using a wiggler?

I’m working on a friends mill that's a little worn and well used so I picked up a wiggler as an edge finder. I’ve not used one before and was a bit disappointed at the accuracy, I’d seen a few things online that said it was good down to a few tho, better if you get a good feel for it.
So it took me a while to track down the problem and it turned out to be how tight the little collet on the wiggler was tightened. I was getting anything from 3.04mm to the edge when it was loose to 3.22mm to the edge when it was tightened up.
There's no obvious way I can see to set a repeatable tightness outside of feel. I looked up a few videos and threads but no one seems to mention this, am I missing something?

Neil,

Welcome to the forum.

I can't help much with your problem because I have never considered using a wiggler as a precision edge or center finder. There are so many much more accurate devices out there. Any chance you can pick up a more accurate type? Although I've switched to a Haimer, I often use a Starrett 'off center' type edge finder and it's usually accurate to .001 or less. I do have a wiggler style and it's still in the box.

This doesn't help you with your question..sorry, maybe someone will pop in with some first hand info.

Stuart

The looser the stylus is in the socket, the less deflection is required to flick off, meaning higher accuracy,
Also allowing a longer pause between progressions to the next division will allow the wiggler more time to flick off which may mean higher repeatability.
You didn’t mention revs used either so I’m assuming your using between 500-1000 RPM?

Wigglers aren’t that precise so f you expecting better then 0.03/0.05 positional accuracy it’s better to use a lever type indicator to be sure

Luke.kerbey said:

The looser the stylus is in the socket, the less deflection is required to flick off, meaning higher accuracy,
Also allowing a longer pause between progressions to the next division will allow the wiggler more time to flick off which may mean higher repeatability.
You didn’t mention revs used either so I’m assuming your using between 500-1000 RPM?

Wigglers aren’t that precise so f you expecting better then 0.03/0.05 positional accuracy it’s better to use a lever type indicator to be sure

Click to expand...

I hope you are using mm.....

The only wiggler worth a damm for accuracy is the Huffam type.

Cromwell Tools - Experts in Hand Tools, Power Tools and PPE .

Spring loaded nylon pivot for the ball gives really consistent action. With care, a DRO and a certain amount of low cunning better than ± 0.0005 thou accuracy is practical when looking for the centre of a workpiece. If looking for the centre there is a certain amount of error balancing so its better to define the centre and work off that, rather than simply finding an edge and subtracting the ball radius. Biggest limitation with these things is the personal and surface finish factors which define just how far you have to go past the true spinning probe alignment before the thing crawls sideways. When centre finding the personal and work surface factors should cancel out.

Theoretically the collet type can do pretty near as well but life is too short to sort out exactly how tight the collet should be, too loose and its unstable. Not to mention the near impossibility of getting a repeatable set up. Before I discovered Huffam I spent ages getting a collet type dialed in pretty good, maybe ± 1.5 thou on a high concentration day, and resolved never to touch the adjustment again.

Bought another set when I needed the sharp pin. Which actually does benefit from friction so it stays put once trued up with finger-nail or (cheap) combination set ruler.

Clive

Thanks for all the replies guys.
It sounds like that's just the nature of the tool then. This mill has not been well treated so I thought some kind of spinning edge finder seemed like a good idea to avoid any discrepancies in the spindle/chuck/collet. Once I’d set it up on any given loosness repeatability was within .02 or .03 at the most so there’s an accurate tool hidden in there somewhere.
I set a gauge block up in the vice, swept both sides with a dial indicator in the collet to find the center and zeroed an edge then brought then wiggler in and found the distance from the edge it spun out at. I guess I could just leave it like that, maybe even put a drop of thread lock on and then at least I’d know where I am for a while though I suspect it’ll loosen over time, it's just a cheap one.
A few tho is good enough for anything I do.

I've been using wigglers for a very long time. They're pretty useful on larger machines where seeing the kick of an edge finder from far away is difficult. Always leave the collet nut loose when locating with a wiggler. The only reason it's even able to be tightened is for when you might want to use the indicator stem (I've never used one in all those years). The collet fingers have enough built-in spring to hold the ball well enough with the collet nut loose on every wiggler I've ever used (all Starretts). Also don't forget that if your spindle has runout your accuracy will suffer - when the high side hits the work there will be extra force exerted on the ball which can cause it to kick early. Really they're not intended to be super-precision locating devices - just intended to get you located ballpark. From there you should measure to a feature and finalize your location if you need higher accuracy.

Clive603 said:

The only wiggler worth a damm for accuracy is the Huffam type.

Cromwell Tools - Experts in Hand Tools, Power Tools and PPE .

Spring loaded nylon pivot for the ball gives really consistent action. With care, a DRO and a certain amount of low cunning better than ± 0.0005 thou accuracy is practical when looking for the centre of a workpiece. If looking for the centre there is a certain amount of error balancing so its better to define the centre and work off that, rather than simply finding an edge and subtracting the ball radius. Biggest limitation with these things is the personal and surface finish factors which define just how far you have to go past the true spinning probe alignment before the thing crawls sideways. When centre finding the personal and work surface factors should cancel out.

Theoretically the collet type can do pretty near as well but life is too short to sort out exactly how tight the collet should be, too loose and its unstable. Not to mention the near impossibility of getting a repeatable set up. Before I discovered Huffam I spent ages getting a collet type dialed in pretty good, maybe ± 1.5 thou on a high concentration day, and resolved never to touch the adjustment again.

Bought another set when I needed the sharp pin. Which actually does benefit from friction so it stays put once trued up with finger-nail or (cheap) combination set ruler.

Clive

Click to expand...

So this truing with a finger nail, it can/will slice between meat of thumb and nail, with all respect, i recommend a scale be used instead of thumb nail .
Gw

Greg White said:

So this truing with a finger nail, it can/will slice between meat of thumb and nail, with all respect, i recommend a scale be used instead of thumb nail .
Gw

Click to expand...

Heh, heh thought that truing with the finger, or thumb, nail comment would get a response.

Done the way I was shown the contact between nail and (slowly) rotating pin is somewhat above the point, perhaps 1/4 of the way up, so there is little danger of impalement.

Advantage is that the point is spinning in free air so any deviation is obvious. Scale or other solid pusher methods generally operate so close to the point that its hard to see whats happening. Much less sensitive control too.

Process is to start by milking the pin almost straight with a swift downwards sweep of the hand guiding the pin between finger and thumb. Then finish off by making as sort of modified fist with the side of your hand and thumb firmly supported by workpiece, vice, table or whatever. Lever your finger against the end of your thumb bringing the nail into contact with the pin to ease it into final alignment. Done that way, it's very hard to actually move the pin far enough to send it berserk unless you deliberately try as you only have about 1/16" or so of movement available at the nail. Hafta be careful where you put your hand down of course.

Direct muscular feedback against (slightly) squidgy flesh makes for very sensitive movement. Try against a tenths thou gauge if you don't believe me.

Be sensible to slide a bit of stripped off wire insulation or similar over the sharp end as protection for the first couple of practice runs.

If you want to live dangerously and spread red around there is always the good old sticky pin. Needle in a lump of plasticine on the end of the cutter. Yikes! Something I'm far too pussy to use.

Clive

This is a cheap no brand one off eBay. Maybe that's part of the problem as it cant hold its center when too loose. The range in testing was 3.04 - 3.22mm to the edge so 0.18mm or about 7 tho, I need tighter tolerances. This is down to the collet adjustment though.
Repeatability from the point at which it kicks out at any given stiffness is within .01-.04mm or about 1.5 tho, good enough for me as long as it dose reliably stay under the .04mm mark and I ran it back and forth dozens of times during the testing and it did stay comfortably within that. I can always use the dial test indicator if there is an occasion I need to be more accurate.

I was running it around 1000 but I did the tests at a range of speeds and found no difference in the repeatability or the kick out point at any given stiffness due to the speed.

I did look at the cylindrical spinning ones, I’ve never used one of those either but given that they work on the same mechanical principle I presumed the accuracy would be the same. I chose one of these spinning types as this is an old mill, an old cheap mill with worn work holders and such and I was under the impression that these spinning ones are unaffected by any run-out, within reason, always finding the center of rotation of the spindle. This is not my set up. The chap has let me at it to finish a project, he knows to do the job I’ll do some minor repairs and dial the thing in as close as can be expected leaving it in a better state than it was.

I don't like the wiggler much at the moment, it just seems too fiddly and time-consuming but who knows, maybe being forced to use it for this job I’ll end up loving the thing and spend the cash on a decent one, if there was a reliable stiffness position as in the better Starretts ones eKretz mentions that would make a big difference and I have used the little pointer for lower tolerance cuts, surprisingly accurate.

Neil

If you are spending hard earned cash the only wiggler worth having is a Huffam. Whatever the label the collet type ones just don't work well enough as the collet held operating principle is unsound. Intrinsically too much and too variable friction.

1000 rpm is way too fast. I run my Huffams down around 100-200 rpm and look for no more than 10° of sideways walk.

Got a shop made (not by me) Huffam clone in a really neat CNC'd case lying around taking up space in the E-Bay one day (but life is too short) pile which I'd let go for very few sheckles. Nearly as good as the real thing. PM me if you'd like it.

The cylindrical ones rely on having a very high finish to the sliding joint for smooth easy movement with an accurate 90° between the faces and cylindrical parts. Good ones are excellent but yet another thing where cheap doesn't cut it. Mine is good, not cheap but I don't use it as I prefer the wiggler.

Clive

YouTube

Thanks Clive, I should point out that I'm in London and I'm not sure where you are but I'll PM you, though I'll have to follow up after work tonight.

I think people are getting "wigglers" and "edge finders" confused.

I don't know about that, but personally I've never had a wiggler locate to more than a few thou' off, and usually they're closer than that. Most of the time within a thou' or two. Repeatability has always been way better than what I'm seeing described above. Maybe time to invest in some better tools. I would look into the offer above.

Wigglers can be accurate. They must be in good condition (not very worn) and made by a good tool builder. Cheap ones and used ones are never a good bet as far as buying. I agree it might be best to find something more accurate though.

I have to agree. I don't think that the term "wiggler" is precisely defined and it can mean different things to different people. I have a set that is described as a "Center Finder". It includes a number of tips and is similar to this:

Single End, Center Finder Set Mechanical 86425279 - MSC

Perhaps there is a way to find edges with it, but I do not use it for that. I use it to locate on a punched or scribed point on the top surface. I usually use the tip with the needle point for that.

For edge finding I use either a "click" type edge finder or an electric one with an LED that lights up when contact is made. Each of these has it's advantages and disadvantages. And it's quirks. Their accuracy depends as much on the skill of the user as on the device itself. Both require a light touch when approaching the edge. And both of them are relatively inexpensive.

https://www.mscdirect.com/browse/lookahead/?searchterm=edge+finder&hdrsrh=true

LED edge finder - Bing - Shopping

My examples are for illustration purposes only. I have not used any of the brands shown and can not give any experience with them. My actual center finders are older and probably no longer available.

I would suggest that the OP post a photo of the "wiggler" tool that he is using so we can be sure we are talking about the same thing.

5 axis Fidia guy said:

I think people are getting "wigglers" and "edge finders" confused.

Click to expand...

That’s a picture of a Starrett wiggler set. If yours is like that then two of the four attachments you have are for edge finding.
They normally come in a set with a pointer, an edge finder, that’s the one with the ball on the end or sometimes also with a tiny cylinder on the end and a DTI holder.

I don't think I’ve ever heard anyone refer to anything as a wiggler that was not one but if there is still any confusion it’s the one with the balls on it. Which are another nice feature as they allow the work to be slightly un-true as well as run out in the spindle without affecting the reading.

It is accurate, or I should probably say repeatable. It’s to within 1.5 tho. Repeatability is not the problem, the problem is it kicks out at different distances from the work based on the tightness of the collet holding the tool, so the problem is no apparent way to set a repeatable level of friction on the holding collet. Meaning registration from the edge is lost ever time the tool is changed, to within about 7 tho, though in truth you could probably tighten by feel to within about 3 tho if you got a feel for it.

Testing, testing, 1,2,3

I've been using wigglers and edge finders for 30 or more years with "good enough for todays job" results but this thread reminded me that I've never actually properly looked at accuracy and repeatability. Despite frequently saying that I'm convinced that the Huffam is the best and most consistent I haven't actually got figures to back it up. So I set them up in an Albretch chuck on the Bridgeport and spent an hour this morning taking readings off the ends of a 1-2-3 block from the 5 micron resolution DRO set-up with slightly surprising results.

Bottom line is I do get best results from the Huffam but the inexpensive collet set isn't as bad as I remember. Maybe a couple of decades or more practice with the Huffam has helped me get better results from the low end one.

I compared the average and scatter over 10 readings for my metric Huffam, a shop made metric Huffam copy, my old mid range collet style wiggler and a mid range edge finder at both 100 rpm and 1,000 rpm. I usually use the Huffam at 100 rpm because its easier to see the moment it starts to walk sideways. When using the Huffam at low speed to find the centre of a part errors cancel out very well indeed making DRO resolution the limiting factor. Assuming that both ends are properly smooth and clean of course.

Bit different if you are trying to measure a single side position of course.

The long side of that 1-2-3 block measures 76.10 mm. I set it up aligned on the X axis of the Bridgeport, used the Huffam to find the centre and set the DRO to zero. So the test edge was at 38.05 mm on the DRO.

The The Huffam ball end is dead on 5 mm Ø according to my M&W micrometer calibrated of a Weber gauge block. Average DRO reading was 40.525 ± 0.015 mm giving an edge position of 38.025 ± 0.015. Subjectively less scatter at 100 rpm.

The shop made copy ball end was also dead on 5 mm Ø. Average DRO reading was 40.500 ± 0.020 giving an edge position of 38.000 ± 0.020. Subjectively less scatter when run at 1,000 rpm. The shop made copy is a little stiffer in the pivot than the real Huffam.

The mid range brand wiggler has a 1/4" nominal diameter ball which measures 6.32 mm. Average DRO reading at 1,000 rpm was 41.245 ± 0.025 and 41.220 ± 0.060 at 100 rpm giving edge positions of 38.085 ± 0.025 and 38.060 ± 0.060 respectively. The collet pivot was much stiffer than the Huffam style which presumably is the reason for the greater intrinsic error. Quite hard to use when run slow but at 1,000 rpm the sideways kick is quite violent. Much more scatter than the Huffam style.

The normal edge finder end is 5.075 mm Ø. Average reading at kick was 40.545 ± 0.015 giving an edge position of 38.013 ± 0.015.

Both Huffam style and edge finder are clearly accurate enough to run into issues with DRO resolution, surface finish and general cleanliness. Best to do tell me three times process if you need accurate but one reading is pretty reliable.

The collet type clearly has greater intrinsic error but, with care its still pretty good. If you take 5 readings you can be pretty sure of results sufficiently reliable to subtract intrinsic error.

DRO digital numbers are slightly silly but it looks like the Huffam claims of finding an edge to 1 thou or so is realistic given reasonable care. Would need to have set-up one of my Heidenhain probes to get scientific level results but that's hardly shop realistic.

Clive

.

Picture order Huffam, Huffam copy, collet style and edge finder.

Clive, this is very interesting. Thanks for taking the time to test and report results.
I'm struggling to picture how the Huffam type works, any chance of some more pictures of it in a dismantled state?

I tend to use a Starret wiggler set for everyday use but break out a DTI or TessaTest for more critical work.

Many thanks

Alan