New Zero Camera

Have you tried taking it apart to see if you can secure the parts that are causing the relative movement?

@veng1 How is your Zeroing Camera setup working out? Any issues? I just got my Supereyes and am about to cut the brackets and get it setup. Any tips?

I have been swamped with other issues and haven’t touched it for a while.

Be sure to use two clamps if your Supereyes is like mine where it tapers rather than a plain cylinder. A single clamp did not hold it reliably as it needed to be fixed at two locations.

I have one of the older, lower resolution versions and need to order one of the newer higher resolution ones to try.

Keep us updated.

I bought one of these cameras and tested its resolution by looking at the gap in my calipers and moving them. At the closest focal distance the field of view (full height of the monitor) is about 0.040". You can easily resolve better than 0.001". When my Nomad arrives I’ll figure a way to get the camera to stay steady one way or another. The idea of using optics for locating seems like it would be invaluable for a tool like the Nomad since you can’t easily use a edge finder or dial indicator.

This link shows a 0.005" gap and a 0.010" gap in my calipers viewed from the USB microscope… https://www.dropbox.com/s/wgmwgrzqjm95xew/2MP%20USB%20Microscope%20pictures%20showing%20sub%200.001%20resolution.pdf?dl=0

I’ve been following this discussion for some time and thinking on it… I’m not convinced that magnification is necessary, high or otherwise… but it would depend on some experiments.

Rather than trying to do as has been espoused here, I would combine the viewing with a laser. Defraction patterns from an edge are very easy to see. With an encouraging geometry, such a system would be much more efficient than simply magnification.

I’ve used a system that only had a laser. One would creep spindle until the edge simply split the laser, part on the stock top and part making a line down the edge. This was, with a bit of practice, better than 0.001" accuracy/precision was attainable (beyond what the Nomad is capable of).

Cheap lasers with a slit that is very small are easily and trivially available… and could be better that he above.

mark

Post a picture of your results and the cost and setup of the system. If its cheaper and/or works better I’d consider it. Any links would be appreciated.

Its hard to imagine anything being easier to implement than the system described in the OP. I’ll find out but I don’t think the camera drift issue will be difficult to solve.

My Nomad arrival is REAL SOON NOW.

My first instinct is to do what I’ve seen and used elsewhere, a cheap laser and some eyeballs - no magnification. That is bound to the be the same cost or cheaper than a microscope and would remove the need of a feed going back to a computer.

Dealing with an edge is well known machining problem. Having learned the old way, I can find an edge in seconds, even without an edge finder. Even so, an optical solution is a very nice, useful “luxury”.

YMMV.

Yes, the problem is largely the mounting of the optical element. The systems I’ve used had a simple machined 6061 mounting and three alignment screws with lock washers.

mark

Very interesting @mbellon!!. I like the idea of a laser. I implemented the whole camera thing including interface with my computer and found that the camera inside the housing was not locked down, so any movement of the cable would sent it out of alignment. After talking to the manufacturers I was discouraged to continue trying with the camera I had. Other projects took precedent and I put the camera aside. At this moment in my learning process, precise zeroing is not that important. It will be when I get to repeatability and when flipping which I will get to soon. The laser idea seems simple and elegant. I will start looking for them. The size of the beam would be something to look out for. I don’t know if this can be chosen. Love this community and all the different ideas that float around. It is refreshing and fun.

I would just machine a bracket out of PVC or Delrin using the Nomad itself. 3D printed parts are good for many things but they generally don’t work well when high precision, high strength or rigidity are needed.

You can find an edge using a piece of paper and an edge finder but I don’t think you can fit an edge finder in a Nomad, can you?

I’ve never been able to get closer than 0.04" consistently by eyeballing, even close up and my eyes aren’t getting any better with time :wink:

There’s just something I love about optical positioning. I’ve always hated even the small amount of time it takes to locate features using an edge finder or dial indicator. I can picture using some kind of three legged pins with cross hairs to locate hole centers. Edges would be pretty easy optically without any jig. I guess there would still be issues for parts with height features that would not allow the camera to be positioned close to the work surface.

I would just machine a bracket out of PVC or Delrin using the Nomad itself. 3D printed parts are good for many things but they generally don’t work well when high precision, high strength or rigidity are needed.

Here! Here. Yes, today’s 3D printers (that are reachable to most of us) do not make parts that hold up to wear and vibration well (and we’ve got plenty of both in the Nomad (and all CNC machines)).

+3 PVC
+10 Delrin

When work holding from the sides, the fixturing has to be cleared. Often that is 1/2" or more… the gripping part. Since the fixtures have to be cleared, a simple, small laser spot is more than good enough. An eyeball - with a bit of practice - will be more than good enough. As long as the work is known to fit on the stock and one can put the spot such that is clears the fixtures you’re all set.

Precision edge find is a different story. That is where an optical system makes sense.

You can find an edge using a piece of paper and an edge finder but I don’t think you can fit an edge finder in a Nomad, can you?

They sell an edge finder for the Nomad. See the product offerings.

One doesn’t need an edge finder, just a (preferably) new end mill. Paper is all to variable (so one has to measure what you’re using). In the old days, we used the cellophane from cigarette packs because they were universally 1 mill and very consistent (I never smoked, but many around me did). I tend to use a piece of commercial flexible plastic… known, trusted thickness.

I hit 0.001" trivially all of the time (and quickly) and my eye’s aren’t very good at all. I use touch like the machinists of old. I can hit 0.0001" with a new end mill via touch too (on a machine designed to do it)… it’s a pain though.

We could put together a tutorial about how to get to 0.001" with ease via touch. It’s not hard, takes some practice, and can be done within seconds once the technique is mastered.

A small laser with the right window should be able to hit 0.001" - all that the Nomad can do. Diffracting off the edge and watching for the pattern, is - theoretically - way better than 0.001" - thousand of Angstroms.

Trying to keep the microscope close to the face is tough. It’s a cantilever and likely to wobble; dust may get into the system; parts often require multiple passes so they aren’t always simple rectilinear shapes. Adding retracting makes it too complex.

mark

P.S.

We’re obviously getting an infusion of ideas here. Brainstorming, none of the ideas presented are right or wrong. We try them out and iterate to an acceptable solution.

I checked the microscope resolution with it offset 2" and I think you could resolve 0.005" even at that distance. It appears to have about a 1.25" field of view when 2" away so you can see a relatively large area even without the microscope anywhere near the workpiece. I think that makes it even more useful than I thought. Its very easy to slide it closer or further from the workpiece and readjust the focus in seconds so you could set it close for the initial setup then move it back a bit to see a wider angle for other operations.

I’m looking into the end of the USB microscope now. It doesn’t appear to be sealed. However I think it would be easy to seal it. The hard part would probably making sure everything INSIDE the microscope assy is secure enough to not vibrate out of position during operation. Its not apparent how the assembly is held together so it may take sacrificing one to figure out how to get it apart without damage so it can be repackaged in a way that doesn’t allow it to move.

When I mentioned machining a bracket out of PVC… I meant a PVC block, not PVC pipe. You can get PVC stock blocks are a lot cheaper than Delrin from McMaster or local plastic distributors and it has a higher coefficient of friction so it would grip the scope better with less clamping pressure. PVC is easy to machine and works great for many applications.

I checked the microscope resolution with it offset 2" and I think you could resolve 0.005" even at that distance. It appears to have about a 1.25" field of view when 2" away so you can see a relatively large area even without the microscope anywhere near the workpiece. I think that makes it even more useful than I thought. Its very easy to slide it closer or further from the workpiece and readjust the focus in seconds so you could set it close for the initial setup then move it back a bit to see a wider angle for other operations.

Impressive. The USB microscope you’ve got is not a cheapie I take it?

Without stereo, how is one sure they are well placed over an edge? The resolution would be the fudge factor. What do you have in mind?

Strain relief on the wiring, as @patofoto noted, is definitely a challenge. As @patafoto noted, the feed back to the computer is dragged around a lot. Thank goodness the Nomad is a fixed gantry so the range of motion is much more limited than a moving gantry design!

I’ve seen (and used) optics and lasers on moving gantry systems capable of many hundreds of IPS in all directions so this is certainly solvable.

I’m looking into the end of the USB microscope now. It doesn’t appear to be sealed. However I think it would be easy to seal it. The hard part would probably making sure everything INSIDE the microscope assy is secure enough to not vibrate out of position during operation. Its not apparent how the assembly is held together so it may take sacrificing one to figure out how to get it apart without damage so it can be repackaged in a way that doesn’t allow it to move.

Agreed. Most optics are not designed for serious vibration. There are very tight limits to what one can do with optics - lenses for sure - when things vibrate. Lens tend to delaminate their elements, the mounts come loose and so forth.

I would totally shocked in the inexpensive USB optics is well designed for the CNC enclosure environment. It is entirely possible to hit a resonant frequency of of “something” and things will degrade over time.

When I mentioned machining a bracket out of PVC… I meant a PVC block, not PVC pipe.

I certainly understood that, that for making that clear for everyone.

You can get PVC stock blocks are a lot cheaper than Delrin from McMaster or local plastic distributors and it has a higher coefficient of friction so it would grip the scope better with less clamping pressure. PVC is easy to machine and works great for many applications.

As I noted, PVC is fine. I’ve used both and IMHO think Delrin would work better. This is a quibble… but I freely admit that Delrin is more expensive and, perhaps, overkill.

mark

I have and have use one of these:

http://www.lasercenteredgefinder.com/main.html

With practice this is as good as a wiggler or touch off - 0.001". The down side is that it sits in the chuck.

A device like this, parallel to the spindle and at a known offset, would allow one to do a very accurate and precise edge find.

mark

Yes, please! I personally would find this SO useful.

Can not find this article. Got only the intro… I am curious to discover these 10 essential accessories. Thanks

The link seems to be broken, it has been for a while, I wish Make: would fix it!

Here’s a direct link to the author’s stuff about zeroing his machine, on his website:
Zeroing A CNC