PDA

View Full Version : Touch Probe for Digitizing



jb2060
17th June 2008, 12:10 AM
Hi All,

I have just purchased a touch probe from Arnie (Mach3 Forum).
http://www.machsupport.com/forum/index.php/topic,5004.0.html

I wanted to get one of these for a while but couldn't really justify the price. At about $110 it seemed very good value. I have a custom made driver board so it may be a little challenging for me to get this working. With the kind help of rodm and greolt I managed to previously get a touchplate working nicely - so maybe I can blaze the trail on this one.

When it arrives and I "attempt" to install it, I will give a further update. If anybody has a similar one already, I would be keen to hear from them.


http://www.machsupport.com/forum/index.php?action=dlattach;topic=5004.0;id=4679;image

rodm
17th June 2008, 12:49 AM
Hi Jason,
It certainly looks the part.
Are you trying to create machine code from digitising a solid object? What I mean is reverse machining.
Art (Mach3) has done a lot of work in this area.
Be very interested to follow up on this one.

Greolt
17th June 2008, 08:20 AM
Jason

I had one of Arnie's probes that I demonstrated at the CNC BBQ.

It is a nice unit for the price. Not difficult to hook up.

Worked well but I did not spend sufficient time with it to get auto probing code working.

Greg

jb2060
18th June 2008, 12:41 AM
Hi Rod, Yes I want to digitize some of the contours and shapes that I have hand caved. Some of the guitar necks I have made previously have been requested by new clients and it is very difficult for me to replicate fully by hand carving again. If I can digitize and rough carve my existing shapes then I can finish them off by hand and get really close to the original. Thanks for the Tip - I will follow up with Art when I get things going... cheers, Jason

Hi Greg, Well it certainly doesn't surprise me that you have already got one of these - I would describe you as a 'power feature' CNCer :) It is also very encouraging that you are happy with the quality too. I would be keen to get some help at the set up stage (unless I can somehow jag a fault-free install). The auto probe digitizing is going to be key for me so hopefully I can share some information back when I get to that stage.
cheers, Jason

jb2060
11th October 2008, 02:55 PM
Hi,

I have finally got around to installing my probe. I realised when I tried to put it into the collet that the shaft was 3/8ths not a 1/4. I had to drill a hole in my mount and then add a grub screw.

Now that is done I am having a little bit of trouble doing the wiring. I am okay with the ground and the tip but the middle connection is supposed to be connected to pin 18 -25 on the parallel or to common ?? I have got no idea what common is and I can't see any access to pins 18-25 on my Xylotex board.

I want to install this so I can keep my height setting touch plate as well. I believe this tool is much better for use on the very fine bits that I use for inlay. I guess I will keep them connected into the current positions.

Greg: I know you have installed one of these but I am not sure it was on a Xylotex board.

Has anyone else out these had any success?

snowyskiesau
11th October 2008, 05:43 PM
Pins 18-25 on the parallel port are earth or common.

rodm
11th October 2008, 05:57 PM
Hi Jason,
Here is my take on it but remember this is mubo jumbo to me.
Pins 18 through 25 on a parallel port are ground. If you have the 4 axis Xylotex board then it goes to the right of pin 17 marked GND. If you have the 3 axis board it goes to the left of pin 17 marked GND.
The input signal from the probe can go to any pin 10 through to 13 as well as 15 but you will be using this for the touplate.
The ground to case it just shielding the probe from EMI. I would take it back to your common ground or if you don't have one then to the ground on your power supply. You can do this on the 4 axis Xylotex board to the right and higher than pin 17 - it says VCC and GND but obviously only connect to GND. Do not use the pin you grounded the probe to as this is the parallel port or computer ground.
I was using the Xylotex drawing at the bottom of the page here
http://www.xylotex.com/XS3525V400.pdf
Now the smart thing to do would be to wait until Greg responds.

jb2060
11th October 2008, 08:53 PM
Thanks Geoff...

Thanks Rod, Sounds right to me (It was the "common" thing that stumped me:-). Your explaination really helped. I will wire it that way and see how it goes.

I am wondering if I can use pin15 for both the touchplate and the probe. This means I will add the probe tip wire to pin15 (where the touchplate currently resides) - add the ground to GND (where the touchplate Ground currently resides) - and add the shielding to the "VCC/GND" GND.

What I might end up doing if I cannot wire them together is to add a stereo jack retrofit to my touchplate so I can interchange them as i need them.:rolleyes:

rodm
12th October 2008, 03:10 AM
Hi Jason,
I think the jack plug might be the best option as i don't think I would wire both permanently to the same pin. Will Mach3 handle both on the same pin - it might cause a conflict.
I don't see a problem with wiring the probe or the touchplate to one of the other input pins though so you might want to consider that.

Greolt
12th October 2008, 03:15 PM
As far as Mach is concerned both the Touch Probe and the Touch Plate are the same thing. Mach is looking for an input from "Probe" or also referred to as "Digitize".

So it is looking for the input pin to change state. Depending on whether it is set to "active high" or "active low"

The way we make the touch plate as simple as it is, is to have the input high so we can change it's state by bringing it low or grounding it. Touching the tool or "closing the switch".

Touch Probe works a little differently. It operates by "opening the switch"

So what I did and what I think is the simplest way to do it, is to have your already installed touch plate connection via a 3.5mm socket at the control box. Plug in to that, either the touch plate or the touch probe. Obviously not both at once.

The less convenient bit to get around the differences in the two devices, is to go into "ports and pins" and set "active low" when using the plate and "active high" when using the probe.

As with most things there is more than one way to skin a cat. This is just one way. :)

Greg

jb2060
12th October 2008, 11:07 PM
Rod/Greg,

It is working ! I will post a picture of the point cloud once I get it into a 3D CAD program. For now - here is a video of it working. I am running a scan across one of my guitar bodies.

rodm
12th October 2008, 11:35 PM
Hi Jason,
Good to see you got it working. Can you show us what it produces and how it does it - I'm really interested in this.

jb2060
13th October 2008, 09:39 PM
Rod,

Here are the results of the scan. I ran it over 1/4 of the top carved surface of the guitar with 2mm incraments 1200mm/min (because of the short bursts it rarely got above 500mm.min) and I ran it 10mm deep. The probe did get stuck every now and then which was a little frustrating and I had to flick it to get it restarted.

It took about 3 hours to get this much. The pics show an angled view of the pointcloud and a side view. I have not created any surfaces from these yet (gotta learn how to do that).

I am going to run it again tomorrow with a much larger grid spacing just so I can get a complete surface.

rodm
13th October 2008, 11:04 PM
Hi Jason,
Very impressed. I can see the potential for that. :2tsup:
You must have your machine sorted out now otherwise you would not have been able to do that.