The Home Machinist!

Drilling into the ball might be enough to loosen it by itself.

You might want to try stronger chemicals. Brake cleaner is not all that powerful a solvent. You want MEK and/or methylene chloride to dissolve stubborn tar. Dilute some paint stripper with enough acetone to make it runny and leave it in there overnight.

Another thought: take a piece of copper tubing that just fits down to the ball, hone the end until it's an airtight fit to a ball similar to the recalcitrant one, grease it a bit, put it down the hole, pull a vacuum on the tube, and try to pull the ball out. You can use a syringe to pull the vacuum.

Good ideas. Not sure how to drill under it without being a huge pain and ruining parts anyway, so might as well do what the one guy I could find on the web with a similar problem, angle grinder and cut off wheel + credit card...

I have MEK on hand, but I figured if it was varnished/thickened oil, the brake clean would work. Another thing to try that I originally discounted, but it's worth it in a final effort before breaking out the TIG.

I already did the vacuum trick, almost as described. But I used a small piece of air tubing (like on small aquariums), heated the end a bit, made a small flare to fit the top of the ball, pulled a vacuum with my MityVac hand pump, and tugged gently, or alternately pulled energetically, tried again and again, the tube just pulled off even when holding a strong vacuum. You could hear the "pffft" when it popped loose. I will probably try that again, but instead of making a suction cup, make a snug "tube" to ball size and see if I can get it to just pull the ball inside (after some MEK) rather than using the vacuum like a suction cup.

I didn't mean drilling under the ball. I meant drilling straight down into the ball via the hole it is supposed to come out of.

I reread the thread and didn't see where anyone suggested superglue. If I missed it, my bad.
Use your tube/rod and make sure it and the ball is clean and dry. Touch a drop of SG to the rod and then push the rod into the ball and let it harden for a bit. Might, MIGHT, be stronger bond than the vacuum. Not too much glue or things could be worse in a second. I suppose that if you were able to drill a center into the end of an appropriately sized rod it would give a small pocket for the glue instead of a flat surface.
The E6000 adhesive is good for metals, but a bit tricky to use.

--earlgo

Ahh, different than what I thought. I doubt that would work for 2 reasons. One, I suspect it's a hardened ball bearing, thus not drilled with anything less than something like a "hiroc" (sp) carbide drill. Those are also very sensitive to misalignment breakage, so getting a setup that can run it right down the front of that spindle casting would be very tough, and one slip off the ball and the carbide drill would break. Second, it moves easily enough with a fairly light spring to make the detent work very nicely. I would therefore guess that even if it's not too hard, it would likely spin freely against the drill torque. That alone might make it come out, but I don't think I would ever get a hole drilled in it. I'm all for trying just about anything new at this point, but unless I miss something, the cards are stacked heavily against this one.

SG is a good idea, and I like the pocket idea. Should also make it a little easier to limit glue transfer. I don't have any E6000, but I do have some common cheap multi tube packs in the freezer. But as you say, a little too much or twitchy hand and it gets worse. That might be my last try before TIG sticking (with cross air flow).

The glue is a good idea. I hesitated to suggest it for fear that some of it would get away and end up gluing the ball in place, but if you put the tube you used for the vacuum attempt down there first and then put the glue-tipped rod down inside it the risk of gluing the ball down should be minimized.

For superglue you don't want a pocket as it works best between surfaces. Grinding the tip of the rod concave to match the ball would be best.

Good ideas, thanks.

BadDog wrote:The ring I'm trying to remove is Quill Housing Item# 113. Comp# 8007118 Part# J-118 "Hand wheel clutch" The problematic ball is Item# 173 shown on my diagram just above the clutch ring.

I don't think it's peened in, just saying it acts like it. It seems to move freely enough in its hole, but will not be dislodged. The spring came out intact, and I can see quite clearly down in the bore with that shiny steel ball just taunting me. No sign of rust or other problems.

I probably won't have the time (more like the patience) until this weekend to try again. With almost assured failure, I will likely move rapidly to one of my mostly highly refined skills, sticking a TIG electrode to a piece of steel...

I wish you had posted links to pictures or diagrams for those of us who do not have specific knowledge of this assembly. What comes to my mind is to run a reamer down the hole to clear it out or turn a shaft to a ball like end with the OD being the same as the ball or a bit over and run that into the hole to maybe round out the hole and 'smash' any high spots the keep if from coming out.

lg
no neat sig line

Sorry, you are absolutely correct. If anyone is interested, the link below shows the guy using the air pressure trick that failed for me. This is the same collar I'm trying to remove.

Clutch collar detent removal with air pressure.

Another good suggestion, thanks. But it looks very neat down in there, no rough machining or bus are visible even under magnification (that I can manage looking down a hole).

There must be something else of which, we have some experience. I don't want you to have to cut the collar. Coming from cheap old welders, rod stick made an impression. There was some way that could be useful.

Remember the "carbon arc" torches? Sometimes the machine was so pathetic, that the only real heat came from that. Weird semi welding/ brazing came from that. The carbon electrodes were clad with copper and their tendency was to fire at the end of the electrode. In this case; a liner superior to copper (which was already considered) allows one pulse to weld the rod to the ball.
Then the extraction and remember sticky rods. Helmet on and all sweaty and welder on. Worked it back and forth to break that weld to the stick. In this case, it is opposite. Extraction and shock value as the rod can be bent or gripped to where extraction impact will be the end of this.

I do remember the carbon arc torches. And my grandfather, a career "boiler maker" had some arc-gouging rods (to young and poor memory to remember what exactly) that were used to cut rather than weld. But along these lines, I think John's suggestion of sticking a TIG electrode sound best and least risk if some of the other new ideas don't pan out. And I have solid experience with sticking TIG electrodes, so fairly comfortable with that approach having less risk than trying to shove a live SMAW electrode down that little hole.