Electronic edge finders - any recommendations?

[kevperro]

Anyone using the laser?

http://littlemachineshop.com/product...2604&category=

[keen]

Hi - I now only use rotating mechanical edge finders -The Starrett one is good - and a paper feeler for the Z. . On my elecronic height setter - sometimes I would get a persistant micro 'hair' of conductive material that gave false readings. Too dangerous for very accurate work.

[bevinp]

Thanks Guys, that was good information. After following the tips and links I will probably go for an electronic with overrun protection. But might buy both.

That mechanical one from H Schmidt looks very good and and seems to have a better design than the Starrett so should perform at least the same. Thanks Michael.

I found the debate (elsewhere in CNC) about exactly when the mechanical edge finders are indicating the correct position, to be a little confusing. The description provided by Starrett is a little ambiguous and probably gives rise to the two differing opinions. One is that the correct position is where the tip snaps into running true and the other is where the continued movement snaps the tip out of true. I am inclined to think it should be the mid position between those two points, with a bias towards the first. And the width of that gap would depend on the friction of the edgefinder tip to body (and spring pressure) and the friction of the tip rubbing against the workpiece. All of which would vary between edge finders, RPM, workpiece material and any lubrication. But I assume the gap would be small and therefore satisfactory for my needs.

No doubt the expensive ($500+) 3D sensors would be excellent but too expensive for me.

And the electronic finders seem to be fragile and most without reasonable cost of repair (but good for salvaging for other purposes!). But their apparent ease of use and consistency of errors that can be accommodated, makes me favour their mid price solution. Now all I have to do is chose the brand and the supplier.

SPI told me their electronics don't have overrun protection. Fowler ads say theirs do and I am investigating the other common names.

Thanks for the help.
Bevin

[bevinp]

Originally Posted by kevperro Anyone using the laser? http://littlemachineshop.com/product...2604&category=

Yes I have the $50 laser and use it for woodwork or bench drilling but it is not accurate enough for mill work. The beam diameter appears too large and because it is a laser, you get the strong reflections which makes it hard to assess when you are on the correct point. Even when I try to align a centre punch mark I get reflections from the side of the indentation and I can't "see" to centre the beam. Hence the manufacturer's advice to paint the surface blue to reduce the reflections and make the beam just illuminate not dazzle. The dimmer (polariser) does reduce the diam/dazzle somewhat but reflection still occurs... it is the nature of a laser beam.

It works better on a sharp edge with a clean vertical surface as this splits the beam into two and makes it easier to assess equal reflections from top to bottom surfaces.

But it certainly does not provide accuracy within a few thou.

[kevperro]

Originally Posted by bevinp Yes I have the $50 laser and use it for woodwork or bench drilling but it is not accurate enough for mill work. The beam diameter appears too large and because it is a laser, you get the strong reflections which makes it hard to assess when you are on the correct point. Even when I try to align a centre punch mark I get reflections from the side of the indentation and I can't "see" to centre the beam. Hence the manufacturer's advice to paint the surface blue to reduce the reflections and make the beam just illuminate not dazzle. The dimmer (polariser) does reduce the diam/dazzle somewhat but reflection still occurs... it is the nature of a laser beam. It works better on a sharp edge with a clean vertical surface as this splits the beam into two and makes it easier to assess equal reflections from top to bottom surfaces. But it certainly does not provide accuracy within a few thou.

That is good input. I'm using an el-cheapo SPI manual edge finder now. I can get repeatable results. I don't know how accurate they are but I suspect they are +/- 0.001

The laser would have the advantage of not needing a turning spindle.

I found directions somewhere on the web that suggested a 1000 RPM spindle speed as the "right" speed for the mechanical edge finder. Tormach says the transition between center & offset is zero. If the part isn't perfectly smooth or if you have an angled surface that probably throws you off more than any error you make judging what to call centered. I've been using the step before it pops out as my reference.

[bevinp]

[QUOTE=kevperro;615304]
The laser would have the advantage of not needing a turning spindle.
QUOTE]

You are right when you say that but since I find it difficult to see exactly where the centre of the beam is, I am not confident that when adjusted, it does not have an error. Consequently when I do use it, I am probably doubling the error.

I do need to wear glasses so perhaps that may be adding to the dazzle. Others may not find the same problem.

[keen]

"I found the debate (elsewhere in CNC) about exactly when the mechanical edge finders are indicating the correct position, to be a little confusing. The description provided by Starrett is a little ambiguous and probably gives rise to the two differing opinions. One is that the correct position is where the tip snaps into running true and the other is where the continued movement snaps the tip out of true. I am inclined to think it should be the mid position between those two points, with a bias towards the first. And the width of that gap would depend on the friction of the edgefinder tip to body (and spring pressure) and the friction of the tip rubbing against the workpiece. All of which would vary between edge finders, RPM, workpiece material and any lubrication. But I assume the gap would be small and therefore satisfactory for my needs."

The best way to obtain high accuracy is to make the cad/cam X/Y axis the part centre line, not the edge. and set the work or stock up central. that means you zero on wobbler spin "out' and do the same on the opposite side, It is very repeatable. and enter half the value in that axis dro - same for other axis. This is very accurate - you can be in position better than a half thou.

[bevinp]

"The best way to obtain high accuracy is to make the cad/cam X/Y axis the part centre line, not the edge. and set the work or stock up central. that means you zero on wobbler spin "out' and do the same on the opposite side, It is very repeatable. and enter half the value in that axis dro - same for other axis. This is very accurate - you can be in position better than a half thou."

That sounds good. Since you are doing two opposing contacts you are automatically compensating for the diameter of the contact pin as well as for any consistent error in the technique. And if you can accurately measure by micrometer the actual width of the workpiece then you will be able to determine the actual allowance you need to make for those occasions when you can only contact one edge. As we say is Australia.. "Bloody good mate".

And the more times you do the two side contact and check the actual measurement, the better average allowance you can determine, even get differing average allowances for different materials. And as the edge finder gets older and has altered characteristics, the two side technique will continue to compensate. As my kids would say.. "Cool".

And sometime when I was a kid I think we would say "Keen".

[keen]

Cheers - wish I had stumbled on this method years ago - so much unnessary grief trying to get high precision battling inconsistant errors. Here is a good double check process - worth the few extra seconds. touch face with steps, zero dro on wobble, up and across, down and do opp side. Divide DRO value by 2 on calculator - enter this value in DRO (check if it needs + or -). Now the double check. Zoom back to the first side and wobble - check it is the same as the value still on your calculator.

[tobyaxis]

Originally Posted by bevinp Toby, Yes I have a mechanical one that I could never get to work properly, until I saw a Youtube of one working correctly. I realised that mine is poor quality and on closer examination I found that the .200" dia end is not square to the body and has significant runout at its end. Also it is out of round by about 0.0005". I thought it was a little cheap when I bought it!!!! I am hoping that an electronic edge finder (a good one) will be easier to use (not require spindle running). Did you use yours much before it broke? What brand was it? Did it have the overrun protection? Bevin

It wasn't mine thank god, it was the cheapo companies. They didn't or couldn't find any qualified help and thought it would compensate for their less experienced employees, LOL. When I was hired as the Leadman they told me I had to use their tools. Needless to say their tools were cheap and already destroyed. I brought my Tools in the next day.

I believe it was a Fowler.

[keithorr]

I use the Fowler above. Works fine. Doesn't solve the problem of poor part design, layout or fixturing. You still need a bombproof process to ensure success.

I use the method above of splitting the difference between sides with an oversize part blank if needed or dedicated fixtures for production parts with a known relative position touch off spot on the fixture (top, front, left)