[Russ Kaiser]

I just finished a router mount for my cnc mill to use as a high speed auxiliary spindle. The router has a 1/4 inch collet and so none of my edge finders will fit.

I did some searching and found that Fisher machine tools makes an edge finder with a 1/4 inch shaft but like all of these devices, it's limited to about 1000 rpm. My router won't dial down that far. Right now, I am planning on using a piece of 1/4 drill rod and feeler gauges to find the edge but this is a slow process.

On my machine, because it is open loop design and I don't have encoders, I have to drive the machine to position using the steppers for the computer to know where the spindle is. You wouldn't need to do this if you were just referencing a single edge, you could get the spindle in position by hand with the controller off and then turn everything back on and set the axis zero, but if you need two references on the same axis, like finding the center of the stock, you have to leave the controller on.

I have a handheld controller and can turn the feed rates way down, but trust me when I say that it's going to be a pain in the butt to work without an edge finder.

Does anyone have a solution?

Here is a link to my router mount thread.

My router mount thread on CNC Zone

[txcncman]

Assume you are cutting parts from oversize raw stock and not finished to size: Use a straight router bit and turn the spindle on and dial the bit to just touch the material then get location numbers from control. Also, don't touch both sides of stock to find center, just measure and do the math. Physically dial machine over with calculated number to center to visually verify center.

[Russ Kaiser]

I am often milling a cavity in the center of a round stock that can't be altered, so touching just won't do.

I wish it was that simple

[beege]

So you mount a pin of a known diameter into the spindle. Make sure it as little runout as possible, or mark where the runout average is. If it's round stock, use a shim of known thickness and "feel" the edge, from both sides along X. Make sure if you marked average runout that the marked area is oriented towards the part. Do the math for center on X, the position for Y zero, do the same for Y, then go back and redo the X if you feel you must. Should come pretty close...

[Russ Kaiser]

But goodness is it tedious to get things dead on.

There's got to be an easier way... I hope.

[txcncman]

Really there is not a "better" way. We just learn to do what we have to do and learn to do it most efficiently. If you want to use an actual edge finder, pony up the cash for a router than can run 1000 RPM on the spindle. I can't imagine you are making any artsy type parts that require more than +/- 1/64" precision. 1/64" is eyeball work. I bet I could set position on your router within 1/64" in less than 2 minutes per axis. How tedious is that?

[christinandavid]

Put part of known diameter against two pins of known position and you can work out the centre position... If you are milling out the centre drill a pilot hole first for quick verification purposes.

DP

[Geof]

But goodness is it tedious to get things dead on.

There's got to be an easier way... I hope.

Have you ever used an edge finder? My experience is that using one is no quicker than using the technique described by beege.

[Russ Kaiser]

Really there is not a "better" way. We just learn to do what we have to do and learn to do it most efficiently. If you want to use an actual edge finder, pony up the cash for a router than can run 1000 RPM on the spindle. I can't imagine you are making any artsy type parts that require more than +/- 1/64" precision. 1/64" is eyeball work. I bet I could set position on your router within 1/64" in less than 2 minutes per axis. How tedious is that?

Hey Tex,

Call me sensitive, but I thought this site was for people to help each other out not to make them sound lazy and or stupid. Of course I know how to do this "old school". I was wondering if there was a "new school" method that maybe I didn't' know about. I am perfectly willing to spend money, but I have limits. I actually have managed to make every meager tool in my shop pay for itself. And my main spindle does turn down to well below 1000 RPMs.

As for being able to do this in under two minutes per axis, good luck. Like I said, you have to drive the guide dowel in the router collet up to the work with a tethered remote making sure you go slowly enough that you don't break something while holding a feeler gauge in your other hand. Round stock, round dowel, flat feeler gauge; pain in the butt.

Plus the stock is semi flexible so you have to consider how hard you are resting against it. It will deflect 0.015 easily without changing the feedback you get from the feeler, so you have to catch the initial contact. Hey wait! 0.015 is about the same as that 64th you claim you can eyeball.

Also, the last time I checked 1/64 was about 10 times more error than I ever would want on my work and frankly this stuff is very hard to eyeball because of the shape and the color. I do have a center hole for reference though.

Here is my setup for this operation, you can see that's not your typical "Artsy Type Parts".




And here is some finished work. A 1/64 error makes a big difference in a musical instrument mouthpiece.

[Geof]

.....Like I said, you have to drive the guide dowel in the router collet up to the work with a tethered remote making sure you go slowly enough that you don't break something while holding a feeler gauge in your other hand. Round stock, round dowel, flat feeler gauge; pain in the butt.....

At the risk of offending you I will say I think you are "holding your mouth wrong". I do this type of edge finding all the time using a piece of paper not a feeler gauge. This gives you a bit of compliance because the paper is 0.003" to 0.005" thick so if you overshoot a couple of thou there is no problem. For non-critical work (+/-0.001) I just use the paper and enter a correction for the paper thickness. The correction thickness I get from pinching the paper in a micrometer to yield about the same drag I feel between the work piece and the probe. For critical work after touching the paper I then switch to a 0.020" feeler gauge and gradually open the gap between probe and work until the feeler will just enter. I don't consider this a pain in the butt and I take a lot less than two minutes per axis to enter the offsets.

[Russ Kaiser]

At the risk of offending you I will say I think you are "holding your mouth wrong". I do this type of edge finding all the time using a piece of paper not a feeler gauge. This gives you a bit of compliance because the paper is 0.003" to 0.005" thick so if you overshoot a couple of thou there is no problem. For non-critical work (+/-0.001) I just use the paper and enter a correction for the paper thickness. The correction thickness I get from pinching the paper in a micrometer to yield about the same drag I feel between the work piece and the probe. For critical work after touching the paper I then switch to a 0.020" feeler gauge and gradually open the gap between probe and work until the feeler will just enter. I don't consider this a pain in the butt and I take a lot less than two minutes per axis to enter the offsets.

Don't worry, I am not offended, but please look at this from my point of view. First, please read my opening question. It is obvious that I know how to do this the way that everyone is describing.

Second, until you try to do this with my machine, on my part, with my stock, I don't think you can honestly say how quickly you can accomplish this task. I use my machine all the time and frankly the remote and computer interface leave a lot to be desired. Please take my word for it, that an edge finder is vastly superior (in my particular case) to a method using any type of gauge material, paper or metal. I have done the feeler method many times because I operated without edge finders for quite a while when I first got started with CNC machining. I know exactly how much simpler edge finders made my life when I bought them... freaking bunches easier!

If paper or feeler gauges were just as simple and accurate, you couldn't sell an edge finder could you? Why spend any amount of money for something more mechanically complicated if something simple does just as a good of a job.

I love this forum, but so often people give opinions without trying to answer the question asked. I simply wanted to know if anyone knew of an edge finding method that would work in high speed application that would fit a 1/4 inch collet. When I asked this I thought it was quiet obvious that I knew the method everyone has so painstakingly described.

Thanks,

Russ

[txcncman]

Hey Tex,

Call me sensitive, but I thought this site was for people to help each other out not to make them sound lazy and or stupid. Of course I know how to do this "old school". I was wondering if there was a "new school" method that maybe I didn't' know about. I am perfectly willing to spend money, but I have limits. I actually have managed to make every meager tool in my shop pay for itself. And my main spindle does turn down to well below 1000 RPMs.

As for being able to do this in under two minutes per axis, good luck. Like I said, you have to drive the guide dowel in the router collet up to the work with a tethered remote making sure you go slowly enough that you don't break something while holding a feeler gauge in your other hand. Round stock, round dowel, flat feeler gauge; pain in the butt.

Plus the stock is semi flexible so you have to consider how hard you are resting against it. It will deflect 0.015 easily without changing the feedback you get from the feeler, so you have to catch the initial contact. Hey wait! 0.015 is about the same as that 64th you claim you can eyeball.

Also, the last time I checked 1/64 was about 10 times more error than I ever would want on my work and frankly this stuff is very hard to eyeball because of the shape and the color. I do have a center hole for reference though.

Here is my setup for this operation, you can see that's not your typical "Artsy Type Parts".




And here is some finished work. A 1/64 error makes a big difference in a musical instrument mouthpiece.

You made yourself sound stupid and lazy. I just reinforced the points.

[Russ Kaiser]

You made yourself sound stupid and lazy. I just reinforced the points.

Who exactly is proving who's point here, Mr. Kirkpatrick?



Onward...


Someone was kind enough to send me a note on camera based edge finders and I am also looking into the laser models that are available.

This company makes a model with a 1/4 inch shaft.

Laser Center/Edge Finders

Pricey, but maybe I can adapter a handgun laser sight if the mounting is concentric and it has a fine enough beam. Most of them are tunable for size and have deflection screws to adjust where they point.

Up until yesterday I didn't know laser edge finders existed and this is the kind of input I was hoping to get.

Thanks,

Russ

[TechGuy5002]

If you need to find the Center of the workpiece. A Coaxial gauge would work great. However you would need to rotate the gauge manually since its limited to about 1K RPM. using a coaxial gauge manually isn't a big deal as long as you can hand turn the spindle. A coaxial gauge has a feeler that rotates either on the outer diameter or the inner diameter, and has a dial gauge to display the offset. It takes about 30 seconds to locate the center with in a couple of mills.

[Russ Kaiser]

If you need to find the Center of the workpiece. A Coaxial gauge would work great. However you would need to rotate the gauge manually since its limited to about 1K RPM. using a coaxial gauge manually isn't a big deal as long as you can hand turn the spindle. A coaxial gauge has a feeler that rotates either on the outer diameter or the inner diameter, and has a dial gauge to display the offset. It takes about 30 seconds to locate the center with in a couple of mills.

Thanks for the input TechGuy - I have actually used one of these before when I was a kid taking shop. I never bought one for myself because there was always some other tool I needed a bit worse.

The height may be a problem though. With this router holder hanging below my quill I barely have z clearance between it and my work for a long mill. I was hoping to find something with a total height of 3 inches or less.

You're right however that dialing it in should be quick since you can watch the deflection and at the 4 points of the compass after a single rotation and figure out exactly where to jog the axis.

I will have to see if someone that caters to the Taig crowd makes a tiny model and go from there.

Thanks,

Russ