edge finder sensitivity

No thats not so. Take a couple cases and see what I mean: If you happen to be at exactly 90 deg to the high/low axis than it will be exactly the same at both tests. If half way between the high/low and the zero/zero (say

45 deg) then the amount high on one side will be the same amound as the low on the other. So any case it will the mean of the two tests. ...lew...

Wolfgang:

You wouldn't need to know, or be, on the maximum high point of any eccentricity to average out the runout in one axis. Let me try to explain it this way. Let's say the edge finder was bent toward the X+ side by +.001, so therefore the X- side would be bent away by -.001. Now when you touched the X+ side, (recorded your reading) then turned the edge finder 180 degrees and touched it again you'd have theoretically have a .002 difference. So half that difference would be roughly the center of the runout and spindle. Now the Y+ side of the edge finder could be +.005 and the Y- side could be -.005, BUT they don't necessarily influence the amount of runout in the X+ and X- sides. Was that clear as mud, or did it help?

One thing that "could" influence the measurement of runout would be the the angle of the stem. If the very tip was touched on one side, and the other side was touched further up the stem due to being bent over... then the true runout would be skewed. Is this type of situation the ball tip electric edge finders would nullify any "bent tip" inaccuracies.

Unka George:

It might pay to indicate your electric edge finder from time to time, and tap it to get it centered if it runs out an appreciable amount. As far as the needed accuracy for your job, that depends on the job. Some can be scaled, some can be within a few thou, and some need to be within .0005. In the latter case I wouldn't count on the accuracy of electric edge finders, I'd either use a regular edge finder or indicator sweep your part and move over half the distance of your part.

anyone have comments on /experience with the ones from Flexbar?

I never felt the need to buy an edge finder more expensive than a good Starrett or B&S. So with the knowledge of me having no direct experience with Herman Schmidt or Flexbar edge finders... my OPINION would be that your money would be better spent in another area. IF you need better accuracy than a good Starrett edge finder can achieve (around ±.0005) then you need to consider using an indicator to find the edge of your part, either by "sweeping", direct indicating the edge, or using a "chair".

In using those you have to make sure the wall you're edge finding is perpendicular to your table or you might get an angular error.

Bottlebob:

Excuse the 'senior moment' on my part regarding the 90 degree babble. You are of course correct.

As to geometric accuracy of the various features: Anyone who reads this ought to understand that attention to spindle squareness to the table, straight tool, work ref. surface finish and squareness is mandatory for accurate results. The latter is particularly important when using the 'chair gauge' you referenced. No sense in measuring the surface finish of a plowed field in micro-inches:-)).

Jig borer operator in my training grounds used a gauge block held to the side of the work. First they swept the work edge and then the gauge block surface with the indicator. When the indicator needle stopped moving the spindle centre line was dead-nuts over the reference edge, say within .0001 or so. The beauty of this method is that no tooling errors are introduced into the process. Theoretically it may be used to achieve any accuracy required.

Wolfgang

Wolfgang:

Is that really your name? It's pretty cool if it is. Anyway, we all make the occasional conceptual boo-boo, it's no biggie.

Actually, I've got a magnetic "chair" and the "corner finder", but I seldom use them for the very reasons you mentioned above. Possible out of square issues.

Correct, and that particular method was what I was referring to when I mentioned "direct indicating the edge". And of course a "tenth" indicator is sometimes used when the tolerance requires it.

I did a little comparison test at work today. The above process, outlined by Wolfgang, I'm labeling as "Direct Indicating of the Edge". Which I compared to a kick-out edge finder, and using a "Chair" tool.

I was edge finding the hardened & ground steel solid jaw on a Kurt vise. I indicated the vise jaw parallel to the machine's ways. Then I indicated it vertically to make sure it was square (the indicator needle did not move by going down in Z for the length of the stem (.0005 Indicator 1 1/2" dia. dial).

First, I "direct indicated" the jaw with a Jo block as explained at the top of this post. It took 4 min. 50 seconds. A lot of that time was used to get the indicator precisely centered vertically AND moving the machine a "tenth" here and there to get the needle to be aligned to the zero within the limits of my vision. For the sake of this test I'm going to assume this process as the most accurate of the three, and will assume the location it displays as true zero within the tolerance range of the indicator and the process itself (indicator is 4 months old).

Second, I used an edge finder with a .200 tip and .500 body. Moving over in .010" increments, then .001" increments, then .0001" increments until it "kicked out". This took 2 min. 38 sec. All these times include the housekeeping chores in the control (Fadal) for setting fixture offsets. The difference in the final zero position of the edge finder compared to the "direct indicated" zero position was .0004".

Third, I used the "Magnetic Edge Finder" tool (upside down chair looking item), with the indicator reading zero. This took 2 min. 5 seconds. The difference between the "chair" zero and "direct indicating" zero was .0003". That could be due to any number of factors, but I would assume there may be a tolerance in the manufacturing of the "chair" device.

This was not meant to be a definitive/exhaustive comparison of these processes, just a quickie test to see if there were any gross differences. The Z travel of the head could have off a slight amount, but I think that unlikely since the machine makes reasonably accurate close tolerance parts on a regular basis. All tool tapers (and spindle taper) were wiped clean before beginning.

I've crossposted this to amc, since it's relevant in there as well.

Bob, I wonder if the Herman Schmidt edge finder, with lapped surfaces, would be any better? I guess Haas has linear ways so sticktion would be minimal? I wonder when your moving in tenths on a boxway machine what the results would be?

Best, Steve

Steve:

I suppose we'll have to await the results of Jon's test with one of the new Herman Schmidt edge finders they just got.

I've put a tenth indicator on the Fadals & Haas, and after any backlash is deal with, they both move a tenth at a time when you are moving in one direction in the mid range of my indicator. At the ends of the travel of the indicator there may be a cosign error due to stem angle.

Ive got a starret and a herman. The herman definetely kicks over quicker. The starrett seems like it has to make its mind up before moving. I wouldn't want to edge find to a tenth, but the herman looks like it would hold it if the conditions were constant. But for some reason...I like the starret better. the herman is big and gay, and the starret feels like an edge finder? lol

One thing, if the starret doesnt get used for a year or 2 i have to "lap" it back and fourth a few times, after 10 years of no use, the herman still feels like i just bought it. SuweeeT!!!

Bob, have you ever thought about picking up one of these:

From what I've seen, it's only a matter of seconds to get a good quality reading for X, Y, and Z. I've never purchased one, because I've got MP700 and OMP40 probes coming out my ears, but it would certainly be my first choice if I didn't have a probe.

Vinny:

Just speculating now, but both the slow reaction and the "stickiness" of the Starrett may be due to the congealing coolant. Since you said that's the edge finder you use more frequently. The Herman may move freely, and like new, since you don't use it.

I'd be interested in the results of a comparison between "direct indicating" and Herman S. edge finding by you.

I've got a Zero Master Joe and they are deadly accurate. I'd originally intended to put it in a collet chuck but the local rep. cut me a deal on one of their HP shrink fits and I'm using that instead.

I also use it for in process part probing.

Joe:

I was THIS close to getting one (holds finger and thumb .125" apart). BUT, how can you be sure the thing is perfectly on spindle centerline every time you put it in. And they say it's accuracy is .0004 on their website. My little test showed similar accuracy with an edge finder. Although, edge finders don't do Z, I'm not sure how often I'd use that feature, since for comparative Z depths I just use an indicator and the control's Z readout.

Once you get it set up in a high quality holder don't take it out and M19 the spindle when you load it.

You'd be surprised Bob. I use mine like a probe with linearized drill cycles for probing code. You can run around a part checking depths in a hurry and probing the bottom of a counterbore or a surface through a hole isn't easily done with an indicator even when you can get the tip in.

You can borrow mine for a while if you'd like to try one out for a while.

BottleBob,

Great report on the comparison of 3 different ways of edge finding. I think this is the first time I have seen this done.

You also saved me some $$$ as I was going to buy the "chair" tool; now I think that the difference between the kick-off and chair tool accuracy is not sufficient to warrant the expenditure.

It's also nice to have my assumption confirmed on the .0005" accuracy of the kick-off tool, and that the point of accuracy IS the kick-off, not the observed concentricity prior to kick-off.

The ultimate test for accuracy of each method would be to locate the edge using each method, move over the work piece by say 1.0000" and drill, bore, ream a .5000" dia. hole. By measuring the remaining ligament between edge and hole the accuracy of each method would be established. Granted this is being anal about this process but I find it interesting.

As it is your tests prove the relative accuracies between processes and, assuming that the error in direct indicating is +/- .0001" based on your new indicator, you have given us valuable NUMERICAL evidence of each process' capability.

Thanks,

Wolfgang... Yes it is my real Christian name.

Gunner, what other middle names do you have, you said ONE of them was... :-)).

Has anyone actually measured the 0.200 tip to see if maybe they make it a tenth or three undersize, so that the kick-off is the correct indicator, and not concentricity? Just curious. Oh, and thanks for the great report.

-- Regards, Carl Ijames carl dott ijames aat verizon dott net (remove nospm or make the obvious changes before replying)

BottleBob poll, how many of the new edge finders does Jon break before he gives up?

Tom

Thanks for sharing the actual numbers.

Unka' George [George McDuffee]

------------------------------------------- He that will not apply new remedies, must expect new evils: for Time is the greatest innovator: and if Time, of course, alter things to the worse, and wisdom and counsel shall not alter them to the better, what shall be the end?

Francis Bacon (1561-1626), English philosopher, essayist, statesman. Essays, "Of Innovations" (1597-1625).

Greetings George, I've bought a few things from Flexbar over the years and everything has been worth the money. Parallels were more accurate than advertised, being to size and parallel within .0001". As far as sine bars go, I have a short one that hardly ever gets used, but the 5 inch one gets lots of use. Very easy way to get accurate angles. One method I use that's real common is to clamp a toolmaker's vise in the mill vise with the angle being set with a sine bar. Then part after part can be machined at the correct angle. Cheers, Eric