Deos the bolt length in the 41 kits impair the accuracy?

Troubleshooting
1

For many stages of the assembly of my 4.1 I found myself wishing that the bolts were just a few mm longer. It often required a lot of force to push the nut and bolt together enough to engage the threads, and by the time the bolt thread was into the nylon locking portion of the nyloc nut, the bolt felt too tight. This is particularly true of the spools which were pretty solid, with friction both in the turning of the spool and the belt passing out through the arm and belt guard/guide.

With the unit built I found extension/retraction problems that others have reported, which can be overcome by upping the force. I have managed several calibrations, but there are big errors in the results. At the moment my X error over 2440 mm is about 15 mm (2/3 inch over 8 ft).

Iā€™m thinking that the solution here is to spend a few Ā£ (insert local currency here) on 40 M3 bolts which are 14 or 16 mm, rebuild the spools and try again.

New bolts should arrive today (I couldnā€™t get matching Torx quickly and have gone for Allen internal hex M3 16mm). I will post any updates once Iā€™ve re-built and re-calibrated .

2 Likes
2

I am interested to hear the results! I agree that longer bolts would be nicer for assembly, but Iā€™m not sure about the impact on accuracy

3

I think from what Iā€™ve read on here that Maslow senses tension from motor current required to retract the belt by turning the spools. Any significant friction in turning the spools may be a problem. Even if Maslow tries to estimate the friction load due to the spools alone, it wonā€™t necessarily be the same throughout a rotation, and may vary with direction. Thatā€™s why Iā€™m focusing on reducing the friction and why I think it may be linked to bolt length causing the arms to be too tight.

4

Anthony Huggett wrote:

I think from what Iā€™ve read on here that Maslow senses tension from motor
current required to retract the belt by turning the spools. Any significant
friction in turning the spools may be a problem. Even if Maslow tries to
estimate the friction load due to the spools alone, it wonā€™t necessarily be
the same throughout a rotation, and may vary with direction. Thatā€™s why Iā€™m
focusing on reducing the friction and why I think it may be linked to bolt
length causing the arms to be too tight.

currently the maslow doesnā€™t try to figure out tension, it deals with length.

It does detect when motor current spikes while trying to move to detect when the
belt is tight, but thatā€™s not measuring tension, just that itā€™s not moving and
applying ā€˜enoughā€™ current.

but yes, reports are that with arms that are able to operate with lower current
limits (aka move more smoothly) the calibation and motion does tend to be more
accurate.

David Lang

1 Like
5

With the work done, retract all is more reliable - using 12mm bolts can distort the belt guards and clamp the spools. I also sanded the spool internal bearing surface. Reducing friction does make ā€œretract allā€ more reliable and at lower force settings. However I havenā€™t a conclusion on the accuracy - I had some very bad Z positions which I also corrected prior to the calibration. I now think Iā€™m ~1 mm out over 1220 mm in the Y direction at both edges, but something like 13 mm over 2440 mm in the X direction.