Mechanical vs electrical edge finder

[SLK001]

Humm, I just watched a video on using the mechanical edge finder. I always advanced the table until the edge finder ran true. If it "kicked out", then I went too far. The video says that the "kick out" is what you want - I say no it isn't, that you've gone too far. I've always wondered how much you were off if the EF kicked out.

 

[woodchucker]

measure your edge finder. Mine says it is 0.200 but it is bigger. If I remember a few tenths bigger. So I believe the few tenths is what gives you that .2 accuracy... if you want better than that, use a test indicator and find the edge.

 

[RJSakowski]

Humm, I just watched a video on using the mechanical edge finder. I always advanced the table until the edge finder ran true. If it "kicked out", then I went too far. The video says that the "kick out" is what you want - I say no it isn't, that you've gone too far. I've always wondered how much you were off if the EF kicked out.

If you advance the indicator until it runs true and take that as your edge, you are using it as a visual indicator. How good is your visual acuity? Can you see wobble of a tenth, a half thou, a thou? If you are using the edge finder in this manner, why not just use a dowel pin? I advance slowly until the edge finder kicks out. I can do this repeatably to +/-.0001". When I have checked the accuracy of locating the edge by approaching a block from opposing sides in the past, I am usually within a couple of tenths.

I just did another check of a 1/2/3 block with a 1/2" edge finder on my Tormach mill. The edge finder positions agreed with micrometer measurements of the block to better than a tenth. That would be for two approaches one from each side so the the difference per approach would be better than a half a tenth. The measurements were repeated three times with identical results.

However, an edge finder is not just plug and play if you want precision results. Attention must be paid to the spindle speed, the length of engagement of the tip and the speed of approach. Too hasty a measurement will produce poor results. The quality of the edge finder also comes into play. My favorite was a 1/2" Enco purchased in 1984. Unfortunately, I trashed it tonight by having the spindle speed set too high. I prefer to use the 1/2" rather than the .2" when possible as it seems more sensitive. That may be a function of these tired old eyes though. I find a drop of light machine oil or gun oil applied to the mating surface of the tip and then wiped off will improve sensitivity.

 

[compact8]

Thanks for all the replies. I think I will stick to the traditional edge finder for the time being. The vendor of the finder ( 4 mm tip diameter ) recommended a RPM of 400 - 600 but I have heard suggestions of over 1000. Not sure if it's related to the tip diameter ( the smaller the faster ? ) . Any comments ?

 

[higgite]

Thanks for all the replies. I think I will stick to the traditional edge finder for the time being. The vendor of the finder ( 4 mm tip diameter ) recommended a RPM of 400 - 600 but I have heard suggestions of over 1000. Not sure if it's related to the tip diameter ( the smaller the faster ? ) . Any comments ?

If all else fails, follow directions.

Tom

 

[rwm]

Humm, I just watched a video on using the mechanical edge finder. I always advanced the table until the edge finder ran true. If it "kicked out", then I went too far. The video says that the "kick out" is what you want - I say no it isn't, that you've gone too far. I've always wondered how much you were off if the EF kicked out.

I think kick out is correct. It is hard to accurately determine "running true" by looking. Having the obvious change when it kicks out is the whole point. Running true to kick out should be less than .001".
Robert

 

[RJSakowski]

Thanks for all the replies. I think I will stick to the traditional edge finder for the time being. The vendor of the finder ( 4 mm tip diameter ) recommended a RPM of 400 - 600 but I have heard suggestions of over 1000. Not sure if it's related to the tip diameter ( the smaller the faster ? ) . Any comments ?

In addition to tip mass, spring tension tension will play a part. I don't know what rpm I was running when mine flew apart yesterday. However, mine had a comparatively light spring which is probably why it seemed more sensitive. I may be able to recreate the rpm setting. EDIT: update, I was able to determine the rpm to b about 3500 rpm

However, different manufacturers use different spring tensions so it's not a one size fits all situation.

PS I was able to repair my Enco edge finder destroyed in the above post and it repeats my position measurement made with a Phase II edge finder exactly, .0000" even after shutting down and rehoming the machine. A testament to the repeatability of the edge finder as well as my electronic homing modification.

 

[Larry$]

The problem with watching for it to "run true" is, that is a judgment call, not an event.
The wobble allows you to see your progress and slow down to just barely moving the table at the point of kickout. Yes, you have gone too far but you have just barely gone too far, 10ths? The "run true" method works as long as your judgment of "true" is in the few 10ths range.
Machining has lots of built in errors and there are always ways to come closer to the magic "perfect." But never get there. A mill, even a high end CNC will have errors. Grinders can come closer to the illusive perfect. As long as I'm on a rant about what is essentially the argument over "good enough," what about the surface texture your method of choice leaves? Most measuring devices use the high spots, but they can compress or rub off.

I think my edgefinder is an import, Brown & Sharp. It seems very nicely made and repeats within a few 10ths running at about 900 rpm. My hands are likely not steady enough to do better no matter how good of a finder I use. I'll bet many of the cheaper finders can do as well. Just do a test of repeatability to satisfy yourself.

 

[SLK001]

The wobble allows you to see your progress and slow down to just barely moving the table at the point of kickout. Yes, you have gone too far but you have just barely gone too far, 10ths?

I always wanted to know how far out I've gone when it kicks out. I also don't rely on the visual. I use my fingernail to check for "the crack" between the two parts. I've looked for a white paper or something on using the edge finder, but haven't found anything.

And don't beat yourself up on running one too fast and it coming apart. I did that to a Starrett at work once and was luck enough to be able to get it back together (it was a communal tool). A hard lesson on running at the correct speed, but one truly learned.

"If the wobble is a blurry,
You're in too big a hurry!"

 

[RJSakowski]

I use a pencil or my fingernail to eliminate the wobble. I usually do this after roughly finding the edge. I back the edge finder off by about .002" and center the tip. Otherwise I would have to go out .010 -.015" to change from offset to wobble state. With a lot of wobble, you can hit the edge prematurely and change to the offset condition so centering the tip reduces that possibility. I usually advance at about .002"/sec. so backing out .002" means that I only spend about ten seconds finding the edge again. The slow advance gives the edge finder time to fully center before contacting the edge and reduces the possibility of overshoot.