I think that I know what the issue is. When the belts are extending the way that the machine senses if you are pulling is to look at the encoder position and see when it changes. When the encoder position changes the machine knows that you are pulling so it will extend more belt.
I think that what is going on here is that since the belt has rotated so the teeth aren’t engaging with the roller when you pull on the belt the roller doesn’t spin and the machine doesn’t know that the belt is moving
In terms of solutions I think that either I can make firmware which will let you manually drive the motor out, or you will have to take that arm apart and flip the belt around
I’ll work on the firmware right now, but it might take me a little bit.
Sounds good. I did start disassembly, but stripped 4 of my M3 screws on the piece that holds the belt stack in place. Must have tightened them too much. Screw extractor set shows up Thursday. Appreciate the help
I’m not sure I understand…which probably also means that the directions are not clear enough. Here’s a quick video showing the correct order. If anyone has an idea for how I can make the assembly guide more clear I would love to.
I think an outline diagram/picture with them numbered would be the simplest way to show the order in the assembly guide. Picture >= 1000 words as the saying goes.
Although the top-left, bottom-right, top-right, bottom-left seems like a good phrasing assuming “top” and “bottom” are established or otherwise clear.
And now I probably have to pull the thing apart as I wasn’t fast enough to catch the belt slack during a part of the calibration process. . . . @bar PLEASE, Pretty PLEASE implement a feature to step one motor back.
I suppose at least now the motors are in the correct order
Was it the heads of the bolts that striped? Send a message with the parts that are damaged to @anna and we’ll get you some replacements in the mail.
Help me to better understand when it would be used. My concern is that adding a button which if you press it at the wrong time will break things seems really risky.
The real issue here seems to be something going on with the arm that you are pulling on. If there is something preventing it from turning freely it seems like giving the motor more power to force it could create more issues. Can you tell what is preventing it from rotating?
So from what I can tell it was put back together and working correctly. It got to about the second part of the grid during the calibration and I wasn’t able to tend to it fast enough. Too much belt was let out and it kind of over lapped itself. I’ll make a video here quick.
Correct yeah once the lumpy bit is there extend all won’t move that motor at all. As you can see from the video with a little elbow grease it does move.
So the way that extend works is that it watches the encoder and when it see the belt passing by the encoder it will give the motor more power so that more belt extends and it sees that you are still pulling so it gives more power yet and the motor extends even faster.
I’m wondering would happen if you were to push some slack back through the rollers before pressing the extend button like this:
That way when you start pulling there is some belt that’s free to move past the rollers and so the system will sense that the belt is extending and make it extend more.
So after power was off for a while I came back and it did let me extend it. That gave me an idea. I extended it fully, unplugged it. Then using a clamp to refuse the belt from going back in told it to retract. The machine then thinks the belt is in, but in reality it’s half way out. Then I was able to finish the pull. Here is a video of what formed the lump.
Man it was really all the way back there at the beginning.
I was thinking something along exactly the same lines. To make it extend further I was going to suggest setting the frame size in the config file to something huge like 10000mm by 100000mm that way it would extend the belts all the way during the extend process
