Hello Guest it is March 28, 2024, 06:12:42 AM

Author Topic: Digitizing  (Read 100287 times)

0 Members and 1 Guest are viewing this topic.

Hood

*
Re: Digitizing
« Reply #50 on: May 29, 2007, 05:31:35 PM »
Looks good Hood.

Brett

It works great but afraid the first attempt at making up an adjuster was a mess, i eventually got it centered to within 1/2 thou but it was a slow process. I am going to do it a different way sometime but when I dont know.

Hood
Re: Digitizing
« Reply #51 on: May 30, 2007, 01:31:02 AM »
Many responses here since my last one!

My probe definitely is not going to try to compete with the high-end ones, but I think it has enough accuracy for the everyday home mill. One test I am doing is running the "Center of Hole" routine that is built in to the latest Mach3 release. I just run the routine, reset x and y using the "Ref All Home" button and then run the routine again to check repeatability. After about the third time, I get errors no larger than .0002, and more often it is .0001 or even .0000 on the Mach DRO's. The first one or two repeats of the test usually shows a greater error than subsequent tests, so that indicates that the first test or two tend to have a greater error, becasuse the error is reduced every time by running the test three times. After three times I never see a difference of more than .0002. Yes, that's 10-thousandths. But not microns!

For those asking about Global Mapper, you can download a trial version that has crippled saves and exports, but it is good for evaluating the program. It is all mapping related, but it can manipulate elevation models in many ways and in reality we are talking about a kind of mapping anyway. When I open a point cloud file it asks which type of geographic projection you want to use. These are standard projection types that are used on maps and they usually take into account things like the curvature of the Earth (which is not enough to worry about in small areas) and how to display a curved surface on a flat one. You can even specify projections for some of the other planets! Depending on the projection system you pick, the point cloud file is opened using (usually) meters as the basic unit of measurement. If your original point cloud is in inches or millimeters it really doesn't matter because we are dealing with units and not inches or meters or whatever.

In the registered version, you can export to a huge variety of file formats for many programs including dxf points or converted to faces.

The trial version can be downloaded at http://www.globalmapper.com/ (click on "Download Trial" from that page). It can also be downloaded from the USGS at http://mcmcweb.er.usgs.gov/drc/dlgv32pro/. This version (they call it "DLVG32Pro" may be less crippled than the "trial" version -- I am not sure about that because I have the full version already and haven't bothered with the crippled versions. The large njumber of file formats that can be read by Global Mapper are listed at http://www.globalmapper.com/product/formats.htm. The point cloud files generated by Mach's wizard is a simple text file that is recognized by the Global Mapper program as an ASCII elevation grid file.

I haven't tried to use Global Mapper for generating a file to convert to G-code yet and haven't given it too much thought. I have so far used it for viewing point clouds with various shading techniques. On road trips I have a laptop computer with a GPS unit in the car with all kinds of mapping software including Global Mapper. I have digital elevation model files (DEMS) for the entire state of Washington (where I usually confine my road trip to). I also have geo-referenced topographic map files and other map format files and aerial photography for the entire state. These can be overlayed in Global Mapper to produce 3D images of the terrain (what it looked like a few years ago when the photos were taken. I can generate a picture of what it is going to look like around the next bend before I go there. Kinda fun. Anyway, Global Mapper is quite a program and could be useful for converting some formats into something useful for machining purposes, especially when dealing with images that you might want to engrave. I really haven't explored this possibility at all.

« Last Edit: May 30, 2007, 01:35:40 AM by TetraLite »
-Mark
Re: Digitizing
« Reply #52 on: May 30, 2007, 02:52:02 AM »
Some new pix of my probe (shaded in ACAD with the housing not shown so you can see inside):

In the above image you can see the springs that push the PC board down against the bottom cap. The setscrews used for adjusting the plane of the PC board thread through the bottom cap and push up against the bottom side of the board. Also note that I have added the jack for plugging in a 1/8" phone plug.

In this picture you can see the holes through the bottom cap for the setscrews and the long (2-56x1.25") screws that hold the probe together.
The setscrews are 6-32x1/4". The probe is shown here without a stylus which screws onto the 4mm x 0.7mm x 30mm bolt that runs through the hub (the part that holds the contact pins). I used the metric screw here because it was a better length than the cloest standard length #8-32 screw, so my stylus must have 4mm x 0.7mm female threads.


And here is a scan of FDR's ear on the front of a dime. Notice the big peak on his ear lobe. This must have been due to a false activation of the probe, perhaps caused by a vibration on my mill's bench. The image was generated in Global Mapper and it extrapolates the elevations between sample points. That is why the sides of the peak on the ear lobe are tapered to meet the surrounding sample points.  I haven't seen it do this before and I think it is because my latest assembly of the probe didn't put as much compression on the spring that pushes the pins against the balls making it too easily opened by vibration. The vertical exaggeration in this image is 10 times nomrmal. The scan was 0.12" by 0.12" with x and y steps of 0.002" making a total of 3600 sample points.

Below for download is my latest ACAD drawing in .dwg format. It is a three-dimensional drawing -- the same that I used to create the images above. If I save it as a .dxf file it is over 23MB. As a dwg it is 868KB. If you can open this 3D file in your CAD program you can measure everything or look at object properties to get dimensions which I have not put on the drawing. There are separate layers for everything.

« Last Edit: May 30, 2007, 04:22:02 AM by TetraLite »
-Mark
Re: Digitizing
« Reply #53 on: May 30, 2007, 03:49:09 PM »
The resolution is amazing!  And the display is very nice after it has been run through the mapper.
How is the probe doing for measuring in X-Y planes and repeatability?

I saw a wizard for finding hole centers but is there an option to use it as a regular CMM function
where the ball diameter is subtracted from all measurements, X-Y-Z.

Even further, how do I go from either a point cloud or controlled data points to a 3-D CAD model?
TIA
Bertho
Re: Digitizing
« Reply #54 on: May 30, 2007, 04:30:55 PM »
I can answer all three of your questions with confidence: (1) I don't know, (2) I don't know and (3) I don't know!

I haven't really done much in the way of testing yet. So far I have just tried to see how well it can scan very small items using a needle point for a stylus. Well, I have done the hole center thing I described above and those tests show repeatability easily within a thousandth of an inch. I get repeatability errors of about 3 ten thousandths maximum (I haven't done a lot of tests). All this is, of course, dependent to some degree on the resolution and accuracy of the milling machine itself, determined by leadscrew slop and stepper motor resolution and whatever else there is to take into account. But it looks to me that if you're not shooting for accuracy better than a couple thousandths or so, this probe should be close enough.

Modified entry: I have noticed that in my scans of quarters and dimes that every other line is shifted a tiny amount along the x axis. This must be, I presume, due to end play of the x-axis leadscrew (what else could it be?). I am glad my machine is that tight! Perhaps it should be noted that my stepper motor settings for "Steps per" (which is steps per inch) are at 32000 which means in order to move a stepper motor one step on my machine I would enter G1 X0.00003125 and the motor moves one step, assuming X is starting from zero. Whether on not the table moves depends on end play and which way it moved the last time it was moved, I suppose. So I am set up for pretty fine resolution.

I haven't seen any options for CMM function either.

I have been wondering about the best way to get from point cloud to 3D CAD models myself. I will be experimenting with that soon I think. Seems like I saw something about that in this forum somewhere.
« Last Edit: May 30, 2007, 05:02:25 PM by TetraLite »
-Mark

Offline Chaoticone

*
  • *
  •  5,624 5,624
  • Precision Chaos
    • View Profile
Re: Digitizing
« Reply #55 on: May 31, 2007, 07:53:45 PM »
Amazing. ;D ;D ;D ;D Looks real good to me.


Brett
;D If you could see the things I have in my head, you would be laughing too. ;D

My guard dog is not what you need to worry about!
Re: Digitizing
« Reply #56 on: May 31, 2007, 10:41:40 PM »

Modified entry: I have noticed that in my scans of quarters and dimes that every other line is shifted a tiny amount along the x axis. This must be, I presume, due to end play of the x-axis leadscrew (what else could it be?). I am glad my machine is that tight!


Actually this is a function of the probe plugin for Mach3. I have probed several parts and I have the same thing.

Mike
Re: Digitizing
« Reply #57 on: June 01, 2007, 07:06:28 PM »
If that is true that would mean that the probe plugin is generating errors in the output file that would amount to elevations being reported at the wrong (shifted by 1 x step) x,y point. Is that what you are saying? 
-Mark
Re: Digitizing
« Reply #58 on: June 05, 2007, 04:07:43 AM »
I have begun assembly of several probes and when I have a few built I will begin offering them for sale on a web page at http://tetralite.com/probe. I have completed the time study and I have decided the price will be $150 plus a flat rate for shipping anywhere in the US plus a PayPal fee (orders will be only through PayPal). I will not be selling to foreign countries. I just don't want the extra hassle of filling out customs forms and calculating the shipping charges. I have enough of that to do with my other product (TetraMouse).

During this short initial run I will be refining my tooling (jigs and etc.). I am considering offering the probe without a stylus, which would mean it would be a do-it-yourself stylus situation. On the other hand, if I come up with a simple all-metal one piece stylus I will offer that. The problem is I can find no info regarding how to attach ruby balls to a stylus for the tip. Anybody know about that?

As soon as I have probes for sale I will put a PayPal link on the above mentioned web page. If I get more requests for probes I will begin another run right away. Lead times will be short -- a couple or three days probably. I will try to keep an up to date total of probes on hand on the web page. If I have to put orders on hold and anyone doesn't want to wait, I will refund their PayPal payment at their request immediately.

The first run is 5 units and two of them are already spoken for. I expect to put them up for sale within two days (June 7th probably -- maybe tomorrow the 6th).
« Last Edit: June 05, 2007, 04:10:15 AM by TetraLite »
-Mark

Offline Jeff_Birt

*
  •  1,107 1,107
    • View Profile
    • Soigeneris
Re: Digitizing
« Reply #59 on: June 05, 2007, 09:10:46 AM »
Why not have them accept a standard styli?  I've bought some from here before http://www.itpstyli.com for my old CMM.
Happy machining , Jeff Birt