Welcome! Log In Create A New Profile

Advanced

Shift problem - sticky bearing?

Posted by AsrroGeek 
Shift problem - sticky bearing?
March 03, 2017 08:25PM
Hi folks - I have a prusa I 3 clone. I've gotten some good prints out of it, but have occasional y axis skip/shift problems. Seams to be related to the model, and the one I'm trying now has very persistent problems.

I think at least part of the problem is that the linear bearings on the Y axis are "sticky". If you watch closely when it's moving in Y - especially +Y you sometimes see the platformkind of shuddering. Moving the platform by hand doesn't feel very smooth either.

The linear bearings look like this:


There are two shafts, with two bearings per shaft.

The shafts seems smooth and straight. I've tried dripping some light oil down into them, but it doesn't seem to help.

Any suggestions? I see lots of vendors for these bearings. Reviews always seem mixed. Is there a better quality bearing available?
Re: Shift problem - sticky bearing?
March 05, 2017 07:39PM
Before swapping out bearing, try loosening both bearing blocks on one of the rails just a fraction so there is a bit of play

A good motion system is designed to have fixed and floating bearings, One rail will be for the fixed bearings and the other for the floating that will shift slightly to account for any miss alignment in the shafts
Re: Shift problem - sticky bearing?
March 06, 2017 05:14AM
Thanks - I'll give that a try. I had loosened all of them, and ran the carriage back and forth a few times to let them find their own alignment, but tightened it all up again. But, as you suggest, if the shafts have some misalignment, they'll still bind.
Re: Shift problem - sticky bearing?
March 08, 2017 05:39PM
Scotty's suggestion (and some white lithium grease) worked, and I finally got a good print on this part. Thanks!
Now to see if I can ramp up the print speed to something saner.
Re: Shift problem - sticky bearing?
March 16, 2017 06:36PM
Ok, so I solved the really bad, early Y axis shift - bed slides very easy & smoothly now. And had one good print on this part.

... and now 5 failures.

bad prints

All Y axis shifts, most near the end of the print. Full print is taking about 14 hours at this point.

Things I've tried

1. Slowed print speed to 40 mm, travel speed to 80 mm
2. Adjusted Y axis voltage - one attempt at 0.8V, rest at 0.9V
3. Thought maybe the Y axis motor was getting to hot (115F with non-contact IR thermometer). Raised unit up on blocks to improve airflow - got it down to about 100F max
4. Replaced Y motor with higher torque unit from steppersonline

Several other models have printed fine in between these attempts, but none took as long as this one - max probably about 8-9 hours.

I have another attempt in progress right now - I swapped the motor driver chips between the X and Y axis.

I've been using the Cura slicer. I've tried Slic3r, but need to do some work on the settings - it's not laying down the first layer cleanly.

I'm running out of ideas here ... suggestions?

Mote info - it's an MKS Gen v1.4 board running Marlin.
Re: Shift problem - sticky bearing?
March 16, 2017 08:44PM
This is odd, the only think I can think of off the top of my heard is are the y smooth rods clean? I know lithium grease while good for linear systems it works as a magnet for dust and dirt. Even though the bearings have shields on them you may have caught something in one bearing that is making it bind at random times. For my linear bearings I pack the with a heavy ball bearing or CV grease before sliding them onto the rails.

Also have you checked your belts and pulley? I know the rubber belts tend to stretch after a wile so you might have less tension on it now after doing lots of 8 hour plus prints. ore the set screw for your pulley might be just a little bit loose causing the skip. Its good to check this before every print.

Also (sorry just thinking of things as I go) Have you checked the stepper driver temp? This could be overhearing and causing the skipping steps, A fan over the control board should fix this

Edited 1 time(s). Last edit at 03/16/2017 09:15PM by scottybfg.
Re: Shift problem - sticky bearing?
March 17, 2017 05:54AM
The rods seem clean - it hasn't been greased very long, and isn't in a dusty environment. When I move the Y carriage by hand, it seems very smooth now. Could be binding, but seems odd to happen that rarely.

I know the set screws were tight for the last print because I just replaced the motor. Think I'm going to get a paint marker and put a dab on the belt & pulley so I can see if the belt is slipping WRT the pulley.

How hot is to hot? The motor driver chips feel a little warm. Not sure how accurate the IR thermometer is on something that small - highest reading I could get in the vicinity was about 100F. They​ have the little heat sinks and a fan - but the X an Y axis drivers are furthest from the fan.

The temp on the new Y motor is about 90F

Thanks for your ideas
Re: Shift problem - sticky bearing?
March 17, 2017 10:00PM
Got a good print overnight but not sure if I fixed something or just dumb luck. Guessing the latter.
Re: Shift problem - sticky bearing?
March 18, 2017 01:43PM
There are a lot of possibilities for shifted layers, you might want to work your way methodically through the following link: [reprap.org]

A slipping pulley is easily overlooked unless you put a witness mark across the end of the shaft and onto the pulley. Too high an acceleration can also be difficult to spot and may only show up as the weight on the platform increases later in the print, or with particular patterns of tool path.
Re: Shift problem - sticky bearing?
March 30, 2017 05:21AM
Such a straight print except for the big jumps would imply that Y axis mechanicals are in good order.

I'd be thinking missed steps, electronics getting hot by the time it reaches that height, the shape of the print at that height - short Y moves

Random - Is your Z sticky/ skipping at that height; Any sign of the nozzle hitting the print and causing the Y to skip a step?

Edited 1 time(s). Last edit at 03/30/2017 05:21AM by Ed would.
Sorry, only registered users may post in this forum.

Click here to login