Tool Post Grinder For My 10" South Bend Lathe

Canuck75

Active User
H-M Supporter Gold Member
Joined
Jan 1, 2014
Messages
233
My 6" Cushman chuck jaws had become slightly bell mouthed and the #3 jaw nose was also .003" different that the others. As I didn't have an internal TPG I decided to try and convert a Ryobi drywall router to this purpose. It is a 4.3 A, 500 Watt, 27000 RPM motor and even if you apply a 50% efficiency rating it would have plenty of power, about 1/3 plus by my reckoning. The motor shaft is 15 mm with bearings spaced almost 5" apart so thought it would be stable and strong enough. I made a frame out of AL, making only what was necessary to keep the project moving, in case it didn't pan out. An initial 3" long spindle using the existing 1/4" collet took a number of tries to get it to run true because it relied entirely on the 1/4" long collet for support. Not good enough! Looking at the motor shaft I realized that the bore behind the taper was only drilled a loose 1/4" and thus not useful to help support the spindle.

I figured that by grinding the motor shaft true internally so that a close fitting tail on the spindle plus the taper would give the required rigidity and especially the accuracy required for extended spindles. This also required a very accurate setup to get the motor shaft aligned precisely with the lathe spindle centre line. This was achieved by using a 3" long 1/4" bar in the nose collet and making sure it had zero runout at the nose and at the collet nut when rotating the router spindle by hand. At this point it was only important that that 1/4" bar was true to the router shaft. Then, putting the DI in the chuck spindle aligning the 1/4" shaft to zero runout along its lenght on the precise centreline in the horizontal and vertical planes of the lathe spindle axis. It is only necessary to swivel the TPG mount to get the horizontal axis alignment, but, in the vertical the TPG mount on the compound rest has to be very accurate to align this plane with the lathe spindle axis. As I was truing the bore in the router motor spindle, this setup had to be as accurate as the DI would read. Any deviation would result is a tapered hole. Note, this same problem doesn't exist when grinding a piece in the chuck as the TPG travels locked to the carriage which presumably travels perfectly in line with the spindle axis. In any case, aligning the TPG by this method, then mounting a 3/16" chainsaw grinding stone in a lathe collet I carefully ground out the hole behind the taper in the router spindle. The lathe was running 1400 rpm one way while the TPG was doing 27000 the other way. It worked fine.

A spindle was made that closely fit the newly ground bore in the router shaft. The original cap nut of 12mm x 1 mm pitch restricted the spindle to 3/8" in order to leave a 1/2" dia shoulder for the cap nut to grip. At this point the concept was proven with only .001" runout at the spindle nose. I didn't want to fool around with balance issues so just used no n-permanent loctite to hold the stone in place. By the way the hole in the spindle end had been drilled undersize and then turned out to 1/4" to keep it concentric as well.

Prior to a grinding attempt I made three identical plates to place between the chuck jaws so that the scroll/jaws could be in compression as they would be when gripping a work piece. Just to ensure everything was correct I ran a DI against the third outer step, which has never been used and thus not worn, to check for runout, which showed .000" difference. I was comfortable that the scroll plate was still in good enough shape and the jaws were being held accurately enough for the grind. The plates held the jaws at about 1-1/8" gripping diameter by design, so I could get a grinding wheel in and yet still contact the gripping surfaces without hitting the chuck body.

The stone was slightly out of true but seemed to run smooth enough at 27000 rpm, so made a setup to hold a diamond point, and dressed (staying well clear), which was scary, but worked. During the actual grind I only took .0005" cuts until the jaws cleaned up. I know they were true after grinding, and that the jaws/scroll were still the same as when manufactured (from the DI check on the 3rd outer step), however the proof would be when a new piece was held in the chuck and then dialed for runout.

A 1" CRS bar showed zero runout at the chuck and .001" at 5" away from the chuck - good enough for me! A 2" bar gave < .001" runout and a 1/2" bar the same. Bell mouth all gone - success!

I have since made 2", 3" and 4" stiffer spindles of 1/2" diameter with a 11/16" shoulder which required a threaded sleeve to enlarge the router spindle as well as bigger cap nuts.The last picture is of a split nut which clamps over the threaded sleeve to remove it if necessary.

I won't caption the pictures as they are self explanatory but would be happy to answer any questions.

Hope some members might find this of use if they try to make a workable TPG the same way I did. Since I had the router and the necessary scrap steel and AL the only extra cost was for the stones.


IMG_6893.jpg IMG_6901.jpg IMG_6970.jpg IMG_6961 (1).jpg IMG_6956.jpg IMG_6978.jpg IMG_6972.jpg IMG_7038.jpg IMG_7031.jpg IMG_7035.jpg IMG_7030.jpg IMG_7059.jpg IMG_7028.jpg IMG_6961.jpg IMG_7060.jpg IMG_7131.jpg IMG_7128.jpg IMG_7126.jpg IMG_7136.jpg IMG_7138.jpg IMG_7134.jpg
 
That's a really neat project. You mentioned 0.0005" depth grinder cuts. Are you measuring that with a test indicator set against your cross slide movement? Or trusting your dials? I've heard of people setting the compound at a very shallow angle (closer to parallel to the lathe axis) to gain more dial resolution & reduce the effective depth of cut. Interested to know how you went about this. Also, do you happen to know what kind of bearings are in those units?
 
That's a really neat project. You mentioned 0.0005" depth grinder cuts. Are you measuring that with a test indicator set against your cross slide movement? Or trusting your dials? I've heard of people setting the compound at a very shallow angle (closer to parallel to the lathe axis) to gain more dial resolution & reduce the effective depth of cut. Interested to know how you went about this. Also, do you happen to know what kind of bearings are in those units?


I have the large dials on the SB toolroom lathe so I used the dial. A small range DI could be used as you suggest as they can easily show 1/4 thou changes. I have heard of using the compound as you suggest and think that is a clever idea. Setting the compound at 1 degree foe example, and dialling .025" of compound travel, will give you .000436 of xtra fine cross adjustment. The problem is you must set the compound angle precisely enough to give you that expected change. You need to know what result you will get from using the compound in this manner otherwise you are guessing. I don't know what class and quality of the sealed ball bearings Ryobi used but they are meant to stand 27000 RPM.

Thanks
Canuck75
 
I see one mistake that most people do when grinding to true up your jaws. The three blocks or spacer that you straddle between the jaws. You have them against the face of the chuck and compression applied. You have it halfway right. You should move the spacer blocks all the way out to the front edge of the jaws and tighten. Then grind your jaws. Doing this will expose the "bell mouth" and allow you to grind it all out.
 
Swinging the compound to 5.7 deg gives a 10:1 effect on your depth of cut, .001 on the compound gives .0001. Your right about getting the angle right but even off a little still gets you consistent light cuts.
My small 3 jaw is really bell mouthed so this is something I need to do.
On a side note, I've wondered if the stock getting warm as its turned causes the bell mouth affect. The thermal expansion must put a lot of force on the jaws.
 
I see one mistake that most people do when grinding to true up your jaws. The three blocks or spacer that you straddle between the jaws. You have them against the face of the chuck and compression applied. You have it halfway right. You should move the spacer blocks all the way out to the front edge of the jaws and tighten. Then grind your jaws. Doing this will expose the "bell mouth" and allow you to grind it all out.


I will set it up the way you suggest and see if my sensitive DI shows any residual bellmouthing. If so I will touch it up. Thanks for the tip.

Canuck75
 
Swinging the compound to 5.7 deg gives a 10:1 effect on your depth of cut, .001 on the compound gives .0001. Your right about getting the angle right but even off a little still gets you consistent light cuts.
My small 3 jaw is really bell mouthed so this is something I need to do.
On a side note, I've wondered if the stock getting warm as its turned causes the bell mouth affect. The thermal expansion must put a lot of force on the jaws.

I can see this technique is going to be one of personal choice. Using your 5.7 degree offset will indeed give you a 1/10 thou depth change for a 1 thou compound dial change, but you are now relying on the compound angle and dial settings for the accuracy you don't trust the cross slide dial to give you. A shallower offset angle and a bigger compound dial change to get the 1/10th depth difference would seem more accurate to me. In any case, I think the important thing here is the discussion. Use the best way to solve your particular situation and one which is satisfactory to you.

On your second thought, I haven't noticed heat in the material being turned to have an effect on bellmouthing. Then again I don't really work the lathe very hard.

Canuck75
 
Back
Top