Thread: Tool Post Grinders - Can A good one be made?

Originally Posted by nearnexus

What makes you think you need to pre-load the bearings?

G'day Rob.

Simon has angular Contact bearings being a 7200. In his case 15 degrees. (Which is what I'd expect for this application). They most certainly need pre-load. They virtually rattle / full apart if there isn't a load to keep the races intact.

When I get back into town, I look up what the force is for a light preload. Every bearing catalog lists it. Then I'll talk him through setting that spring, to that force for its given gap. Its as high-tech as compressing the spring under a drill press over a set of bathroom scales.

Being such a small bearing, bathroom scales wont cut it. I'd guess 100 - 200 N range. Might have to borrow the brides cooking scale, or substitute a dead weight.

This test it and adjust it routine, that you read on the net. That must have been developed by Rollie's cousin. A angular bearing that's dead loose, will develop just as much heat as one that's over loaded. If you don't have enough preload to ensure that the balls actually have enough point contact to make them roll. Sliding friction will kill them. I believe that factor is about 7:1, rolling friction verses sliding friction. Hence the heat if an unloaded angular contact is loose.

How would you know, if you started it up and it runs hot. Too loose, too tight? Unless you know that it has pre-load to begin with, its as bout as accurate as hooking up a 3 phase motor. 50/50 % at best.

Regards Phil.

Originally Posted by Machtool

G'day Rob.

Simon has angular Contact bearings being a 7200. In his case 15 degrees. (Which is what I'd expect for this application). They most certainly need pre-load. They virtually rattle / full apart if there isn't a load to keep the races intact.

When I get back into town, I look up what the force is for a light preload. Every bearing catalog lists it. Then I'll talk him through setting that spring, to that force for its given gap. Its as high-tech as compressing the spring under a drill press over a set of bathroom scales.

Being such a small bearing, bathroom scales wont cut it. I'd guess 100 - 200 N range. Might have to borrow the brides cooking scale, or substitute a dead weight.

This test it and adjust it routine, that you read on the net. That must have been developed by Rollie's cousin. A angular bearing that's dead loose, will develop just as much heat as one that's over loaded. If you don't have enough preload to ensure that the balls actually have enough point contact to make them roll. Sliding friction will kill them. I believe that factor is about 7:1, rolling friction verses sliding friction. Hence the heat if an unloaded angular contact is loose.

How would you know, if you started it up and it runs hot. Too loose, too tight? Unless you know that it has pre-load to begin with, its as bout as accurate as hooking up a 3 phase motor. 50/50 % at best.

Regards Phil.

Hey Phil,

Any chance of talking a bit louder so I can hear too ! I had no idea when I installed the Bardens in my 44. All I could attempt to do was set the preload nut at the depth it was originally using a depth micrometer. Whether that produced the requisite preload would be anyone's guess. Nowhere in any of the technical literature I have from Dumore, is the setting of preload discussed. A return to factory for expensive servicing sort of thing I imagine.

BT.

G'day Bob.

Your 80% of the way there all ready. https://www.woodworkforums.com/showth...85#post1873785

http://www.bscindia.com/catalogue/FAGPrecBrgsCat.pdf

That's one of the best bearing engineering catalogues going around. I call it the red cover FAG book. I'd draw your attention to page # 93. I'm a mile off in my estimate of 100 - 200 N. I never deal with stuff this small. Part of what I was saying about getting back to town. But I'll hammer it out from the hotel. I have that Fag Cat saved.

According to the above, light preload in a B7200 C (15 deg) is 32N. Because its a pair. And Uncle Newton said each force has and equal and opposite attraction. Long, story short, you need 32 N opposing each bearing. That would be 64 N, squishing those bearings apart, and applying the preload to that pair.

Practical me, would forget about that 9.81 acceleration thingy, And just call that 6 kg's, remembering it's only light preload, and you blokes are never going to be doing 80,000 rpm. So a bit of extra preload for mum and the kids wont hurt. I actually err on that side. I'll reverse engineer a spindle, for the maximum preload it will sustain at it's maximum speed. So if I get a spindle that will only ever see 6,000 rpm. I'll do medium + 20% for a 7014 quad set.

You mentioned 84,000 rpm or there about's some where earlier on in this thread. That's a bogus number in so many ways, but a basis in an other. Not blaming you, all the catalogues use it. Who use's just one bearing. That's the max tax for one bearing, under extra light spring preload, oil mist. Start stacking them together in pairs, triples or quads, add in grease, you can often end up at 20% of that value.

Getting back to Simon, /and/ or you. What ever space that spring gizmo takes up. I wants to to be applying 6+ kg's of force, in the space it takes up.

That gets back to what I was saying before. (prior post), You only have to measure the stack hight of the spring, at that displacement to know how to set it.

Regards. Phil.

Originally Posted by Machtool

G'day Bob.

Your 80% of the way there all ready. https://www.woodworkforums.com/showth...85#post1873785

http://www.bscindia.com/catalogue/FAGPrecBrgsCat.pdf

That's one of the best bearing engineering catalogues going around. I call it the red cover FAG book. I'd draw your attention to page # 93. I'm a mile off in my estimate of 100 - 200 N. I never deal with stuff this small. Part of what I was saying about getting back to town. But I'll hammer it out from the hotel. I have that Fag Cat saved.

According to the above, light preload in a B7200 C (15 deg) is 32N. Because its a pair. And Uncle Newton said each force has and equal and opposite attraction. Long, story short, you need 32 N opposing each bearing. That would be 64 N, squishing those bearings apart, and applying the preload to that pair.

Practical me, would forget about that 9.81 acceleration thingy, And just call that 6 kg's, remembering it's only light preload, and you blokes are never going to be doing 80,000 rpm. So a bit of extra preload for mum and the kids wont hurt. I actually err on that side. I'll reverse engineer a spindle, for the maximum preload it will sustain at it's maximum speed. So if I get a spindle that will only ever see 6,000 rpm. I'll do medium + 20% for a 7014 quad set.

You mentioned 84,000 rpm or there about's some where earlier on in this thread. That's a bogus number in so many ways, but a basis in an other. Not blaming you, all the catalogues use it. Who use's just one bearing. That's the max tax for one bearing, under extra light spring preload, oil mist. Start stacking them together in pairs, triples or quads, add in grease, you can often end up at 20% of that value.

Getting back to Simon, /and/ or you. What ever space that spring gizmo takes up. I wants to to be applying 6+ kg's of force, in the space it takes up.

That gets back to what I was saying before. (prior post), You only have to measure the stack hight of the spring, at that displacement to know how to set it.

Regards. Phil.

Hi Phil,

I appreciate the help your sending our way. I understand what your saying, place 6Kg on the spring and measure it's displacement. Then put the whole assembly together and compress the the spring by the same displacement. I guess it all comes down to the best way to measure that initial displacement which what you are essentially doing is determining the spring constant.

Looking forward to having a go at that.

Cheers,

Simon

Originally Posted by Machtool

G'day Rob.

Simon has angular Contact bearings being a 7200. In his case 15 degrees. (Which is what I'd expect for this application). They most certainly need pre-load. They virtually rattle / full apart if there isn't a load to keep the races intact.

Regards Phil.

Hi Phil,

The Waldown also has angular bearings, and as I adjust them with it running, there probably is a small amount of preload. I've never considered it significant, but you are correct.

The Waldown uses the old add/remove paper shims method.

When the factory assembles these things, they would probably have a simple set procedure such as adjust out the end play, then nip it up another X degrees sort of thing. That or maybe X grams spring loading on a set length (balanced) lever on the spindle.

If you know the bearing friction measurement required, then it should be easy to calculate and use the last method (lever/weight).

Worth Emailing them for advice.

Cheers Rob

OK. I had a play with the preload spring last night after work. Setting the preload will be interesting. Considering the preload is very light, the actual spring is quite stiff, meaning the spring displacement for the required preload will be very small, creating a significant uncertainty in the preload measurement.

I'm about to go out and rig something up that will allow me to measure displacement with several weights and hopefully construct some form of graph. I suspect the displacement to be less than 0.1mm at the required preload though.

On another note, I have been thinking about belts etc. I had it in my mind to just buy one of these which I saw in AB's TP grinder thread:

http://www.ebay.com/itm/20-5M075-Jas...item2331c48f61

But I'm also wondering of the merits of a round PU belt. Something like this:
http://www.ebay.com.au/itm/12-Feet-D...item3f32185039

Anyone had experience with using these as a drive belt for smal load applications?

Simon

[QUOTE

But I'm also wondering of the merits of a round PU belt. Something like this:
http://www.ebay.com.au/itm/12-Feet-D...item3f32185039

Anyone had experience with using these as a drive belt for small load applications?

Simon[/QUOTE]

Yes. I have used this type for a light load application, but don't remember what it was. It performed OK though. An alternative method recommended to me at the time to join these was to heat the blade of an old stainless steel table knife and hold the two ends of the belt one on either side until the plastic softened then slide the two ends off the knife blade and press them together. This helped to keep them aligned and made a neater joint which required less trimming.

I have also used a hollow version of this type of belting on a very small lathe. This used small metallic joiners with a barb on each end which were inserted into the hole in the centre of the belt. Sorry, cannot remember the name of this product, but it worked well. It did not have to transmit much power though.

Frank.

Hi Frank,

The method you describe is similar to the one shown here on youtube, which is what got me thinking. I am mindful of the fact that the belt may be asked to zip along at quite some speed and so a nice seamless (homogenous) join would be required in order to reduce vibration. I have almost finished my first two pulleys for this, it would almost be cheating using a round belt, given the simplicity!

A round belt would be nice in the fact I would have enough room to make a one piece stepped pulley allowing mechanical speed changes without removing the pulleys. Not enough room to do that with a flat belt, well not unless I undercut the inside to allow mounting….

Simon

Finished a couple of pulleys. Time will tell if they behave. I will still need a couple more different sizes but these should at least get me going to see how it all works.

I also did come measurements. It turns out the the spring compresses 0.06mm per 6Kg or there abouts*. The thread on the preload adjuster is 42 tpi or a pitch of approx 0.6mm. So one turn would compress the spring 0.6mm. I want 0.06 which is 1/10 of that. So the plan is once the threaded adjuster just takes up on the bearing, a further 36deg or so turn should get me pretty close. What do people think?

*Realistically I think there would be 15% uncertainty in this given the limitations of my measurements.

Edit: Bob, are you going to bother stripping your spindle assembly and measure your spring? Be intersting to see what results you come up with.
Edit: I just redid the measurements. I got them wrong, but they have been corrected in the above.

20150625_122711.jpg

Simon

Hi Simon, I used to get F's for my maths, at that stage I thought was for "Fantastic",, but since have learnt a lot more, so your maths sounds about right.
Nice work on the pulleys
Kryn

Originally Posted by simonl

Hi Frank,

The method you describe is similar to the one shown here on youtube, which is what got me thinking. I am mindful of the fact that the belt may be asked to zip along at quite some speed and so a nice seamless (homogenous) join would be required in order to reduce vibration. I have almost finished my first two pulleys for this, it would almost be cheating using a round belt, given the simplicity!

A round belt would be nice in the fact I would have enough room to make a one piece stepped pulley allowing mechanical speed changes without removing the pulleys. Not enough room to do that with a flat belt, well not unless I undercut the inside to allow mounting….

Simon

Hi Simon,

I didn't get your link to the round belt video, but note that, while they will work in a standard V belt groove, they are said to work better with a 60 degree groove. For power transmission they should bear on the sides of the groove, not run on the bottom. I got the specs for the pulleys I made from the Model Engineer's Workshop Manual, I think - not sure if I still have it.

Neither of the applications I used them for was a high speed set-up, so I have no personal experience how they perform in these conditions.

There was enough stretch in the plastic belt to change grooves on the lathe headstock spindle pulley without having to release the belt tension, unlike the round leather belts on a small drill press my father owned where tension had to be released from the belt to change speeds.

Regards,

Frank.

Simon,

How wide are your pulleys? Dumore used a 18" x 1/2" wide belt. And how did you apply the load and measure the spring compression? Any informative photos?

As for pulling mine to bits again, maybe. I have one project I need to bowl over before I start fooling around with another.

Bob.

Hi Bob,

For the measurement setup, I used the drill press to provide a reasonable flat platform and I used one of the slots on the table to dangle a bolt with weights through it. I used one of the bearing endcaps from the grinder to house the spring and a washer, just to keep everything nicely aligned. Given the spring sits up against the outer race of the bearing, I disassembled one of the old bearings and pushed the outer race into the cup, then sat the spring on that bearing race. I then placed a washer (I ground the washer flat on both sides) on the spring. The bolt passes through the washer, spring & bearing outer race and pokes through under the drill press table. The bolt head nests pretty well on the washer. I faced the head of the bolt to enble a "reasonable" surface to measure.

For the "no load" zero point I took the micrometer depth gauge (that's already come in handy!) and carefully measured the depth of the bolt head below the top of the end cap. I found some hand weights and a few other odds and ends that fitted nice and centrally on the bolt to provide a "reasonable" concentric loading on the bolt. I then remeasured the depth of the bolt head.

I had been a bit concerned at the very limited range of measurements I took, so I did some more up to a mass of 10Kg. The spring compression is reasonably linear and at 10Kg, the compression was about 0.1mm which is consistent with my previous observations.

I really had no idea what width the flat belt originally was so I allowed for a 3/4" belt. The useable pulley width is a tad more at 20mm or so. I may have made them too wide!

20150625_140102.jpg20150625_140158.jpg20150625_140250.jpg

Simon