Lowrider V2 wandering to one side moving Y

I’m having the same problem as in the other thread… I have tried my best to make sure everything is square and level. I can drive it from one end of the table no problem but if I make a script to run the carriage back and forth 20mm at a time over and over the entire carriage gradually shifts over until one side is rubbing on the table. The only thing I can find that looks off is the side that I bolted the power supply, emergency shutoff, etc to seems to be sagging under the weight tilting the wheel towards that side. Is this the cause?

I think I can get some good cuts by letting it rub on one side so at least it’s straight. Should I try cutting a replacement out of 3/4" MDF ?

 

 

 

I have noticed this too although I’m using Unistrut and that may be compound my issue. I’ve spent a lot of time trying to pin down this issue over the past two weeks and trying to figure out the best way to move forward. I’m considering adding some sort of a bracket to brace the bolts that hold the wheels. The flex is quite noticeable. Washers on the exterior of the plate between hexhead and wood may help? Or maybe a nice press fit between the bolt and an aluminum y plate?

1 Like

Aluminum is quite flexy too and for the same weight there isn’t a lot of advantage unless I’m mistaken.

I was playing with the piece before final assembly and the ‘dog ears’ are quite flexible due to them being so narrow so I am thinking washers will not help very much. I fiddled with the DXF to beef them up and add a section I can screw the electronics to. Not really looking forward to making this change if it’s not the problem though.

I also made a DXF so I can bolt the electronics to hardboard instead to save weight I might try that first.

That said, it’s also possible to use this to advantage. The edge my tubes are rubbing on is a factory MDF edge so it’s perfectly straight and might actually improve my accuracy. The only problem is the non-straight wheels are probably going to introduce other artifacts.

Attaching the files… haven’t tried cutting them yet they may have issues.

Y-Plate-Stuff.zip (2.93 KB)

Some people have put a shelf track down for the wheels to run in which will keep it tracking straight. Check the squareness of the table as well.

1 Like

I’ve figured it out! I’m hoping this will be the case for you as well. The issue I believe stems from the mounts that the wheels are connected to. If you look down from the top ( I held a machinist square to the face of all the wheels) It was apparent whats been going on. Depending on the fit of your through holes (plate and printed part and the tightness of the bolt for the z rail bearing guide your wheels will face in or out causing all types of mixed results depending on other adjustments. as soon as I backed off fully on all the bolts for the zrail bearing tension my y axis started tracking differently and I was able to make it down the full length of my table without rubbing. Here’s the issue I see now, when you fully loosen one bolt and tighten another you are going to loose rigidity and who knows how it will act with cutting force applied,( Maybe not much). I at least now know that I can continue with my build and that it wasent all for nothing… I didn’t really go the most frugal route… I bought my SS tubes online and today found a place in town that could have saved me 160 buck :(… Anyway I hope this insight helps.

1 Like

Jeff, yes my table is square.

Douglas… I think you might have it. This matches with my fiddling. I cut out a thin hardboard electronics mount from the DXF I posted and it’s nice and light but it had no effect on my wheel angle. I did more checking and both back wheels are vertical but both front wheels are angled top out. Checking top down using your method all wheels are angled in due to the clamping bolt with the front much worse than the back. Previously I had a lot of trouble clamping my horizontal pipes. The pipes are exactly the same length but if I clamped them with the same amount of pipe sticking out then the front and back wheels were different distances from each other. So, I changed to clamping so that the wheels were the same distance but the pipes stuck out different amounts. I think it’s because the clamping bolt needed to be tightened differently front and back. Short version is, there are so many variables it’s extremely difficult to get everything adjusted.

If I make some C shaped shims I can clip over the clamping bolts I’m thinking I can make the wheels parallel without messing anything else up… but yeah… otherwise it’s best to not tighten too much and keep the plastic part parallel with the Y plate.

 

 

 

 

1 Like

Definitely alot of variables. You should post your setup with the shims when you get them installed. I was thinking about designing an adjustable arm that would in some way span the distance between the two wheels and some how allow for fine tuning of each individual wheel direction. On both sides looking from the top view the outside edge of the wheels where angled forward. In some cases when I thought I had fixed the issue by adjusting the height of the unistrut the gantry would make it down and back the y axis but would eventually going to one side consistently.

 

Just tried the shims… works great on the one side where I was only out about 2 degrees. I noticed it also fixed the problem with the up/down misalignment as well so they are related. However the other side of the carriage that had the heavy board with the electronics was WAY off… I think maybe the PLA warped over time. I’m wishing I could go back in time and reprint in poly-carbonate. The amount of shimming necessary for that side caused the bearings to bind on the pipe since it shifted them.

I think you have the right idea making some kind of spacer for the end of the bolts. It would be so much easier if the bolts were the same length. Maybe I’ll have a go at modelling one up if I have time.

 

 

Ok… made up the bolt spacer and it makes it really easy to align your wheels. Just turn the m3 bolt in the printed part until they are lined up and they should stay that way since it also acts as a brace.

https://www.thingiverse.com/thing:3532333

I’m thinking it’s going to be a never ending battle trying to get everything lined up. It’s really easy to test though… just cut some slots in foam along the Y axis about 3mm deep and .1mm at a pass. If you carriage is wandering you will see one side of the top of the slot be rounded instead of right angles.

After I get everything back together with the wheels straight I will run this test and if it still likes to wander one direction I will just print up another set of endstops that stick out 1/4" less and put them on the side the carriage likes to run away from and have it always run with the tubes rubbing along the straight edge of my table.

 

 

1 Like

What material are your plates made out of and how thick?

It’s the 1/2" MDF plates that you sell. I can’t say for sure if it’s the plates or the PLA parts… it seemed to get worse over time as I didn’t notice it when I first setup my lowrider.

Just a followup… I was able to eventually get the wheels as straight as I could measure and stay that way through multiple cuts… I had to reprint the braces in PC+ filament, the PLA kept warping. The wander is much less pronounced now but it is still there… probably because there is always going to be a little error in my measurements. I glued a 3/4" wide strip of hardboard to the edge of the table and snugged the carriage up so it was rubbing on it and the cuts came out perfect.

So, for that last little bit of accuracy I might need a metal wheel track like Jeff suggested… or even cheaper route a groove in a couple strips of MDF to do the same job… my torsion box will need to be only 3 3/4" thick to leave room for a track.

 

I was having the same issue with my 1/2” MDF y-plate and wheels bowing inward, causing some wobbliness to cuts in the y-axis. I’ve made a couple modifications that seem have fixed the problem for me:

  1. This first fix is probably enough to help most people. I added metal braces between the wheel bolts on the inside of the y-plates to keep the wheel bolts straight. According to the DXF file for the y-plate, the wheel bolt hole centers are 308.2 mm apart. So I drilled 2 holes 308.2 mm apart using a 5/16” drill bit in a spare piece of thin angled steel I had laying in my garage. I added a couple washers to the wheel bolts before this piece to keep the wheels from rubbing on in.

(Note: I would recommend using something like a flat aluminum bar instead of an angled piece, because the angle interferes with the gantry on the x-axis, causing me to lose a bit of cutting area near the edge.)

  1. This next step is probably overkill, but it was a fairly cheap mod so I went for it. I’m sure you can see from my photos that I’ve replaced my skate wheels with some 3D printed wheels that I designed to ride in 1/2” aluminum u-channel. (I used the same bearings from the skate wheels.) I bought 2 - 8’ pieces from my local hardware store for $40. I used clamps to hold the channels accurately in place while attaching them to the wasteboard with screws and took my time to get them as parallel as possible. (This is crucial to keep them wheels from binding)

I have a 2.5w laser attached to my gantry and I ran tests last night, burning a laser line down the entire y-axis the length of my table. I checked it against a straight edge and it was extremely accurate…I can’t wait to get back to milling some things with my upgraded machine!

Hopefully this can help someone else…let me know if anyone wants the files for the wheels, I’d be happy to share.

4 Likes

I’ve uploaded the wheel file to GrabCAD, search for: lowrider cnc track wheels

That looks like a pretty good solution, @xidcrisisx.

You can get nearly all of the benefits of this sort of system with only one channel. The second one will ‘overconstrain’ the axis and can be prone to binding (as you mention). With just one channel you only need to worry about it being straight and don’t have to bother with being parallel.

3 Likes

Thanks for your advise, I will remove one of them if it gives me any trouble…though I will need to account for the height difference of the non-track side…maybe you could even get away with leaving it in the channel, but only attaching it with one screw on the end so the aluminum channel can pivot slightly instead of binding if it needs to?

Had the first tests on my temporary setup today, and got similar results.
In my case it’s moving more ore less to left or right, depending which direction my Y-axis moves.
Going to try some of the measures regarding alignment of the wheels tomorrow, I expect some improvements, but now good enough to be a final solution.
I like “xidcrisisx” solution, I had similar thoughts long b4 I actually started the build, using angle steel as a rail and V-shaped wheels. I guess the advantage of such an setup is that dust has a hard time to build up on the running surface.VMS-6x2PS angle-wheel-new

2 Likes

Does not show the actually left and right bouncing, just the temporary setup, and first test.
Get the steel delivered next week, to build a steel frame as at some point there should be a plasma running on it too.

1 Like

I built my Lowrider about 18 months ago, got it working but didn’t have a chance to use it properly until recently. I was dismayed to find it not tracking properly, even after building a squaring jig. It would track about -ve X 5mm after going about 300mm +ve Y. But would track back to zero (home) OK. And issue was repeatable.

Logged on to forum to seek help and found this thread. So grateful - immediately diagnosed my issue. I verified it was a wheel (roller) alignment issue. I looked at the solution - they seemed a bit involved.

For what it is worth, I simply mounted aluminum angle section (25 x 20mm x 1.6mm thick) parallel to Y. Rollers track to the inside of the vertical portion of angle. Cheap and easy. Works perfectly - have perfect tracking in Y now. No evidence of rollers binding . See in photos.

Thank you to forum contributors!!


11 Likes