Cincinnati Hydrashift Lathe Spindle Bearing Help Needed

I'm replacing the spindle bearings on my Cincinnati Hydrashift 12-1/2" x 36" manual lathe, but have some issues that hopefully someone here with more bearing knowledge can chime in on.

First, some background:

1. The spindle bearing setup is a box design, i.e. three bearings total: two opposing tapered roller bearings, one at the nose, and one in the middle of the shaft, with a deep-groove, single-row ball bearing at the left end. The middle tapered roller bearing is set axially with an adjustable locking collar nut, which controls the pre-load on the two roller bearings. The ball bearing is fixed to the spindle, and essentially "floats" axially in the bore in the headstock, allowing for thermal expansion of the spindle. Here's a drawing of the spindle from the Cincinnati factory manual and parts catalog:



2. This lathe is an earlier machine, built in 1963 (I think), s/n 1W295M-23. It is the standard model, not the toolroom or 3,000rpm version. The spindle rpm range is from 26 to 1,800 (two-speed motor).

4. Spindle bearing lubrication is by a low-pressure oil pump with distribution lines from an un-regulated splitter block to the individual bearings. Lubricant is Mobil DTE Light. Oddly enough, the left end ball bearing on the spindle does not have a lube line, so it must rely on splash lube.

5. Cincinnati did not give specifics on the spindle bearings in their manual, only referring to them by their factory part numbers, but I'm assuming that the different part numbers for the tapered roller bearings (callouts 8 and 13 above) in the standard lathe versus the toolroom/3,000rpm lathe models implies a higher precision rating for the latter.

6. When I pulled the spindle from my machine, the tapered roller bearings looked to be the original Timkens. In addition to the usual "TIMKEN" and bearing numbers stamped into the race, the cone on the center bearing was ink stamped "163" and "3", which I'm guessing means a January 1963 build date, precision Class 3. The cone on the right bearing (next to the spindle nose) was similarly ink stamped "262" and "3".

7. The ball bearing had probably been changed at some point in this lathe's life, based on the marks on the mounting sleeve on the spindle. What I removed was an NDH (New Departure Hyatt) 3212, which is a standard 60mm x 110mm x 22mm size. Other than the "3212" and "NDH", it was unmarked, so I'm assuming it was ungraded, probably ABEC 1, and likely not a precision bearing. This one also had a seal on one side, which I suspect is incorrect given the oil flow path in the headstock.

I've been able to find replacements for the two tapered roller bearing cones and cups, which is something of a relief as the middle bearing cone is a "W" suffix, meaning it has a notch ground into the ID for a taper key to lock it to the spindle shaft, and I didn't relish the idea of grinding my own notch into a plain ID bearing. The middle bearing cone and cup are Class 3 like the original. The right end matched cone and cup bearing set that I found is Class 0, so one precision grade higher than original.

Here are my concerns:

1. Will mixing precision classes between the bearings be an issue? One potential problem is that the Class 0 bearing is marked for the high spots, but the Class 3 bearing is not. So there's no way that I know of to align the two bearings to minimize the runout. Or is there?

2. Cincinnati shows different part numbers for the left end ball bearing between the two lathe models, but they didn't show any specs as to the precision class. Should I go with an ABEC 5, or is an ABEC 1 acceptable? My inclination is to go ABEC 5, despite the significant cost difference, but a three-point box mounting is a much more complicated dynamic assembly, and I'm not sure how increasing the precision class (i.e. reducing the radial runout) at that end will affect - good or bad - the spindle accuracy, harmonics, bearing life, etc.

3. The Timken website has some great technical resources, and I've been wading my way through it, but I can't find any information about bearing break-in with oil-lubrication. Timken has a section on the break-in process for grease lubricated tapered roller bearings, but makes no mention that I can find for oil-lubricated. The original Cincinnati manual does not mention it either, other than to change the headstock lubricant after the first 60 days of operation.

I contacted Timken technical support last week to get their input on these questions, but have not heard back, so either I'm down the priority list (understandable), or they're just slow with these inquiries.

The left end ball bearing gets oil from the oil coming off of the gears and traveling down the shaft to the bearing. I have a 1920s Cincy that doesnt have any oil pump. I started it with the cover off and there is some splash but oil flowing off the gears and across the shaft was the main distributor of oil.
As far as mixing classes of bearings I think rpm's make a difference. Advice on that will arrive. I think I would at least buy a bearing as a matched set if you cant get the precesion. Dont buy the bearing and race separately.
Your in bearing hell. Been there. I hope you get some good help but do your homework and buy precision if available and next best from there. Some bearings become obsolete.
I cant see where runout within the tapered roller bearings would be a problem with the collar adjusted correctly. Are the old bearings bad.

I agree that the ball bearing on the left end doesn't require much lubrication, so the splash is probably sufficient. But since that is the one spindle bearing that was previously replaced, I'm wondering why.

The Class 0 bearings are only available as a matched cup+cone set, or at least that's the only way I've found them. The Class 3 and lower precision bearings can be purchased as separate cups and cones.

I did speak with a very helpful technical rep from one of the major bearing manufacturers (not Timken), and he didn't think that mixing the bearing grades would be a big concern, only that I wouldn't get much benefit from the more expensive Grade 0 bearings. He said that given the age of the machine, the design of the headstock, the level of precision of the lathe, and and rpm speed of the spindle, that the ball bearing precision (ABEC 5 vs. ABEC 3 or ABEC 1) was probably less important than (i) using a bearing with a steel cage, and (ii) using a bearing with a tight radial internal clearance, such as C2. I hadn't given much thought to plastic vs. steel vs. brass cages, but he said that he'd go with a steel cage for a bearing on a shaft that is driven by a gearbox such as in this lathe.

Even with the correct preload on the tapered bearings, if they are not of sufficient precision, the spindle will rotate in something less than a circle - the less precision, the larger of a wobbly circle.

The old bearings are bad enough that I want to replace them. I've done every other bearing on the other three shafts in this headstock, and given the condition of the lube that was in there when I bought it and of the other bearings, it's time for these to go. It just won't pass the "will it make good parts test" anymore.

Could you tap into your low pressure oiling tubing. And add a line to the ball bearing. Or add an oil slinger by the bearing for assured oiling. Run at a slow speed before putting the cover on to watch what the oil does. Previous ball bearing replacement is telling.
On my lathe the end bearings have a recess below the bearing that holds a little oil . Are yours similar. Like a little oil troft.
It backfired because condensation accumulates in the bottom and ruined/rusted the bearings and races.
It sounds like your making it right.

As far as bearing break in for bearings I would run it at a slow then a medium speed for a couple of hours and change the oil. To flush any particles and let things seat. You may then check your adjusting/locking collar to see that end play hasn't changed.

As far as bearings not rotating in a complete circle with proper preload it would have to be a real poor bearing or the race not seated. I spent 25 years in the printing and envelope manufacturing business. Hundreds of bearings and shafts in each machine. Never seen that, I would cut a piece of cardboard or plywood to cover the top and run it and lift it to watch what goes on.
Last post on break in I said slow and medium. I would also crank it up to top speed for a few minutes after some slow and medium run time. Then change oil.
Being thorough is the only way to work. It's not a 5000 rpm lathe.
I dont see it mentioned much here on the forum but a cheap mechanics stethoscope is a great tool for hearing what a bearing in doing. You just place it on cast iron close to the bearing. We use to use a long 3/8 socket extension. It works.

mllud22 said:

Could you tap into your low pressure oiling tubing. And add a line to the ball bearing. Or add an oil slinger by the bearing for assured oiling. Run at a slow speed before putting the cover on to watch what the oil does. Previous ball bearing replacement is telling.
On my lathe the end bearings have a recess below the bearing that holds a little oil . Are yours similar. Like a little oil troft.
It backfired because condensation accumulates in the bottom and ruined/rusted the bearings and races.
It sounds like your making it right.

Click to expand...

The distribution block just so happens to have an unused port, so I was thinking about adding a line to feed that left end spindle bearing. Maybe pinch the end a bit to reduce the flow compared to the existing feed lines to the other bearings.

As far as I can see, the housings around the bearings do not allow the oil to sump. The drain holes are at the lowest points in the cavities around the bearings.

As far as bearing break in for bearings I would run it at a slow then a medium speed for a couple of hours and change the oil. To flush any particles and let things seat. You may then check your adjusting/locking collar to see that end play hasn't changed.

Click to expand...

That's close to what I was planning, too, absent any different instructions from Timken. I have an infrared non-contact thermometer, so I'll keep tabs on bearings temps as well. I've also added some magnets to the sump in the headstock, just before the return port to the hydraulic lube pump, since Cincinnati only installed a strainer, not a filter, in this system.

Cincinnati was big on shaft nuts with the set screws and feet to lock them on the shaft threads, and I've read about them coming loose and causing damage. So I always install them with some medium strength threadlocker on the set screws, but you're right, checking them is a good precaution.

mllud22 said:

As far as bearings not rotating in a complete circle with proper preload it would have to be a real poor bearing or the race not seated. I spent 25 years in the printing and envelope manufacturing business. Hundreds of bearings and shafts in each machine. Never seen that, I would cut a piece of cardboard or plywood to cover the top and run it and lift it to watch what goes on.

Click to expand...

Perhaps I should have been more specific, but the wobble is likely measured in increments finer than my 0.0001" test indicator will read, so it's not something I'll probably be able to observe. More like what was discussed in this thread: https://www.practicalmachinist.com/vb/general/timken-tapered-roller-bearings-truck-bearings-my-spindle-218992/index2.html?highlight=spindle+bearing+truck

I dont see it mentioned much here on the forum but a cheap mechanics stethoscope is a great tool for hearing what a bearing in doing. You just place it on cast iron close to the bearing. We use to use a long 3/8 socket extension. It works.

Click to expand...

The stethoscope was what led me to the bearings. I've been having surface finish problems with this lathe, and after eliminating all of the usual suspects (tool selection and condition, compound and cross slide way and gib wear, etc.), I put a scope to the headstock and listened to the spindle bearings. Ugh. Removing the chuck and rotating the spindle slowly by hand revealed some brinelling or binding in the bearings.

I assume you have a manual by the picture you posted. The little 90 degree cast iron keeper is ground at a 45 degree angle 1/2 inch wide.You will get a bearing with no notch in it. You will have to do it. Cincinnati will send you one, but it will cost of maybe more than your machine cost. They dare you to buy it.. If you don"t have a manual, I do if I need to look for you. It won"t have the bearing numbers. Timken said to set bearing at .002 cold. With class 3 bearings may be different. When i rebuild spindles, I put a touch of good grease on the bearing for run in. Heat gun is a must. In the field with big lathes, we use a heat gun and ran the bearings by heat. The bearings will fry at 140 degrees. Do a slow run in and keep checking the heat.

Good luck !

cmike said:

I assume you have a manual by the picture you posted. The little 90 degree cast iron keeper is ground at a 45 degree angle 1/2 inch wide.You will get a bearing with no notch in it. You will have to do it. Cincinnati will send you one, but it will cost of maybe more than your machine cost. They dare you to buy it.. If you don"t have a manual, I do if I need to look for you. It won"t have the bearing numbers. Timken said to set bearing at .002 cold. With class 3 bearings may be different. When i rebuild spindles, I put a touch of good grease on the bearing for run in. Heat gun is a must. In the field with big lathes, we use a head gun and ran the bearings by heat. The bearings will fry at 140 degrees. Do a slow run in and keep checking the heat of the heat.

Good luck !

Click to expand...

That special notched ID of the cone on the middle tapered roller bearing is what had me worried, but I was able to find two sources (neither one was Cincinnati, thankfully) for that special "W" suffix bearing, new-old-stock in Class 3 precision, same as the original. Expensive, yes, but not as bad as I expected. As it turns out, I've been struggling more with finding the right deep-groove ball bearing than with the precision tapered bearings.

Thanks for the offer of the manual, but I was fortunate in that this lathe had the original factory manual and sales brochure from the selling dealer tucked in the door of the electrical cabinet when I bought it.

By "heat gun", I'm assuming you mean a thermometer? Or are you warming the bearings for some reason?

I'm used to working with old military manuals, where every detail of the parts is spec'd, so this Cincinnati manual is a bit lacking. It would be nice to have the bearing details, but then Cincy wouldn't have gotten the service parts business back when they were still interested in after-sale support.

Thanks for the post - good to see another person here in Michigan with some Cincinnati lathe knowledge!

Could you please list the cup and cone numbers that you ordered ? I have the same lathe and suspect bad bearings as well. Id like to keep the lathe operating until i get everything ordered. The tail bearing on mine was engraved abec5 by the way

Chains956 said:

Could you please list the cup and cone numbers that you ordered ? I have the same lathe and suspect bad bearings as well. Id like to keep the lathe operating until i get everything ordered. The tail bearing on mine was engraved abec5 by the way

Click to expand...

Glad to. These bearings should be the same for the 10" and 12-1/2" Hydrashifts, standard version (not the toolroom or 3,000rpm models).



I have an Excel spreadsheet of all of the bearings on this lathe, since Cincinnati didn't include that info in the manual. If you'd like a copy, send me your email address in a PM.

Thanks for the ABEC spec info about the left end spindle bearing in your Hydrashift. Do you also know the manufacturer and part number of that bearing?

Is your Hydrashift a standard model, or a toolroom/3,000rpm version?

The tail # was 4248 i didn't see a manufacturer name it may have been on the other side .