Thread: I have the Power (Surface grinder)

I think (thinking out loud here) that if the chuck is ground so that it is flat with respect to the spindle (that is, the distance from the spindle to the point on the chuck directly beneath it is always the same) then anything placed on the chuck and ground will always be parallel so doing that test after the chuck is ground will not tell you anything - unless there is something really weird happening related to wheel height. Doing the test with the chuck as is does tell you what the chuck is doing as by grinding the washers you establish that datum surface on the top of the washers so any variation in thickness should represent the error in the distance from the chuck to the spindle.
(Still thinking)
If the chuck were ground when new and this test was done after some time it would show any wear present but that could be in the ways or on the chuck. Positional errors would be best found indicating with a DI. Form errors could only be found by indicating off a known flat surface (for example a small surface plate known to be flat, positioned over the chuck). If the DI showed a variation when indicating off the surface plate then you could conclude some sort of wear of the machine rather than the chuck.
(More thoughts)
So - the washer test (pre ground chuck) should show errors in the chuck but may be compounded by machine errors. To show error in the machine you need a known flat surface - although if a chuck is used for that purpose, the surface can't be generated on the machine as it will disguise any error present. Using the chuck is also problematic as if it was flattened on another grinder, the very act of bolting it down (or clamping on the machine producing the flat surface) may distort the surface.

Note to self - Things for tomorrow:

• Check Schlesinger for any insights
• Indicate off surface plate

Michael

Originally Posted by Michael G

I think (thinking out loud here) that if the chuck is ground so that it is flat with respect to the spindle (that is, the distance from the spindle to the point on the chuck directly beneath it is always the same) then anything placed on the chuck and ground will always be parallel ....

That being the case, all surface grinders would grind perfectly parallel, and there wouldn't be such a thing as 5 washer test

It's normal procedure, certainly as far as I'm aware, to always grind a chuck when it's mounted to a machine, then periodically afterwards if it gets removed, beat up for some reason, etc etc. I'd expect the likelihood of that chuck having a factory grind on it would be almost zero. It would have the grind on it from the machine whence it came. I know you told me where it came from but can't recall now. The previous machine could be a total POS, who knows? As I said, i'm afraid what you've done now really hasn't told you anything. It's a different world down at these tolerances! Even if the chuck was perfect, just for argument's sake, it's what comes off it when paired with the machine that's important.

I just checked in on my way to turning off the sound of the computer as I'm off to bed, a long day at work tomorrow, so literally just grabbed the first Google hit in the list of procedures when installing a mag chuck to a grinder. I'd expect other sources to state the same http://www.walkermagnet.com/Collater...uckinstall.pdf Don't be afraid to grind your chuck, it's perfectly normal and just what's done on installing. It is however a PIA on machines like ours, as heat is the enemy here. It's a case of grinding a little, letting it cool, grinding some more, letting it cool, etc etc etc. It takes a while, the feeds are fine, and you'll have tired arms at the end of it!

Anyway mate, good to see the girl turning, at least you'll have plenty of wheels to feed her!

My head hurts! I'm going to have to re-read the last couple pages again. Very interesting though, SG tolerances are a whole new ball game. Anyway, I will say this, good work Michael it looks great. I had never heard of the 5 washer test, but then why would I, I don't own a SG! I'm going to read up now.

Simon

The easiest way to check for parallel over the length of the chuck, after taking a lick off it... is to grind something big enough to cover the length of the work envelope, and then check it for parallel, I used a length of PFC. ( parallel flat channel ) it has to be big enough to be fairly rigid.

There is a youtube video somewhere of measuring the test piece...

Found it.. Also the thread discussing the same stuff currently under discussion... http://www.woodworkforums.com/showthread.php?t=134610



Ray

One way is:

grind the chuck...

Get a known good parallel as long as the mag chuck.. Put it on perfectly matched small parallels located at each end.... Indicate with an indicator as you move the table..

As you would expect the emphasis is on the parallels you are using for a reference being true parallel..

Originally Posted by Pete F

Originally Posted by Michael G

I think (thinking out loud here) that if the chuck is ground so that it is flat with respect to the spindle (that is, the distance from the spindle to the point on the chuck directly beneath it is always the same) then anything placed on the chuck and ground will always be parallel...

That being the case, all surface grinders would grind perfectly parallel, and there wouldn't be such a thing as 5 washer test

Precisely - a 5 washer test is no good if the chuck is ground parallel. It only works to show the extent of the deviation from flat

Originally Posted by .RC.

Get a known good parallel as long as the mag chuck.. Put it on perfectly matched small parallels located at each end.... Indicate with an indicator as you move the table..

As you would expect the emphasis is on the parallels you are using for a reference being true parallel..

True - and that was what tripped me up when trying to evaluate to Schlesinger as I was indicating with a 1/10th DTI and that was enough to show variation over the scraped surface of the surface plate I was using (I have a Wing 10x4"). Using the side of a combination set rule I saw enough swing to suggest that the ruler was only straight within a few thou over 12".

This is a photo of the chuck after a 1/2 thou pass. It looks like jobs were mainly put a little forward of centre and then dragged off to the left. I took another 5 or 6 tenth's off and it is virtually "smooth as" (I removed the chuck after the 5W test, cleaned off the bottom of it and the mating surface with acetone and the oiled it - I figure if I'm chasing 0.01mm then I'm in bits of grit territory and it needs to be clean)
P1020605 (Medium).JPG

Exhibit 2 is the page from Schlesinger for grinders such as this (although bigger - and for convenience, 4/10ths is equal to 0.01mm).
HSG (Large).jpg
Of the tests listed 1a and 1b are just for leveling, 2 is of interest, 3 should be spot on after grinding the chuck, 4&5 are not applicable, 6 and 7 may be of interest, 8 is n/a again and 9 again may be of interest. The last test without the figure number is probably most applicable as it deals with results - but as my chuck is only 10" (250mm) long, so if scaling down I'm only allowed 1/10th variation in parallel over that length. The block I ground for the J&S was 150mm long and I had around a tenth variation on that (limit of my micrometers) so with the wind in the right direction I think I can claim that I'm near where I should be.

Michael
(For those wondering about Pete's comments about having wheels to feed the grinder, he kindly found some for me and even delivered in person. You meet some nice people on this forum)

Another test fraught with danger is putting a precision level on the ways and moving the table from end to end and seeing if the bubble strays...

I did that on my 540 and got some nasty results on the cross axis... Then realised the grinder is on rubber pads and moving the saddle over made it compress the rubber and droop...

Not that heat is a problem for me but what i have picked up when grinding a chuck-

Use a smaller wheel, both in dia and width,
Dress the wheel coarsely, ie move the diamond across the face quickly.
Light cuts, (thats a given!)
Turn the chuck on, in case it pulls down to the table differently.

My 5 corner test came out within about 2um after grinding the chuck. I have not directly tested the travel against a known reference, although i have tested parts i have ground against the rock. A level on my table does not change at all when the travel is at either end or front or back. I think the size of machine helps to remove the other variables RC pointed out.....

Ew

Michael I'm away and cutting/deleting comments is awkward on this ipad, but I'm afraid you've lost me in regard the 5 washer test. I really can't see how that told you anything until the chuck was ground. Once it was ground it should indicate how parallel the machine is grinding and the answer should be "very". Maybe I missed a point where you said the chuck was proven before mounting. Even if it was, it really means nought as the chuck is ground in situ anyway, just as you've now done.

I emphasised parallel above, as I don't think the washer test will tell you if it's grinding flat, and you could be producing perfectly parallel bananas and the washer test would look good. Again I could be missing something as I've just spent the day flying from one side of the continent to the other and back again, and am, in a word, shagged. If so I apologise for either or both of the points, however I think you'd need to check flat using another process mentioned.

Incidentally, you make the process regarding the wheels sound much more saintly than reality. In fact Michael purchased some excess wheels from me and had the money in my bank account before I even knew I had an account! He then waited very patiently for me to finally make it to ADL and then endured me invading his privacy while he should have been packing, to answer a bunch of stupid questions while I ewwed and ahhhed at his extraordinary workmanship before I scurried off, perilously close to completely overstaying my welcome, lathe file in hand! I really am looking forward to seeing that cylindrical grinder turning and thought the castings were brilliant.

Originally Posted by Pete F

Michael I'm away and cutting/deleting comments is awkward on this ipad, but I'm afraid you've lost me in regard the 5 washer test. I really can't see how that told you anything until the chuck was ground. Once it was ground it should indicate how parallel the machine is grinding and the answer should be "very". Maybe I missed a point where you said the chuck was proven before mounting. Even if it was, it really means nought as the chuck is ground in situ anyway, just as you've now done.

I emphasised parallel above, as I don't think the washer test will tell you if it's grinding flat, and you could be producing perfectly parallel bananas and the washer test would look good. Again I could be missing something as I've just spent the day flying from one side of the continent to the other and back again, and am, in a word, shagged. If so I apologise for either or both of the points, however I think you'd need to check flat using another process mentioned.

Incidentally, you make the process regarding the wheels sound much more saintly than reality. In fact Michael purchased some excess wheels from me and had the money in my bank account before I even knew I had an account! He then waited very patiently for me to finally make it to ADL and then endured me invading his privacy while he should have been packing, to answer a bunch of stupid questions while I ewwed and ahhhed at his extraordinary workmanship before I scurried off, perilously close to completely overstaying my welcome, lathe file in hand! I really am looking forward to seeing that cylindrical grinder turning and thought the castings were brilliant.

I am in full agreeance with you on this Pete.
It is truly amazing that nothing is too much trouble for Michael and his hospitality towards visitors is unsurpassed.
Apologies Michael but it had to be said as you probably can't see it.
A top bloke.

In all sincerity
Phil