Thursday, July 17, 2008
RepRap Built - Z axis is stubborn
The collapsible stand below it was made from an old laundry hamper frame (the flexible mesh bag used to drape over the top like a garbage bag folds over the lip of a garbage can). Drilling holes for the four steel rods to slip into was all it took, it's structurally sound, and easy to transport. I like it.
The RepRap is not yet functional, though. Problems must pop up to allow us to conquer them and feel superior, right? :)
Well, anyway, the main problem I had was getting the Z-axis to work. The stepper motor kept skipping, not having enough torque to turn the belt.
I tried running the motor from a different driver board (the y axis one, which works fine for the y axis), and got the same result.
I took off the belt and checked how hard I had to hold the motor coupler before the motor skipped. I compared this to the y-axis motor and found that the z-axis skipped much more readily.
This suggested to me that the motor might be the problem... So, I disassembled one of the four posts that comprise the z-axis, a seemingly daunting task, but something that wasn't so bad, in order to remove the z-axis motor and coupling from the machine. I also removed the y-axis motor and coupling, and ran them both side by side, applying the same grip test. Now, the two motors were identical!
Then came my hand against my forehead. Of course they were different before, there was a different amount of friction in the z-axis posts and y-axis rod. Because the z-axis had more friction, it required less of my hand-grip to make it skip. Wow... I should have thought of that! :)
Ok... So it's not the motors, probably.
I then checked how hard it was to turn each z-post, and lubed everything I could. Two posts were still difficult to turn. One would easily turn 90 degrees or so, but on each side of that it was difficult to turn.
To make a long story short, I narrowed it down to washers and nuts at the bottom of the post. When these were loosened, the post rotated freely. When they were tight, it would rotate a certain angle (at first 90 degrees, later about 180 degrees?) and then become too difficult to turn. Rotating in the opposite direction produced the same symptoms.
Take a look at the following video (ignore the audio, though). It shows the bottom washer on the subject post as I turn it. It's hard to see, but it seems to bind at the extreme angles. At first, and when I made the video, I thought the washer was bent. It turns out the washer is sufficiently flat.
The bottom side of the corner bracket supplied to me by Bits from Bytes is not flat, however. It instead, has lower edges and a higher middle, presumably from the casting material shrinking as it cures/dries.
Yeah... That Dremel's an efficient instrument on this plastic. As you can see, I ground about twice as much out as I wanted to. :)
After putting the washer and bolts back on, the post turned quite easily.
The other stubborn post happened to fix itself while I was doing all of these troubleshooting steps. It was probably the lube that helped.
In retrospect, I'm not exactly sure what fixed the first post. I could have just been squeezing the corner bracket too hard, or I could be right on about the angled-corner-bracket-edge thing.
After assembling everything again, I was able to successfully move my z-axis under computer control. Whoohoo! :) :) :)
* Last night, some weired things were happening with my stepper driver boards. When i sent a signal from the reprap-host program to run one stepper motor, two boards would light up and start trying to move their motors. At one point, when I told my extruder to move along the +x axis, I got a 45 degree +x and +y movement instead. This happened after I accidentally unplugged one of the boards with my chair, but I've since double-checked the wiring. Did I short something with the flailing wires as they were pulled loose? Did one of the stepper motor board's driver ICs overheat (the one with the big heatsink gets really hot)? Maybe I'm getting electromagnetic interference with my bundled wires? Who knows?
* My z-axis pulleys were installed upside down, I believe. The lip that is supposed to prevent the belt from falling down due to gravity is on the top instead of the bottom. The belt falls off after about 15 seconds of use.
Unexpected simultaneous movement of X and Y is probably caused by an intermittent sync wire.
Sync wire, 'eh? That seems like it should be pretty easy to check. :) Thanks!
Also, search on Nophead's blog for his DC to Daylight post. Essentially he found that a GM3, especially a brand new GM3 emits tons of electrical noise, so a filter (can be as simple as a 10 nF cap on the motor leads) will help, as will keeping motor wires short and away from control wires.
I had the same problem, but a cap and a little bit of brush wear in (after only a few minutes of GM3 use) fixed my noise problems completely. Check your grounds too; maybe you've got a broken or intermittent ground wire somewhere.
Links to this post: