3D Digitizing Probe???? Who has one and whats the low down on it?

[dray]

I want to replicate a few pieces I have. How long does the probe take to probe something?

Are there any other alternatives for replicating in 3d?

[Brady Watson]

It depends on the size of the part and the resolution you wish to capture. I've scanned many parts on the Bot. The longest scan was a 18X30" 3D carved panel that took 6 days to completely scan. I can now scan the same part at 100X finer resolution with my laser scanner in about 6 hours. It all depends on what you need. Overall, touch probes are simple, reliable and economical. As with anything there is no free lunch and the price you save in the initial cost of the probe, you pay for in the time it takes to scan. For most people this is not an issue. Just scan overnight or over the weekend. The probe that ShopBot sells is good & will last a long time as long as you don't do anything stupid with it...like try to use the Zzero plate with it!

-B

[mitchjr]

Where did you get the laser scanner and how much was it ???

[Brady Watson]

I got it at Walmart. It was on sale for $49


Seriously...It's my design & I'm not at liberty to discuss the system or price. You can contact me offline if you need laser scanning done as a service.

http://www.ibild.com/3d.htm

-B

[cjohnson]

Danny,
I bought a probe but I damaged it trying to learn how to use it. ShopBot replaced it at no charge because of a possible glitch in the earlier version of the probe software. I believe that the current probe software is safe to use but I am waiting on the user manual to be written before I take another shot at it. I think that if I had to do it over again, I would do more research on laser scanning. It looks very interesting.
cj

[dray]

I found some links on making your own 3d scanner. Im gonna give that a shot this weekend hopefully. The problem is that the pieces I want to scan in are pretty big so I think Im gonna try and put a 3d scanner together.

I read 4 different essays on it last night and the principles are pretty basic. Theres also a guy here north of orange county that made a big one by himself and he uses it to reproduce furniture and molding profiles etc. I think Im gonna go have a look at his and get a feel for it if I have time tomorrow.

Anyways if anyone else is interested heres a link

http://www.muellerr.ch/engineering/laserscanner/default.htm
(good concept here but he needs to do it in an enclosed dark box)


8381

8382

[steve4460]

Hi Danny

So if I have the indexer motor hooked up to my bot then I could use that to mount it under my table pointing up then attach a rod with my laser mounted to place the camera and the part on the table put a box over it , and then use the bot software to control the motor turning the laser .Then comes the hard part to figure out what to do with the images . Let us know when you find the software for that .
good luck
bot on

[dray]

Heres some software in python
http://www.splinescan.co.uk/downloads.php


and heres the whole ball of wax from stanford
\newurl
http://graphics.stanford.edu/projects/mich/


=P

[dray]

sorry heres some free windows software that will bring xeroxing your butt to a whole new level
http://www.brucerayne.com/software.html

[steve4460]

Thanks Danny
That last one looks realy good I will give it a try .
This way I get to use my indexer driver and motor even more .

[andre]

This is peaking my interest greatly. I can understand why brady does not want to share how he built his. Danny that web site doesn't explain how to build it. Brady if you would not mind sharing, where would be a good place to start building one. where in the heck would you purchase a laser?

[steve4460]

Hi Andre

Try one of those laser levels at menards.
bot on

[Brady Watson]

Gents,
Let's not confuse high-end industrial hardware collecting data points via CNC, with 'build your own hardware store scanner'. It's not the same. At all!

If you want to do a neat science experiment with a laser line generator and webcam, go for it! It can be fun, but in the end you will be most like be frustrated. I have tried/built/experimented extensively all of the ones mentioned above and then some. Your results will be a hardly recognizable blob. But...don't take my word for it. Try it out for yourself. You gotta do your own R&D to find these things out.

There are low-end 'commercial' scanners out there like the NextEngine and Roland laser scanners that also leave a lot to be desired. They are just OK in their scanning ability and resolution is no better than .002". That sounds small, but it isn't in laser scanning. I can pull in data 100X finer than that. I have evaluated everything available under $40k and I really didn't like any of them for one reason or another...3D cameras, lasers, structured light etc. I could have bought 2 Alphas for what I have wrapped up in my system...really! So it's a bit more than a webcam, turntable and hacked laser level. What's interesting is just how many of my customers own NextEngine or Roland scanners, yet still have me scan their parts for them.

Aside from having a good scanning system, you need the software and expertise to cleanup the data and manipulate the scan. Not all laser technologies work the same, and depending on the one that you choose, you may be doing more cleanup in software than you anticipated. Don't overlook the power of a good touch probe...I know it isn't as sexy as a laser, but it works and and it's cheap. There's no free lunch with scanning. It's either a lot of time (probe) or a lot of money (laser) to get into the game.

-B

[paul_n]

yet another to look at:

http://www.intricad.com/index.html


Paul

[dray]

Ill spend some time tonight researching the different types of technology for laser scanning.

it does seem that the camera/laser ones give poor definition. I am unsure if this is because of reflective surfaces or unstable turntable or what.

I think that a laser or mulriple laser range finder types may also work.. Lemme do some research

[Brady Watson]

Just an FYI...The LaserRanger by Aculux is no longer sold on an individual basis. You need to buy them 25 at a time...Figured I'd mention that in case one of you stumbled on that as a possibility. LR1 is @ $1250, LR2 @ $2200.

-B

[Brady Watson]

Paul,
Have you used the intricad/Mauch system? Have any pics of your scans?

-B

[myxpykalix]

I know next to nothing about laser scanning although i'd like to have a good one. But one thought occurred to me. Rather than have one of these laser levels project a straight line on your object wouldn't it be more advantageous if you could have a grid pattern with dots projected onto it instead? And if you could somehow adjust that grid pattern? Would that be a different lens or could that be adjusted thru the computer? If for no other reason it would give you a "map" to use a touchprobe with. Then again it might be easier to use the shopbot probe.

[Brady Watson]

There's absolutely nothing wrong with the SB probe. In fact, it is very easy to use & it should be the starting point before one invests in a laser system. Without doing this, how do you know how much better the laser is? In most cases, the SB (or home-made) touch probe will run circles around a laser in terms of overall ease of use, lack of spurious noise & data anomolies. A Roland Picza touch probe is also an excellent little device. Even though a laser will scan faster than a probe, the process of getting a good scan from one object to another, may take you 5 to 10 tries in many of the systems out there. If you have the time to monkey around with EVERY scan that comes thru the door that's one thing...I'd rather run a touch probe, set it & forget it...knowing the scan is going to be nice when it's done. Microscribe & Faro arms seem like a great idea until you actually use one...you need good hand/eye coordination (or more accurately hand/eye/foot coordination) to use it AND you physically have to be there for the scanning to get done.

I used the SB probe for larger 3D objects and the Picza for smaller objects for about 4yrs. Each are very good in their own way, and in the end you get a good scan. Before everyone jumps on the laser bandwagen (it's a bumpy and often disappointing ride, BTW), explore what a touch probe can do for you 1st. Several of us have put together our own touch probes for under $25 (it's just a switch afterall) and all of the probing magic is already built into SB3 by ShopBot. Plug it into input 1 and go to town. The SB version is much nicer than the one I cobbled together since it has optical switches inside of it that never wear out.

-B

[paul_n]

Hey -B....

Nope, still just 'window-shopping'. You and I talked about Mauch'ers system over a year or more ago, and it looks like it might be time to take -another- look.
Not having an Indexer, or even planning one, a flat-bed scanner / probe is of more interest to he, at this time.

Paul

[bill.young]

The most promising DIY laser scanner project I've played with so far is DAVID at http://www.david-laserscanner.com/

In my experience none of the laser-and-camera setups will give you a scan that will be nearly clean enough to create a part file from, but if you're looking for something that you can use as a basis for creating a 3d model in something like Rhino then DAVID shows some promise.

Bill

[dray]

I am thinking of laser ranger mounted on z axis using a similar routine as the probe except it will not have to go up/down , just add those calculations into the probe routine from the output of the laser ranger. I think it can be scanned in fairly quickly if rastered.. the high end models can take in 10,000 points per second I believe

Doing more research as I have time

[drodda]

I have a digitizing probe that I bought from shopbot still new in the shipping box. If someone is looking to pick one up at a slight discount.


-D

[dray]

Has anyobne tried using a laserranger or other laser range device?

[bruce_clark]

Danny,

Yes, I have. Matter of fact, I did a "retrofit" of a desktop CNC machine with an Aculux range finder and custom control software. It was faster than a touch probe and the data was decent, but the problem with the Aculux is that areas that cannot be "read" will give you a result of zero, so you get a lot of "holes" in your surface mesh.

If you want more info, email me.

Bruce

[jseiler]

Another method for 3d is normal map photography.
I've not had much time to play around with this concept, but it looks very workable for shallow reliefs (most of what botters want to do anyway).

http://zarria.net/nrmphoto/nrmphoto.html


I've also had some time to play around with the david scanner. Its really sensitive to camera settings and light in the room. My scans have ranged from pretty good to really poor depending on moon phase. They just upgraded the software again and now you can buy software to stitch slices together.

Another one to look at might be:
http://www.camtronics-cnc.com/
No word on how good it is yet.

John

[Brady Watson]

John S.
Please post some pics of the scans that you got with the David. I'm curious to see if it is any better than when I tried it. Yes...lighting is critical.

-B

[andre]

Paul
have you used intracad? Looks interesting. What is the final result of a scanned object? is it usable for detailed toolpathing?

[andre]

Another question
What software is needed for the shopbot probe?
How do I take the 3d image and manipulate it?

[Brady Watson]

Andre,
Paul and I both looked at this system a year or so back. Neither of us own it. It's only a few hundred bucks...get one and try it!

You can scan to an SBP with the probe and output a SBP and cut a duplicate OR convert the SBP to a DXF, bring it into ArtCAM and do what you want to it before you toolpath it...OR just bring the DXF right into your CAM package (MillWizard, Cut3D etc) and toolpath it.

-B

[dray]

Bruce.. can you post a pic of the scans you got with the Aculux laser ranger?

I think this is the best option. I have a box of THK linear guides and ballscrews. I am thinking of making a simple scan table or box.

All of the camera/laser scanners pretty much send the same output of blobs no matter how nice the webpage is. There are too many possibilities in that method , lighting, surface reflection, contour ,lens, shadows etc etc.

But a laser ranger type combined with a stable x/y gantry just might be the ticket for getting the best results.

For my applications and turn around time a probe is just too time consuming.

I am less concerned about the price and more concerned about time/detail.

[garyb]

From a reasonable price scanner..........



8383

[myxpykalix]

Gary,
You weren't in the movie Goldfinger were you? Tell us some details on this scan above please? how did you make it, hardware, software, ect?

[garyb]

No but wouldn't the midas touch come in handy at times.........
This scan is a one that comes with the scanner, I just loaded into AC and grabed the image.
As far as the rest, well mum's the word right now but for those attending the KS camp, well there's your teaser..............

[Brady Watson]

Gary,
That's not a bad little device...although it would burn me up if I couldn't scan metal with it. It's nice for scanning humans though...if you can get them to pay for it!


-B

[Brady Watson]

Here's one I did recently for the company that makes these for West Point Academy:


8384

8385

8386

[garyb]

Very nice Brady, is that for the lapel badges?
Scanning human parts is easy its convincing people to stretch out on the bot table for cutting would be the trick

Actually this scanner was developed for the automotive industry so scanning metal is right up its alley

here's another



8387

[Brady Watson]

Gary,
Yes it is a lapel pin. Good eye!

Nice ball & claw. I'm in the process of making a 4th axis scanner for stuff like that...Too bad I can't clone myself to get more stuff done! ...scary thought, I know


-B

[dray]

Gary what scanner is that?

[dray]

LOL I was just scanning through the boards... There are some really odd people on these boards I was looking up the 3d scanners on cnc zone. Waaay back in the archives some of the shopbot people here ask alot of questions and people chime right in but they but never return the favor, they use other people's information and call it their own.

I have never understood these kind of people.

Anyways.. Ill do research myself and post it freely again.

[drodda]

Guess your going to have to go to the KC camp Danny?

[dray]

I am going to the jamboree on 17th. I dunno what kc camp is

[jseiler]

Brady,

I pretty much scrapped all my 3d scans, but in the next couple of weeks, I should be able to take another stab at it. I had a hard drive crash and haven't recovered everything off my old hard drive yet.

I did post on the david laser scanner forum and someone responded with some things to try. I've not tried those things yet. Its really sensitive to room light (even from the backlight of your lcd monitor), laser quality (low power is ok if the beam is thin and doesn't have blobs...I have 2, one is high power but it beam line is pretty thick, and a cheap one with a blob on one side but it leaves a nice thin line).

I probably didn't get anything better than you did, and I certainly don't get the detail you get from your system, but I suspect that if someone wants general 3d shape and is willing to smooth the scans by hand in 3d cad, the results would be ok for the price.

John

[odulfst]

I'll be making a model of my car w/ a laser scanner used for GIS studies by NASA. Supposedly it can do 1 mm res at 1/4 mile with textures. that was my friend who is learning to use it throwing out numbers though. shoul dmake a nice model of the car for wind tunnel testing.