edge finder speed

Advert

New Replies

Report Bug

Help/FAQs

edge finder speed

This topic has 11 replies, 10 voices, and was last updated 7 December 2020 at 11:32 by Clive Foster.

Viewing 12 posts - 1 through 12 (of 12 total)

Author

Posts
6 December 2020 at 14:54 #511990
paul rayner
Participant
@paulrayner36054
just out of curiosity what speed do you run your edge finder at and why.

I run mine at either 600 or 1000 rpm, seems to flick off better at 1000 rpm, it's a vertex made unit, and my mill is a warco super major, so only 6 speeds,

Reason I ask is I have noticed on youtube there are some videos that appear to run them slower than this and others much much faster.

Interested in your thoughts I,m NOT an engineer by the way, just a man with a shed that likes to play

Paul
Advert

6 December 2020 at 14:54 #27810
paul rayner
Participant
@paulrayner36054
6 December 2020 at 15:01 #511992
Colin Heseltine
Participant
@colinheseltine48622
Tend to run mine at around 90rpm. I just felt happier with it running slow. If the concensus of all the experts on here is to run faster they I will try that.

Colin
6 December 2020 at 15:39 #512000
Howard Lewis
Participant
@howardlewis46836
Being ABSOLUTELY bone idle, mine tends to run at 340 rpm for most of the time. The Wiggler seems to work OK>

Howard
6 December 2020 at 16:32 #512007
David Colwill
Participant
@davidcolwill19261
I run mine somewhere between 500 / 800.

I have had a few disasters by going over 1000 (forgetting to check the speed setting) and watching the end of the edge finder describe a rapidly increasing orbit around the spindle.

Fortunately the cheap spring sets sold by Machine Mart IIRC and others, contain suitable replacements.

Regards.

David.

PS

I now have a Haimer 3d taster, so will try to run this at 0 rpm
6 December 2020 at 17:00 #512010
John Hinkley
Participant
@johnhinkley26699
I seem to be in the minority here. And not for the first time.

I run my edge finders – either electronic or mechanical – at no more than 100 rpm, I would estimate. The co-axial centre finder at about 20 rpm. I can see no advantage in running them any faster, but then I wasn't an engineer by profession, either. ALL my bad habits are self-taught.

I stand to be corrected, however; not that I'll probably take any notice!

John
6 December 2020 at 17:28 #512016
Clive Foster
Participant
@clivefoster55965
Huffam specify 200 to 500 rpm for their wigglers.

I run mine at around 150 rpm and watch for the sideways walk to start rather than relying on the flick off. I claim that to be marginally more accurate on single edges. But I would wouldn't I.

Running faster gives a much more positive, and quicker flick.

I suspect that if you are using the centre finding mode on a DRO by finding both edges the higher speed method might actually be better than mine.as the over-travel before flick could be more consistent. Huffam say lesst than half a thou over-travel to fick so its all a moot point anyway.

Clive
6 December 2020 at 17:29 #512018
Baz
Participant
@baz89810
I seem to get the most consistent results at somewhere between 400-600 rpm.
6 December 2020 at 17:36 #512021
Martin Kyte
Participant
@martinkyte99762
I get the impression from reading that the ball ended edge finders were intended to be used by judging the point when the ball end ceased to move by eye. The flicking to one side technique seems to be a practice that has crept in over the years. When 'doing it by eye' slower speeds are a neccesity. The flick out technique requires the finder to have been pushed past centre to a small extent so a small error is always present for that method. My engineering gut feeling is that a faster speed when using flick out is advantageous on the basis that the device has a certain amount of friction and the faster rotation would increase the sdeways force a little couple with the action takes place more suddenly and gives less time to turn the handle that little bit more. In milling operations I cannot say it seems to matter too much, certainly not to the extent of getting something running truly in a lathe when much more prescision is required. I can't see it being much more accurate than fag paper methods, maybe even worse. Maybe someone fancies doing some experiments using dead stops and the like on a big machine with good DRO's.

regards Martin
6 December 2020 at 19:50 #512030
colin brannigan
Participant
@colinbrannigan54160
Around 1000 for edge finders and 300 for wiggler type.

Colin
7 December 2020 at 08:20 #512082
Tony Pratt 1
Participant
@tonypratt1
Posted by Martin Kyte on 06/12/2020 17:36:57:

I can't see it being much more accurate than fag paper methods, maybe even worse. Maybe someone fancies doing some experiments using dead stops and the like on a big machine with good DRO's.

regards Martin

Martin, the much admired Geo H Thomas covered this exact subject in his book 'The model engineers workshop manual' from page 29 onwards.

Tony
7 December 2020 at 11:32 #512127
Clive Foster
Participant
@clivefoster55965
Martin

Huffam claim better than 0.0005", half a thou, accuracy for their version of the wiggler when used in flick to one side mode. They say the exact distance by which the probe needs to be pushed past centre before it flicks is indeterminate and dependent on conditions but will always be less than that. Under constant conditions repeatability is excellent. I'd say the scatter is less than half the claimed accuracy but all this is well into the range where sensible measurements on our sort of workshop machines isn't possible.

Realistically accept within half a thou omnidirectional error for single edge determination. Significantly less than half that if you have a DRO and use the centre finding function to work off both sides. Theoretically the error then becomes the variation in scatter as all the systematic ones should cancel.

Huffam claim better than 0.0002" accuracy can be got by using the line of light method looking along the contact point between probe and workpiece. When the probe is slightly offset from true running the light passing through the gap between probe and work will flicker. The out of line probe alternately closing and opening the gap for each 180° degree rotation. When the light stops flickering the gap is said to be 0.0002" or 2 tenths of a thou.

Needs better eyes than mine!

Suggests the minimum offset needed to flick is about 0.0002".

But the Huffam is better engineered than the common ball in a collet set ones with much lower friction so, presumably, needs less offset to flick.

I think you'd be working hard to get reliable results of that order using the fag paper or pull a 1 thou feeler gauge out methods.

The great thing with the wiggler is that its quick, easy, very clear and pretty darn close.

Even the cheap set ones are very repeatable. As I recall things I simply dialed in a 5 to 10 thou compensation when I was using a cheap set version by adjusting the effective ball diameter. I found the error held well so long as I didn't fiddle about with the adjustment or change things. If I did change things I needed to re-calibrate. As I recall it the cheap set was reliable to approaching a thou and a half, maybe better, back then. I suspect I'd do better now having learned to exploit the Huffam.

Clive

Edited By Clive Foster on 07/12/2020 11:32:44
Author

Posts