Cheap and Dirty Zeroing on Custom Fixture w/ End Mill

[GJeff]

I'm in the process of making a custom vacuum table that will have accurate stops for positioning stock and custom spoilboards. The spoilboards will be 1/4" polycarbonate because that's what I have lying around and it seems like they'll work.

What I want is a way to quickly get the machine zeroed to the vacuum table at startup. My homing switches are not it, obviously, and I don't own any fancy pants probes. I'd like to build something into the vacuum table itself to make zeroing it easy. For instance, an empty cylinder where I can lower an end mill, then use calipers, measuring the distance on X and Y between the shaft of my end mill and the cylinder walls to ensure that it is centered in the hole (assuming vacuum table has zero rotation).

Any other ideas? Ideally I want to be able to do this without tool changes. I don't want to use wobblers or anything like that. I like the idea of calipers.

Similar Threads:

• Cheap,flat,sacrificial work holding fixture
• Just In Bobcad-Cam V25 Mill/Lathe/Plasma + tons of EXTRAS. CHEAP CHEAP CHEAP!
• Zeroing End Mill Question
• Cheap & dirty mister setup
• Quick and Dirty X3 Coolant enclosure..cheap too!

[Jim Dawson]

For non-precision work (+/- 0.005 or so) I just put a couple of shallow divots in the fixture with a small center drill, one on each end. Then just use a center point in a collet to locate the divots. I think the calipers would take longer to set up. If you indicate the fixture, then you just have to locate one divot.

[GJeff]

I'll be looking for +/- .001 precision or better. I'm thinking making a 3 axis touch off plate may be the way to go. I'll make it to work for stock corners or to fit in the grooves of my vacuum table so it'll be multipurpose. I can put an insulating layer in it, or just make the base out of delrin. Use an aluminum cutout on top of that that's connected to one wire of an ohmmeter, attach the other to the end mill, sink it deep enough so that the shaft is what makes contact, manually creep to the edges until I hear the beep and take the average. Then do the obvious for the Z. Think that'll get me the precision I want?

[Jim Dawson]

That will work fine for the Z, but the question is exactly where on the edge of the endmill are you contacting in the X and Y axis? Just measuring an end mill with a mic is challenging enough to get an accurate measurement. It has to be turned just right to get an accurate measurement across the maximum diameter. Not saying it won't work, maybe with some practice you can get the accuracy you want.

When I want real accuracy I use a spindle microscope.

[GJeff]

I don't have a spindle microscope yet, and that would require a tool change

I typically use smaller end mills, so it should not be hard to make my gauge tall enough so that I can contact off the shaft of the end mills above the flutes and neck. For longer end mills I would manually rotate them back and forth to find the x and y positions where the edge of the flute just makes contact. I think I get similar or better accuracy using that method just be feel (no ohmmeter) than I do with a concentric mechanical edge finder.

Or maybe I should just stop being a cheap ass and buy a TTS probe. I've heard such bad things about the moderately cheap ones, though. And the good ones like a renishaw are $$$$$. Do you use one or do you always prefer your microscope?

[Jim Dawson]

I have a cheap probe (around $100) out in the shop but I have never used it so can't offer any advice there.

Normally I use an edge finder, that will get you to 0.001 or better. I don't normally need to locate anything closer than that. That's pushing the limits of the mill anyway, at least for most work. There are ways to get closer if needed, but it takes time and careful planning. I do have 1um scales on the machine. If I'm doing a full profile on a part I leave about 1/16 or more for trim on the raw material, this eliminates the need for super accurate setup.

I really don't use the microscope that much, most of the time it's not needed.

This is how I set up the microscope, it's integrated with my software. I could set it up to do an auto edge find, but that would require actual work



The mint mark on a penny


Screen shot

[GJeff]

Using image processing to do automatic edge detection with a microscope? I like it. I wonder if it would take you more time to program than the entirety of the time you'd save. Most likely, but that doesn't necessarily mean it wouldn't be worthwhile for pride/fun. The engineering side of the brain wants to maximize efficiency as if we'd be doing the process for eternity.

I always try to avoid needing an accurate setup and let the machine produce the accuracy, like you said by doing the profile or something similar. My current project requires that I make a custom spoilboard, place my workpiece on it (1/2" hardwood sheet), mill the surface, flip it, then mill that surface within +/- .001 XYZ relative to the other surface (for my OCD, not out of necessity). The method I used before was making a pair of .25 holes through the workpiece into spoilboard, then flipping the workpiece and using pins to realign those holes. It wasn't the most accurate in acrylic, and it'll be worse in wood. Instead I want to machine some shape proud of the top side of the workpiece so I can flip it and have it self align into premade holes in the spoilboard. The problem with that method is the machine has to be already be indicated to the spoilboard to start - which is challenging considering my lack of tools and that the spoilboard is a soft material, polycarbonate. So I wanted an easy way to indicate the machine to the vacuum table, which would then be indicated to the spoilboard by virtue of a similar aligning method that was machined into the bottom of the spoilboard which fits in the grooves of my vacuum table.

It would all be trivial if I could turn on the machine and do it all in one go without ever losing position, but that is not the case.

[GJeff]

Spoilboard may not be the proper terminology. A custom vacuum fixture might be a better way to say it. I need to be able to re-use the same spoilboard/vacuum fixture many times, over many days, so being able to indicate it is necessary.

[Jim Dawson]

It really wouldn't take that long to program an edge finder, I already have all of the algorithms written for another product I sell. I would just have to drop the modules into the CNC software and tie it all together. About a day of work, but I haven't had a need to do it.

I normally use 3 dowel pins to locate the work edges on the fixture if I'm doing multiple parts. 2 dowel pins on one edge and 1 dowel pin on another, this locates the work square in two axis. Just locate the dowel pins one time, then all you have to do is slide the workpiece up to them. This works if you are not profiling the entire part. Another way to do this is with shoulder screws, same as dowel pins, but allows you to remove them for full edge profiling.

Another thing you could do is put a ''key'' on the back of your vacuum table that fits the T-slot on your mill. It doesn't even have to be aligned that well with your vacuum table to start with. Machine the vacuum table to the key once it's installed on the machine. You could even take a very light cut on the T-slot edge to make sure it's in line with your X axis, but not really needed.

[GJeff]

Knowing me I'd make a key and then never trust it, so I'd be indicating the edge of the vacuum table every time it came out. I should be able to do 95% of my work with the vacuum table in place. Only if I need the 4th axis should it come off is my hope.