V-wheel modification

I have been fighting with my v-wheels for a long, long time. I had previously built the Shapeoko 2, a small model, and I have since upgrade to the X-carve (1000m). I gone back and forth with inventables.com a few times because I wasn’t satisfied with the way v-wheels were working. I simply have not gotten smooth linear motion on the makerslide. I’m been through all of the discussions (I think), and I’ve tried most, if not all, of the obvious solutions. A few of the more extreme things that I’ve done is to take apart all of the v-wheels (including the replacements that inventables.com has sent to me), and measured all of the bearings, wheels, rib spacing and spacers. I bought myself a descent set of micrometers, and I have spent countless hours in the garage making measurements and spreadsheets, squaring my machine and taking apart/putting together v-wheels. In the end, I’ve narrowed the issue down to a few issues. I will premise this by saying that I have not recieved parts that are within the quoted specs. I’m sorry, they just aren’t. I’m looking at a box of 100+ bearings, and I can see that the inner race pieces are different thicknesses, and my measurements tell me the truth.

Issue #1: Do the two bearings both spin on the same axis point, or do they have eccentricity? To test this, I loaded two bearings with a spacer in-between on the M5 bolt and tightened down a nut without the black v-wheel. I but the bearings between my fingers, and then rotate the bolt with an allen wrench. I can frequently feel the bearings rotating in different planes. Perhaps press-fitting the bearings into the v-wheel lines up the bearings? Well, it feels hit or miss. I can pick two bearings that appear to rotate perfectly together, put them in the v-wheel, and then rotate the v-wheel and I can feel it “sticking” at different points in the rotation. It’s not smooth. This poses a problem once these are tightened down on the rails. This begs the question whether or not the sides of the v-wheels that each bearing is press-fit into are centered.

Issue #2: Are the two bearings binding? The purpose of the 1mm spacer is to give enough distance between the two bearings so that they rotate freely. In most cases, the GOOD spacers are 1.01 and 1.1 mm in width. I recieved several v-wheel assemblies that had spacers that were < 1 mm, like, 0.7 and 0.8 mm, and that had burs on the edges. These burs catch on the black seal of the bearings. So, I measured all of the rib widths inside my v-wheels (0.97 to 1.07 mm). Technically, all my wheels should not be binding because the spacer is thicker than the rib. Nonetheless, I still had binding, and adding additional shims (very thin) to the inside of the v-wheels hasn’t helped in most cases.

OK. So, I tried an experiment. I put two 1 mm washers in between the bearings inside a v-wheel. That’s a HUGE gap of 2 mm, where the spec is supposed to be 1 mm. Perfectly smooth. Why not put two spacers in? This could work for both the x and y axes of the x-carve. The spacing of the rails could keep the v-wheels always running true despite the extra space of the bearing sticking out on one side, but this would risk potentially a bearing popping out of the v-wheel assembly. On the z-axis, I do not think this solution is at all acceptable. (I’ve read in another post that any lateral motion of the v-wheel is a real problem fo rthe z-axis.)

So, I went on McMaster-Carr’s website to look at different bearings, and I saw a lot of pictures of open bearings. This gave me an idea. What if I remove the black plastic seal from the inside tracks of the bearings, the ones that face in on the v-wheel? The bearings are press fit into the wheel, and I think it would be unlikely that the ball bearings would be exposed to dust and dirt. With the black seals gone, the spacer (no matter what the size) nor the bearings themselves will have a chance of binding, right? Am I missing something? I removed a seal on an extra bearing so you can see what it looks like inside.

https://drive.google.com/open?id=0BydezB1Pca7eeWthaldSaTJrUnM

Will this work? Will removing the black seals on the inside races of the v-wheel prevent binding?
Will it be too risky in the sense that it would allow dust/dirt to fowl up the ball bearings?

The rim on the inside should be slightly thinner than the washer to put a mild load on the bearings, if the washers are too thin the bearings can move and the side thrust support is weak, if they are really loose it increases backlash.
The rim has to bear all the load and is only 1mm wide
In production the bearing seats should align to make the wheels turn evenly. The tolerances are important although plastic will keep deforming until the bearings come loose. If they are made in a two step production both sides will probably not align.
The original spec from thet SO2 is 2.7kg per wheel, many machines will put a lot more on the wheels.

Openbuilds had Polycarbonate wheels for a while but took them out of production as they failed under load,

I think it is time for a better wheel design altough the available space on the X-Gantry does not allow much more material.

My guess is that if you make the wheel for smaller diameter bearings like the 5x13x4 instead of 5x16x5 you get 3mm material on the inside rim and a stronger wheel. Closer tolerances would also be benficial. The bearing is weaker but still way stronger than the delrin wheel. I would stick to Delrin and not move to Polycarbonate. With more space on the Y and Z you can make the wheels a lot sturdier.