• Spring 2024 meetup in Calgary - tentative date Saturday, April 20/2024. Other regions are also discussing meet ups. If you want one in your area get going on organizing it! discussion
  • We are having email/registration problems again. Diagnosis is underway. New users sorry if you are having trouble getting registered. We are exploring different options to get registered. Contact the forum via another member or on facebook if you're stuck. Update -> we think it is fixed. Let us know if not.
  • Spring meet up in Ontario, April 6/2024. NEW LOCATION See Post #31 Discussion NEW LOCATION

New Digital Edge Finder

Tom Kitta

Ultra Member
I may try this - found exactly same one for $14 CAD -

There are others available - all well under $20 with fast shipping to Canada. Main difference is 10mm metric ball - which is 0.393701 in. So for extra 80 CAD you get imperial ball. This is so far the most extreme example of price gauging by Accusize. Usually they are less 50% more. Heck sometimes as close as 10% - I almost considered buying set of square 5C collets from them but decided to go with aliexpress and saved $20.

Thanks for the tool tip - I get mine in three weeks.
 

PeterT

Ultra Member
Premium Member
Tom beat me to it. I'm pretty sure I've seen that same edge finder that Accusize re-markets for a lot less cost, but that's another issue.

A disadvantage to most of those IMO is an overly large diameter shank & ball. Yes you can put the EF in a 3/4" collet, but rather limited in any other kinds of smaller tool holders. I use this Borite model. USA made, half the price, smaller diameter shank & edge surface, same quoted repeatability. I can insert in a much wider range of tooling, even a drill chuck. That speeds up setups. Also the measuring end is a smaller 0.200" diameter cylinder. That allows it to get inside smaller diameter holes or smaller vertical spaces for (center finding) half measurements. For conventional edge finding, works the same way - make contact enter nice round +/- 0.100" on DRO and you are done. All subsequent measurements are relative to that edge.

 
Last edited:

Tom Kitta

Ultra Member
Yeah, the Chinese version is a bit "massive" 20mm for metric and 3/4 for imperial. Both use LR1 batteries (well technically they are the same thing, just one imperial the other metric).

The Borite stuff looks great and a good price as well, made in USA also a bonus as its well made and imperial.

20mm is the max ER32 can handle - not exactly an issue as ERs are mostly metric so 20mm works far better then 3/4 as far as accuracy goes.
 

PeterT

Ultra Member
Premium Member
Little plate. 2”x2”x1/8” chunk of plastic, something you can use with epoxy. Glue it to a similar-size chunk of aluminum to make a sandwich. Aluminum needs a small tab to attach a wire. Clamp sandwich to mill, face to a known thickness. You now have an insulated gauge block, connect an LED and a battery so the LED lights when your tool bit touches the block. Zero the DRO, remove the block. You now have a known distance bit to table.
I've seen those & thought about making something like it, but I wasn't sure how good an electrical conductor typical carbide inserts are. Any experience on that front?

For my own (non-CNC manual milling) work, I've never seen much value in using an EF for vertical pre-calibration. Too many variables +/- a couple thou swapping tooling in & out, not to mention time. On a Bridgeport knee type mill with DRO its actually quite easy. Insert the cutting weapon of choice, make light contact with work surface, lock the quill in position, zero the vertical DRO, remember to back off the knee, done. If the work or cutter edge is sensitive, lay a shim of plastic or brass & enter the thickness, same thing. I also have a quill DRO so I often zero it too. It will show any quill drift during machining. All that said I have yet to work with a DRO where you could count on exact part measurement especially through an extended machining range. It always pays to re-measure & re-calibrate DRO as you approach final dimension, at least for +/- thou type accuracy. At least this has been my experience.
 

whydontu

I Tried, It Broke
Premium Member
I had to run down to my shop to find out, it never occurred to me since I rarely use carbide. So, trusty analog multimeter in hand, I checked. (Using my 1975 analog meter as a nod to Susquatch)

Yellow TiN coated, conductive
Shiny in-coated, conductive
Black PTFE-coated, as expected, non-conductive.
Grainy in-coated parting tool bits, non-conductive
 

Susquatch

Ultra Member
Administrator
Moderator
Premium Member
@Tom Kitta , @Tom O & @PeterT .

Time may prove me wrong on this one and I might have wasted some money. But it was a calculated risk I took with my eyes wide open. (well, maybe some rose coloured sunglasses on).

I saw the Ali express finders and a few on Amazon too. Accusize also sells two models. The one I got is the better one with 0.0002mm precision. I also bought a Fowler with 0.00001" precision which converts to 0.0003mm at $35. 1/3 the price of the Accusize. So I bought one of those too. (I did say I splurged on this - remember?)

BUT the accusize has a ball that can also be used for vertical edge detection. The fowler is only rated for side detection.

The fowler has a 0.2" cylindrical tip. The accusize has a 0.4" ball. I wanted one of each to play with.

I started out wanting to experiment and I was willing to pay a little for the lessons I might learn. One of those lessons might be "Don't spend money on meaningless precision".

That said, I have no regrets about the money I may have wasted on my Mitutoyo, Starrett, and Interapid precision tools. Not that accusize is in their league, but I'm absolutely certain that most of the other metrology instruments I have acquired from various sources is mostly pure junk.

I've been having good experiences lately with Accusize so I figured a hundred bucks for this kind of equipment was a reasonable risk to take given that they actually provide a number. Most of the stuff on Amazon and Ali would not even provide a number. I don't like that, but we will see how that all shakes out with time.

The specs for the $20 unit that Tom Kitta referenced say that it has 0.003mm precision. That's an order of magnitude poorer than the fowler.

Maybe that doesn't matter in the end, but for whatever reason I thought it was worth the extra coin to get the Accusize. Time may prove that wrong and perhaps even foolish.

My DRO is capable of sensing at the 1 micron level (0.00004"). So I used it to evaluate the two units. Sure enough, both were consistently accurate to +/- 1 micron over 10 samples minus a few I Fugd up. I had speculated that the difference in the two ratings might be purely a function of the point contact of a ball VS the line contact of a cylinder. That doesn't appear to be the case - at least not within the 1 micron resolution of my DRO.

Right about now you might be asking yourself why anyone would care. In fact, that's what I'd be asking if this was your thread. My answer is simple - I don't know what I don't know but in general I believe better is better. I can always use better where it isn't needed but I can't use less where more is needed. My curiosity also needs to be regularly fed.

Anyway, I remain very impressed with both units, but I favour the accusize unit because it will also do Z (if I can figure out how to do that) and because it also has a beeper and I REALLY like that!
 

Susquatch

Ultra Member
Administrator
Moderator
Premium Member
Little plate. 2”x2”x1/8” chunk of plastic, something you can use with epoxy. Glue it to a similar-size chunk of aluminum to make a sandwich. Aluminum needs a small tab to attach a wire. Clamp sandwich to mill, face to a known thickness. You now have an insulated gauge block, connect an LED and a battery so the LED lights when your tool bit touches the block. Zero the DRO, remove the block. You now have a known distance bit to table.

This is how I zero the cutter on my little CNC engraver.

I had similar thoughts. I even thought perhaps I could use the actual end mill that way.

Right now I'm leaning toward using the edge finder itself in a VBlock. What I'm not sure about is the conductivity of various types of end mills.

But I also have a bigger problem that needs to be solved first. How to get repeatable Z height. No point measuring anything I can't repeat. I'm not buying that TTS system. WAY TOO MUCH MONEY. But @Darren had a good idea about using R8 end mill holders. I can afford a few of those.

Anyway, your thoughts are good ones. I feel like there is a light at the end of the tunnel on this issue.
 

Susquatch

Ultra Member
Administrator
Moderator
Premium Member
I'm not sure how much effect the drawbar torque would influence Z height. You'd have to experiment. I have always set Z manually. I dont have a DRO on the Z.

Starting over, i think i'd be looking more closely at the TTS or other type of quick change setup. I really like putting everything in the ER40 collet chuck for speed, but with the power drawbar i can swap fairly quickly now.

I'll do it as soon as I get a 3/4" end mill holder. (I don't have anything else that big - my chuck is only 5/8. Pretty sure an R8 collet will move so no sense trying that. With a holder, I think I can just zero Z, tighten the drawbar and then check how much Z moved - if any.
 

Susquatch

Ultra Member
Administrator
Moderator
Premium Member
You'd leave the endmill in the holder and number the holders. When you remove the endmill from the holder itself, you'd have to reset the Z offset in the DRO. So to swap from your 1/4" to 1/2" endmill, yo would be swapping out the holder by undoing the drawbar.

So, it turns out that the tool library function only applies to the Lathe configuration. It doesn't exist for the mill configuretion. Too bad, but C'est La Vie.

However, it does have 200 sets of auxiliary zeros. I may be able to use a few of those. It just won't be as easy because they only work in absolute mode.

More fun playing in the new sandbox.
 

Darren

Ultra Member
Premium Member
So, it turns out that the tool library function only applies to the Lathe configuration. It doesn't exist for the mill configuretion. Too bad, but C'est La Vie.

However, it does have 200 sets of auxiliary zeros. I may be able to use a few of those. It just won't be as easy because they only work in absolute mode.

More fun playing in the new sandbox.


double check again. I can't see them not having that in mill mode. If its true, that cancels my plan for getting a 4 axis for my mill to replace the 2 axis Mitutoyo. Damn
 

Susquatch

Ultra Member
Administrator
Moderator
Premium Member
I'm not sure how much effect the drawbar torque would influence Z height. You'd have to experiment. I have always set Z manually. I dont have a DRO on the Z.

Starting over, i think i'd be looking more closely at the TTS or other type of quick change setup. I really like putting everything in the ER40 collet chuck for speed, but with the power drawbar i can swap fairly quickly now.

I do have a 3/4 Weldon Arbour. That should be every bit as good as an end mill holder so I tried it. On the basis of just one test, the difference between snug and tight is 2.8 thou and the difference between snug and torqued down really tight is 3.1 thou (just another 3 tenths) which makes sense to me. I did ten samples to see how consistent the R8 taper and I both are. Tight (my normal torque application) varied between 2.7 and 2.9 thou. So +/- a tenth. Not bad for most things.

Now I just need to figure out how to do the Z Axis in a way that translates to tooling, get some R8 end mill holders, and then figure out how to use the alternate zero function on my DRO.
 

Darren

Ultra Member
Premium Member
If you able to get under 1 thou accuracy on a Bridgeport type mill, you're doing pretty darn good already.
 

Susquatch

Ultra Member
Administrator
Moderator
Premium Member
double check again. I can't see them not having that in mill mode. If its true, that cancels my plan for getting a 4 axis for my mill to replace the 2 axis Mitutoyo. Damn
Will do. I'll find it in the lathe function and then switch to the mill mode. Stay tuned.
 

Susquatch

Ultra Member
Administrator
Moderator
Premium Member
double check again. I can't see them not having that in mill mode. If its true, that cancels my plan for getting a 4 axis for my mill to replace the 2 axis Mitutoyo. Damn

Sorry Darren, it's only there in the Lathe mode. Not only that, but it's only for x & y.

But I am thinking I can use 10 or 20 of the alternate zero settings to do the same thing. It's a bit more complicated but there are 200 of them and they apply to all 4 axis.

I'm heading off to the house right now. CEO's orders. But I'll read up on it there and test it tomorrow for you.
 

PeterT

Ultra Member
Premium Member
Surface finish can have large affect on EF accuracy & repeatability. For fun, grip your best 123 block in the vise, get an EF null reading, zero the DRO. Wind out the lead screws, go to a new spot on block, rinse & repeat. No peaking at the DRO, only rely on the EF light/beep. How many of the DRO 0.00000" digits to you get to keep? Now consider if the typical part you are machining will have the same Ra & squareness as your 123 block.

Even if the EF is 0.000000000 to the edge, that just conveys the center of the cutting tool. But the EM center is not the EM edge, its cutting OD also has +/- tolerance too. Usually we are milling a surface relative to a datum so no getting around breaking out the mic to verify & adjusting accordingly. Drill chucks & drills & collective gripping of the two are typically crude resolution by comparison. I know you know all this stuff, just reinforcing the concept. My own rule of thumb for manual mill/latheis one-thou is pretty good, actually hard to achieve on more complex parts. If finer tolerance is required, at least repeatably, its probably grinding & lapping territory.
 

Susquatch

Ultra Member
Administrator
Moderator
Premium Member
Surface finish can have large affect on EF accuracy & repeatability. For fun, grip your best 123 block in the vise, get an EF null reading, zero the DRO. Wind out the lead screws, go to a new spot on block, rinse & repeat. No peaking at the DRO, only rely on the EF light/beep. How many of the DRO 0.00000" digits to you get to keep? Now consider if the typical part you are machining will have the same Ra & squareness as your 123 block.

Even if the EF is 0.000000000 to the edge, that just conveys the center of the cutting tool. But the EM center is not the EM edge, its cutting OD also has +/- tolerance too. Usually we are milling a surface relative to a datum so no getting around breaking out the mic to verify & adjusting accordingly. Drill chucks & drills & collective gripping of the two are typically crude resolution by comparison. I know you know all this stuff, just reinforcing the concept. My own rule of thumb for manual mill/latheis one-thou is pretty good, actually hard to achieve on more complex parts. If finer tolerance is required, at least repeatably, its probably grinding & lapping territory.

Ya, I have no delusions about my ability to actually make things anywhere near that accurate.

But I am a believer that my measuring equipment should be at least an order of magnitude better than my machining equipment and my skills.

FWIW, I already did almost what you suggested. Even to the point of not peeking at the DRO. I was actually amazed to discover that all digits were exactly the same - even the very last one. Made me wonder what wasn't working or what did I do wrong.

My own musing about it goes like this. My DRO is way better than I ever expected it to be. Both my electronic edge finders are as good as their manufacturer said they were. And both my DRO and the edge finders are way better than me or my mill.

Which is exactly the way I would like things to be.

But I confess that I am a little out of sorts over the DRO. Maybe unsettled would be a better way to express it. I like to understand things. I don't understand the DRO. How does it achieve that level of precision? How can it be so repeatable? How come my Gibbs, the weight of my table, the ambient temperature change, and a dozen other factors don't mess with it? I kinda feel like I'm dreaming and I'm gunna wake up tomorrow and realize that there is a little man inside there who is just making up numbers to mess with me.
 

Susquatch

Ultra Member
Administrator
Moderator
Premium Member
Ok Darren, no tool library in Mill Mode. BUT.....

It does look like you can designate any number of the 200 available SDM storage locations to serve the same function (actually a bit better). There are three benefits of using the SDM locations.

First, they are relative to the ABS position. If you change the ABS location, the SDM locations change accordingly. Ie set abs 2 inches right, and all the sdms change 2 inches right.

Second, setting an SDM location saves all 4 axis, not just the x & y. The tool store only saves x and y.

Third, they can be saved by moving the axis, or they can be entered directly on the keyboard.

Using them is pretty easy. Just hit SDM, then the number you want (anything between zero and 200). The selected SDM number displays right in the SDM Screen box. Pressing either ABS or INC exits SDM mode.

I believe they are intended to keep track of features on the part relative to your part zero. But it seems that the part features are not really parts or tools - just a 3D location in space relative to ABS. So the DRO doesn't know or care if it's a part or a tool.

Go ahead and ask questions. Answering them will help me understand them too.
 

Tom Kitta

Ultra Member
@Tom Kitta , @Tom O & @PeterT .

Time may prove me wrong on this one and I might have wasted some money. But it was a calculated risk I took with my eyes wide open. (well, maybe some rose coloured sunglasses on).

I saw the Ali express finders and a few on Amazon too. Accusize also sells two models. The one I got is the better one with 0.0002mm precision. I also bought a Fowler with 0.00001" precision which converts to 0.0003mm at $35. 1/3 the price of the Accusize. So I bought one of those too. (I did say I splurged on this - remember?)

BUT the accusize has a ball that can also be used for vertical edge detection. The fowler is only rated for side detection.

The fowler has a 0.2" cylindrical tip. The accusize has a 0.4" ball. I wanted one of each to play with.

I started out wanting to experiment and I was willing to pay a little for the lessons I might learn. One of those lessons might be "Don't spend money on meaningless precision".

That said, I have no regrets about the money I may have wasted on my Mitutoyo, Starrett, and Interapid precision tools. Not that accusize is in their league, but I'm absolutely certain that most of the other metrology instruments I have acquired from various sources is mostly pure junk.

I've been having good experiences lately with Accusize so I figured a hundred bucks for this kind of equipment was a reasonable risk to take given that they actually provide a number. Most of the stuff on Amazon and Ali would not even provide a number. I don't like that, but we will see how that all shakes out with time.

The specs for the $20 unit that Tom Kitta referenced say that it has 0.003mm precision. That's an order of magnitude poorer than the fowler.

Maybe that doesn't matter in the end, but for whatever reason I thought it was worth the extra coin to get the Accusize. Time may prove that wrong and perhaps even foolish.

My DRO is capable of sensing at the 1 micron level (0.00004"). So I used it to evaluate the two units. Sure enough, both were consistently accurate to +/- 1 micron over 10 samples minus a few I Fugd up. I had speculated that the difference in the two ratings might be purely a function of the point contact of a ball VS the line contact of a cylinder. That doesn't appear to be the case - at least not within the 1 micron resolution of my DRO.

Right about now you might be asking yourself why anyone would care. In fact, that's what I'd be asking if this was your thread. My answer is simple - I don't know what I don't know but in general I believe better is better. I can always use better where it isn't needed but I can't use less where more is needed. My curiosity also needs to be regularly fed.

Anyway, I remain very impressed with both units, but I favour the accusize unit because it will also do Z (if I can figure out how to do that) and because it also has a beeper and I REALLY like that!

There is the stated accuracy and then there is "real life" like I do not care what Accusize states - their stuff is exactly the same as stuff from China I got just in imperial. So if Chinese state 0.003 then it is probably closer to +- 0.005 in both cases. I could be wrong, and it could be +-0.002 but 0.005 is quite good already. Remember that is +- two thou. Accusize just re-badges stuff. I am sure if I looked a lot I could find imperial one or I could ask seller for imperial one - I bet $5 more.

The DRO scale step is a micron. That does not mean the measurement will be that precise. I have all 5 micron scales and they are better then a BP needs. Again, this does not mean anything more then a "step is 5 micron". Its like your calipers have 0.01mm step - does not mean you can measure reliably with your calipers down to 0.01mm.

I am unsure of the Fowler unit - but a lot of their stuff is "better China" re-badged - at least for a while now.

Also I am unsure of the Z stuff - I guess when I get my Chinese indicator I play with the Z as well. For actual precise Z I have a little cylinder with 0.001mm step scale that measures the tool height. To be honest I never yet needed this - I may use it for the CNC stuff.

For tool library stuff to work with height precisely (or good enough) you would need 40 taper tooling (or 30) not R8 and I don't think MT3. Also multiple holders with multiple pre-set tools. This could be OK in some applications but seems a bit out of the way of most hobby stuff - there was some discussion about this on this board a while ago by Dabbler.
 

Susquatch

Ultra Member
Administrator
Moderator
Premium Member
There is the stated accuracy and then there is "real life" like I do not care what Accusize states - their stuff is exactly the same as stuff from China I got just in imperial. So if Chinese state 0.003 then it is probably closer to +- 0.005 in both cases. I could be wrong, and it could be +-0.002 but 0.005 is quite good already. Remember that is +- two thou. Accusize just re-badges stuff. I am sure if I looked a lot I could find imperial one or I could ask seller for imperial one - I bet $5 more.

Perhaps you missed the fact that I actually measured it. It was at least as good as they claimed, perhaps better.

It meets my personal criteria which is to ensure that my measurement equipment is always an order of magnitude better than my equipment and me.

I think accusize has changed. It seems like they have started sourcing some stuff locally. Even if not, they have a clear leg up on you and me. They speak Chinese, they buy in volume, and they definitely care about customer satisfaction. They have gone way over board for me several times now. Even when it wasn't their fault. The improvement in their quality suggests they have swung deals that I could not swing to buy better stuff than I can buy on my own. I have no proof of that, but I can say that everything I bought from them in the last 2 years has been first rate. I can't even say the same for KBC!

I am happy to give them credit for what they have accomplished, and they are earning my respect for it.

Bottom line is that I'm happy with it, it met my precision requirements, it delighted me from the first moment I used it, it arrived in just a few days, and the beeper function is awesome! To me it was worth every penny I paid for it! I have no regrets!
 
Again I see accuracy thrown around when reading specs and referring to them in performance our equipment.

Regardless of the perceived accuracy of your scales and your DRO's display, what your machine is actual capable of is the determining factor and I'll bet that if you can get down to 1/10 of thou. you are doing really good. Anything beyond that is truly wishful thinking despite what your DRO says.

That said, a simple probe that is capable of a 0.0001 is more than accurate enough for 99.99% of us. For the few that have machines capable to go beyond that that gets into a different discussion.
 
Top