I started the calibration process to measure the chain length and the left motor chain wrapped around, spooled up and jammed before I could figure out how to stop things. I tried to use the home controls to unwind by pressing move left and then I smelled horrible things. I was able to get the chain released by undoing one of the screws but now the motor will only turn CW. I’m guessing I burned out the motor driver board? Is it possible to fix with an off the shelf part (I’m handy with a soldering iron) or do I need to order a replacement board from somewhere?
it is a surface mount chip, if you are comfortable messing with them, you can
replace just the chip, otherwise contact Hannah about getting a
where can we source the chip?
Is there a way for the firmware to detect a wrap situation based upon the motor feedback?
If not, I wonder if we can incorporate a microswitch that sits above the motor such that if the chain starts to wrap, the chain will hit it and close the switch and this is then detected by one of the arduino controller inputs to stop the machine.
That’s a cool idea. Might be tough to find an input on the arduino if you are implementing an auto z-axis zero.
A chain guide that is a close fit to the sprocket should also work to peel the chain off the sprocket before it wraps.
Man I wish I had AutoCAD at my current job, but here goes:
where the large grey circle would provide a backing for the chain to keep it feeding parallel with the sprocket.
Mind you, I think that @mrfugu’s idea of putting the roller right next to the sprocket works to avoid skipping as well as keep the chain from wrapping. I have been fortunate to only have had the chain start to wrap once while I was right next to it an able to pull it off (this was before I implemented MrFugu’s roller placement), but I am pretty sure that the roller would keep it from wrapping on itself.
found the chip on ebay
The price is tempting, but beware of counterfeit chips. Too, the delivery time is on the order of a month…
at $6 a piece, I would go with a known business like Digikey. Not sure if that’s convenient if you are outside the US, though. Also, I know that shipping can make small purchases much more expensive, but Digikey used to take mail orders with a check enclosed and would ship those orders for free. I used to do that when ordering things like 4 feet of heat shrink tubing that would be mailed in a four foot long box and cost 10 times the product’s worth to ship.
Octopart lists international distributors…
Hmm, I’ve never used them before, but looks like a great service. Thanks for turning me on to it.
I’m going to order a new board and add a chain guide. This forum/community is awesome… So responsive, thanks everyone!
You’re not the first one in that position, by any means!
The ‘Extend…’ button for the chain calibration starts out with a countdown to give one time to scurry over and be near the motor before it starts moving.
On the motor spacing calibration that would get tedious after the first movement. I usually use a small 30mm extension for the first one, long enough for the chain end to dangle if things go right, and not too hard to untangle if not. After there’s a bit of chain dangling, jams are less likely. I’ve also tried hooking a piece of ‘stretchy string’ to the dangling end to manage the chain while it extends instead of piling up below or needing me to manage it. That’s kind of a bother, though.
You could parallel the input? Not like you are going to be doing an auto z-axis zero while the sled is moving… But I do like a guide that also prevents wrapping… better to avoid the problem than detect the problem.
There are more AUX pins available, just need someone to write the code to serve them. The switch closure could call the same function as the ‘Stop’ button from GC does perhaps. Or more elaborate code could send an alert for GC to display.
or set off an audible alarm!
I’d like to get a copy of log during a chain wrap event. Maybe it can just be software based.
How does the motor distance measurement control the “Pull tight”? It seems to me we already have some sort of overload protection (or is it a no more move detection) in the code that could be added at more places.
The code just looks like it turns the right motor for a given amount of time at a given speed. No feedback.