Repair a static grinding wheel balancer way

[metric_taper]

I purchased this tool last summer used off eBay. It does not have a manufacture name. It has too much sticktion in the bearings. Each disk is supported by a shaft, and two ball bearings marked 608ZL. So the Z is for the metal shield, and L is some polyamide ball cage.
I'm hoping someone here as taken one of these apart and can confirm what I'm thinking in the attached video.

This unit only works when the misbalance is excessive. So I can't get a wheel balanced close enough not to vibrate.

 

[Weldingrod1]

The wheels need to spin freely and their axles need to be parallel for it to work right. At least that's what I learned the hard way making version 1...

Sent from my SM-G892A using Tapatalk

 

[Cadillac]

I have one near the same. Mine is a Anderson. If you take the covers off it should expose loose balls wrapped around the shaft. At least that’s how mine is designed. No grease just a light coat of oil and they spin effortlessly. Takes a looong time for a wheel to settle. I prefer this style over the parrallel rods on a three point base which I also have. The Anderson seems more sensitive.

 

[benmychree]

Loosen the end caps up slightly and see what happens.

 

[metric_taper]

I have previously removed the end caps, and I removed the exposed shields on the bearings, then gave them a hard spray of starter fluid to remove grease that may have been coagulated. There was no improvement. That was my original thought on why there was so much friction.

I think that this had been dropped or via shipping there is ball damage on one or more of these disk ways. They will spin for a while, two for a long time, and two for shorter time.

I was looking at replacing the 8 bearings with some roller blade/skate board bearings, that are silicon nitride balls, with the steel ball races. And the shields are non contact.
I know this should work good for balancing, but there is something wrong with it.

 

[benmychree]

Perhaps remove the shields on the outer ends of the bearings to facilitate flushing out.

 

[metric_taper]

Perhaps remove the shields on the outer ends of the bearings to facilitate flushing out.

That requires my original question of taking this apart. To get to that, that nut on the shaft has to be taken off, and I think the shaft pressed through as this is done. There just is not enough room to make this easy.
I was hoping someone has done this to theirs to confirm the disassembly. I don't want to make this into a $500 scrap pile, as that's what I paid for this used. Darn things are very expensive new. This one was fairly clean, but I suspect this is why it was sold. .....it don't work good.

 

[Cadillac]

It looks like what you said in the video. Put a rod in the hole get a thin wrench and break free the nut. Use a brass punch and tap on shaft towards that hub with the hole in it. It will drive the bearing out then do the other end. Bearings shouldn’t be a tight fit in casting. I’d maybe try tapping on the not so free spinning shafts to see if you can align the bearing/shaft assy better?
Mine has grooves cut in the casting and a corresponding groove on the shafts with balls filling the groove. I’ll try tomorrow to disassemble and take a pic.

 

[metric_taper]

They used some sort of thread lock on the nuts. The 6mm rod just bent. So I used a vise grip to clamp onto the disk.
It seems the tolerance of the hole in the disk and the machined mate on the shaft are a slop fit.
I can see why it didn't work well as a static balancer as the wheel was running an eccentric, so only large out of balance loads would rotate with any reliable heavy point going down.
I did replace the bearings on the all four shafts with some low friction roller skate bearings.
But I need to figure out how to grind the disks so they are concentric.
The other odd thing that 6 mm hole was not drilled all the way through, and that mass that was not drilled out, became the heavy side. So I drilled all 4 shafts through, and matched the counter bore, so I hope this reduces the inherent imbalance of the system.
These look like quality units from observation, similar to ones that used to be made by Sunstrand. They have no manufacture marking, and I believe now a poorly made copy. At $500 it was an expensive eBay purchase for me. But they go for much more there and new, in the kilo dollar amount.
So I'm still working on fixing these.