How to remove a 'press fit' steel ball bearing?


I'm doing some work to my Russco turntable. I need to replace the ball bearing in platter shaft. This bearing is pressed into the end of the platter shaft and cannot be just 'popped' out. It has to be forced / pulled out. Being pressed into the shaft, there is no way to get a grip on it to pull it out. Someone suggested that I heat the end of the shaft which would cause it to expand so the hole in the shaft would get bigger and the ball bearing would drop out. I tried that to no avail. Someone else tried to explain a removal method but I think he never really understood how this bearing is set. So, I'm going to try to explain this so that it is more clear.

Imagine that you are holding in your hand a 6" long steel rod of about 1" diameter. You turn that rod so that you are looking directly at one of the ends (you are seeing a 1" diameter circle). In the center of this end a hole of about 3/8" has been drilled. Into this 3/8" hole a ball bearing has been pressed so that approximately 1/2 of that steel ball bearing protrudes from the end of the shaft. How do you remove it?
kwillisjr
OK here's my shot. Get the sharpest drill bit you can find, and drill a whole in the ball. Then thread this whole. Or maybe better; search out a machine shop; explain your needs and see what they have to offer.
Ideally, you need a collet type internal bearing puller.
See the following link. This is obviously larger/more than you need but it gets the idea across. There are also applications where a simpler two-jaw internal puller may be appropriate. Think of these as similar to that shown in the link with just two opposing hooks instead of the collet.

http://www.reliabilitydirectstore.com/SKF-TMSC-series-Internal-bearing-puller-kits-p/skf-tmsc%2030-60.htm

Sorry, I don't know how to paste an active link on this site.

Hope this gives you some ideas.

Steve G.
Go to your car mechanic and ask to use his hydraulic press which is to install and push out sealed bearings.
Post removed 
Post removed 
You're gonna have to post a picture. I think a lot of folks still don't understand it.

I tend to agree with those who say that, unless the existing bearing is flatspotted, don't bother. In particular, if you're doing this because of the tubesforever/ceramic bearing posts, I'd think long and hard about how much good this tweak will really do versus the possibility of screwing up your bearing. T4 is so obviously FOS so many times, I'd, frankly, do the opposite of what he suggests.

IMHO, et.
Elizabeth understands the situation, the others don't. This is not a bearing assembly where there are a number of steel balls in a race. This is a single steel ball pressed into a shaft.

Usually, the thrust bearing is made from a softer material, so it's the part that wears. Most likely just changing or resurfacing the thrust will be enough to restore the bearing's operation. So, check the ball and if it's OK then do as Elizabeth says... hug yourself!
Even when it had a flat spot at the end of the ball bearing, you don't need to remove it. You can use diamond lapping compound to smooth out the flat spot. Then you can use a small teflon, nylon, or brass flat disc placing at the bottom of the bearing housing. Now you have the ball bearing flatted spot contacting the teflon disc as rotating (contacting) surfaces. My old VPI Scout non inverted bearing got a flat spot at the bottom and it still works fine.
Nghiep beat me to it. If Kwillisjr came back and said that the ball was worn, I was going to recommend resurfacing it in a similar manner. I just used 600 up to 2000 grit sandpaper to polish the end of my Delphi's motor armature/bearing to an killer finish. It was pretty worn but turned out very nice. There are some "after" pics of the bearing in my virtual system.
Differential expansion coefficients of different materials is often used to assemble things like this. If you were to heat or cool the assembly the ball might just fall out. Sometimes during assembly one part is heated up and the other is cooled. This makes assembly easy, but disassembly is near-impossible.
I think I understand the situation,and if so then I say it is impossible to drill the ball out because the ball will spin around when the tip of the drill bit contacts it or soon after it bites on the ball.I say leave it alone.If you are firm about it just see a machinist to make a new one .Probably easier task!I've seen steel security bars that are hacksaw proof.How?The manufacturer places rods inside the steel tubing,any attempt to cut it through fails simply because as soon as the blade hits the rod inside the tube it makes it rotate and inposiible to run through the whole piece.Simple and intelligent solution.
Best of luck
George
Thanks for your responses. I actually have two platters and bearing assemblies at the moment. This is not the only table I've ever seen that uses a single ball in the end of the platter shaft. It is the only one I've ever seen that actually uses a second ball in the bottom of the bearing well as a thrust bearing. One of these bearing assemblies came from a table that was bought in the early seventies, used on weekends for a couple of years, and then left sitting in the intervening years. Because of the bearings just sitting there, both platter and thrust bearings developed flat spots. The other bearing assembly could possibly be polished out (and it sounds like that may be the preferred option). So, anyway, I'll think about it and see.
You could try some compressed air. I've used it to remove thrust plates down at the bottom of a bearing well a few times. The air blows out any oil that holding the ball/plate in place and gets in behind the ball, loosening it and popping it out. Hey, its worth a shot!

Fran
In automotive applications it usually takes some serious heat from an oxycetelene torch to release frozen studs and pressed fittings-- certainly more than butane or map gas. If the spindle is all metal you might get away with it.

The noninverted VPI TNT spindle bearing is similarly designed, but the bearing just falls out.
I first tried heating the shaft to no avail. Heating the shaft seems to have hardened the oil that had seeped past the bearing over time making it much harder to remove. I finally got that one out using a tungsten carbide bit in a dremel tool. My recommendation is "Do Not Heat the Shaft". I had another Russco platter needing the same work as well as both thrust bearing holders. For those I used a rotozip tool with a thin metal cutting blade. I very carefully cut a slot in the bearing and used a large handled screw driver to twist the bearing until it popped out. It took maybe 15 minutes to do all three.