviously so it spins the other way. They are in parallel, not series, so I think they should be fine running like that.

What would the belt do to drop steps? Too tight/too lose? I didn't do the belts up extremely tight, but they aren't loose either.

I think all of the axes actually loose steps, which would point to more of a software problem (Maybe the parallel port priority idea?), although the 2 sides not running together is a bit odd. The way the steppers are wires, each of them should do the exact opposite of each other.

I didn't consider interference on the controller board, might try covering it in some foil and grounding it.

Cheers,
Dan

[Tweakie.CNC]

Hi Dan,

May be an idea to change your design at this stage so you only have two steppers on that axis and a rod joining two pulleys at the other end of the belts. Wiring two steppers in parallel and then slaving them with another pair of steppers wired in parallel will never work well. (The reason relates to the accuracy of microsteps and the difference in the size of each microstep between one motor and the next).
Incidentally, that particular axis is the Y axis and the one with only two steppers is the X axis (this becomes important later on when you may want to use your machine to raster engrave and the axis with the least mass needs to move the fastest).
Good luck with the foil idea, I suppose stranger things have happened. 

Tweakie.

[stirling]

Dan

What would the belt do to drop steps? Too tight/too lose? I didn't do the belts up extremely tight, but they aren't loose either.

As I've said, it is likely to exacerbate the possible problems of driving two motors from one driver. These problems include but are not limited to: microstep mismatch (as Tweakie has said), resonance problems - two motors can really screw with a driver's anti-resonance circuitry - and messing with the drive's current limiting circuitry. Part of the reason for these problems is that when two motors are mechanically coupled as yours are, then each motor is alternately driving/being driven whilst the other is being driven/driving. A driven motor is a generator: result: chaos.

However, even if the two motors had their own driver, mechanically coupling two motors together is likely to produce a combined torque value that is actually less than a single motor would produce. This is because they will more than likely be continually fighting each other mechanically.

Ian

[larryc]

Hi things I built a machine 3 yaers ago and I used one stepper for the x axis and one stepper for the y axis. the stepper for the x is 660 oz inch the other is 445. The gantry is aprox 100 pounds. I have a 7 to1 belt reduction. 3 to1 from motor to a jack shaft then 4 to1 on the second stage or from jack shaft to drive gear. It is rack and pinion drive. that gives 7 to 1 eduction for every 3.17 inches traveled. My machine has never missed a step. with the two motors as you have on each side you are lucky that the gantry doesnt twist it self apart. if each motor had its own driver then you would have no problem but with 2 motors an one driver or motor controer You are getting feed back lucky you dont blow the controller.  Better check to see if the magic smoke hasnt left the controller. When the smoke escapes it will not work again.....

[Things]

No, I made sure that all the stepper driver chip ratings were not exceeded. These are only small steppers, they dont have to move anything close to the 100 pounds you mention.

One thing that crossed my mind was, the holes in the pulleys are a few mm too big for my stepper shafts. Could off centre pulleys tightening and loosening the belt cause the steppers to miss steps? My thoughts say yes, although I'm not sure what the correct method for properly fitting them on the shafts are.

Somehting else just came to mind. At the moment, between the stepper shaft and the pulley, there is a bit of silicone tubing that helps keep the pulleys centred. it could even be a case of the pulleys slipping on the shaft.

How are you supposed to deal with shafts being too small for the pulley?

Cheers,
Dan

[Tweakie.CNC]

Sorry Dan, I don't wish to cause offense, but it sounds like your build quality may be letting you down.
Unless you get the mechanics perfect (and I mean perfect) then you will not get the machine accuracy you are trying to achieve.

Tweakie.

[Things]

Nah I know it's not going to be the best machine ever, however I've come this far, not going to let some off centre pulleys stop me