Best Centering device - coax, zero set, etc

I broke out my blake coax to quickly locate a couple centers and I notice that the arm was moving up and down almost 1/4"!! I don't mean the normal up and down movement of the body. it was centered so there was no body movement at all. that arm was moving like it was attached to a bent shaft.

So I got to checking. spindle, 0.0000 indicator run-out. brand new techniks holder, first check on the shoulder + .0015" TIR. hmm.. maybe that's it. check inside the collet seat to make sure. .0003TIR. (how can the shoulder be that far off? i digress, the holder is good.)

so put a collet in with a dowel pin, and I get about .0005 TIR.. about right considering the extra length from the collet seat. But then I put my blake coax in the collet leaving enough room to indicate the shaft, and now I'm throwing up almost .002TIR JUST BELOW THE COLLET? Then I check down on the metal piece just above the body and i get a little over .004TIR

WTH..this thing is brand new. (and over 400 bucks!) It sits in my box, and probably has been used less than a dozen times. never been crashed. I'm positive. (unless someone took it out my box and used it without my knowledge.)

So here-in lies the problem. I just drilled about 400 bolt holes using that thing to find center. Thankfully there is .020 positional tolerance, but it appears to me that any error/bend in the shaft will be multiplied at the indicator, since you now have to move the machine out of position to offset the bend. (should be a 1:1 move I beleive.) Who knows how much that value was, but I'm putting it around .008 out of position. But I yet to measure the part. This explains so much why a couple jobs I did in the past, just "felt" like they were ever so slightly off, even though it was just barely or impossible to measure.


Obviously a zero set is going to be better because you are directly linked to the spindle axis, but what do you guys use when you cant get in there. or you don't have enough room to spin your indicol. I guess I could use my 3d Taster to touch the sides of the hole. do that twice per hole and your gonna be very accurate. but even then, sometimes its to big. So throw some years of experience at me. what Have you guys found to be the best. and has anyone experienced on of these blakes to be bad almost right out of the box? (i should of checked it. out of warranty now)

What is the difference between what you Blake says and what a regular old test indicator say? Put the Blake in, find center, don't move XY axes, put the DTI in check, what's the difference? Should be way less than .001".

Do the same test, but move the Axes one at a time, according to your Actual Position, check deviation there.

A Blake basically works by pushing an indicator up and down its spindle. Far as I can see what matters is that the face that the butterfly runs on be properly perpendicular to the machine spindle. Its a zero sensing device so any internal spindle TIR and other errors inside the device effectively resolve themselves as a fixed change in the feeler end position relative to a nominally perfect Blake with absolutely no inherent errors. Obviously said errors are assumed to be small. As the feeler goes round with the spindle any TIR produces a fixed offset relative to the feeler end regardless of rotation angle. So when the indicator is zeroed out the reading is true.

If the face the butterfly runs on isn't truly perpendicular to the spindle the fundamental error is the same as using an Indicol with the head out of tram.

Clive

The only problem I've had with the Blake is people abuse them and they get a dent on the body that causes a riser. Then every time the butterfly passes that riser the pointer jumps, at a hundred rpms it's a lot where at ten rpms it's not so bad.
If you have a bent shaft the body will not slide up and down as it's supposed to and if it gets coolant on it it can stick also.
We used to have one for every 2, 3 spindle machines; for centering a hole tested they were as accurate as a regular .001" finger indicator.
I'm wondering if the acorn nut is just not tight enough for it to work, sometimes things need adjustments instead of deciding to throw the tool against the farthest wall.
Dan

Clive603 said:

A Blake basically works by pushing an indicator up and down its spindle. Far as I can see what matters is that the face that the butterfly runs on be properly perpendicular to the machine spindle. Its a zero sensing device so any internal spindle TIR and other errors inside the device effectively resolve themselves as a fixed change in the feeler end position relative to a nominally perfect Blake with absolutely no inherent errors. Obviously said errors are assumed to be small. As the feeler goes round with the spindle any TIR produces a fixed offset relative to the feeler end regardless of rotation angle. So when the indicator is zeroed out the reading is true.

If the face the butterfly runs on isn't truly perpendicular to the spindle the fundamental error is the same as using an Indicol with the head out of tram.

Clive

Click to expand...

That makes sense to me... but my hole are about 6 thou out in 2 axii...honestly still trying to wrap my head around it. Going back at it in the morning.

Motorsports-X said:

...honestly still trying to wrap my head around it.

Click to expand...

That sort of thing is certainly guaranteed headache territory.

Tests that occur to me :-

1) Spin it up without a finger. Assuming the butterfly is in contact with the ring, so the dial gauge has pressure on it eliminating lost motion, zero movement proves the ring and internal spindle to be perpendicular within the capabilities of the gauge.

2) Set up a suitable hole as per normal. Align with the Blake dial facing front and centre. Check alignment with the dial facing 90° left, 90° right and straight behind. If the Blake is defective or damaged error will vary with dial orientation. Might be possible to figure out WTHIGO directly but I'd move direct to test 3.

3) Lock the tables and bore a hole. Test with the Blake. Hole will be central to the spindle so forms perfectly aligned test ring. Blake should read zero when spun. If it doesn't the polar position of the error can be mapped when turning by hand.

Test 3 is sufficient on its own to prove device error but I think the other two provide some potentially useful insight.

If it is a bent spindle it should be possible to correlate measured TIR with error as a Blake indicator is, supposedly, correct to nominal calibration with a 2" finger.

Good luck.

Clive

You say that "the arm was moving up and down almost 1/4"". I assume that you mean the horizontal arm that is used to prevent the body of the Blake from rotating. That arm should not move any appreciable amount in any direction. The fact that it is and the amount (1/4") says that something is seriously wrong. This motion does tell you that the Blake is not working properly. 1/4" is 250 thousandths and that arm appears to be about 8" long. The surface that the indicator arm rides on is about 2" in diameter or 1" radius so a ratio says that it must be wobbling by about 250/8 = 0.031". That is a lot.

The first thing I would do is to look for obvious problems, like a burr on the mounting stud. Rotate the Blake in your hand and see if you can feel any roughness: it should be very smooth.

Put a 10" piece of drill rod or shaft in the same collet that you use to mount the Blake. Check the run out near the collet and at the far end, about 8" from the collet. If it is only a few thousandths, then your collet is not at fault. If the problem is in the collet you should get about that same 1/4" of movement at the end of that piece of rod.

If these simple checks don't give you any answers, I would use a DI to try to see just where the center of the wobble/rotation is. Mount it in a good collet for this. You can check the motion of the body of the Blake at several points, top, middle, and bottom. See if you can find any point where it does not wobble. Plot this motion on a graph and it should tell you where the wobble is centered. This may be at or even above the mounting stud. But it can be inside the Blake.

I suspect that you do have a bent mounting stud. Or a bad bearing inside the Blake. It should be a warranty issue and I would contact Blake or the seller before attempting any repairs on your own.

I have used Blake indicators frequently. I always eyeball it in to my bore or boss by spinning it by hand first. If you are way off on centerline it will bounce up and down. I try to get it within .005 or .010" before I turn the spindle on which I have never had an issue with. Getting centerline within a 10th is always fast and easy for me with the Blake. I used to dial in the 10 cell 5C collet fixtures from Beere and the Blake made checking all 10 positions quick work.
OP, how familiar are you with the Blake set up and operation? You know to use the straight stylus for IDs and the curved for ODs? Do you know how to adust the preload for different Diameters?

All of the above is why there are no coax's in my tool boxes.

JH

James H Clark said:

All of the above is why there are no coax's in my tool boxes.

JH

Click to expand...

Agreed.


So I figured out what happened.


I talked to a friend (machinist)that worked with me at the last job,l and I told him how pissed I was that my brand new Blake was giving me errors. . He said you know, back when we were at Blackwell, the programmer that I let use it once, turned it on with it to far out of center and that it didn't have enough room to compensate. He said he remembered because the guy actually asked him if he might of damaged it. That makes sense, these things can bind and bend the shaft if you have them to far out when you turn the spindle on. He told me that he responded that he didnt think it would hurt since the feeler can move out of its own way.. but he didn't know it stopped at center.

So at least I know what happened now. I cant believe the guy never said anything to me. (at the time i was working in the office, and only out in the shop occasionally) I'm about to call him and ask for 60 bucks. Anyways. I ordered a centro from Haimer yesterday. Ill send the blake out for repair and then put it on ebay. I really was seriously, seriously disappointed in American Craftsmanship for a good few hours there. I take very good care of my tools and try to buy high end stuff, so I couldnt for the life of me figure out what was wrong.

Blake has its place in the shop, As long as it is not abused and used with some degree of common sense it is a valuable tool. I have two and never had one give me bad info, one is @30 years old. Don't expect your new Haimer to allow you to indicate a center punch point.

I have a blake, only used it 2-3 times in 8 years or so. It is way too long. To use it you need to lower the knee, that is too much work even though I have a Servo power feed on the knee. I can center it up so fast with an indicator I don't even take the blake out of the drawer anymore.