New (to me) VMC

[ostie01]

Maybe you could try some of these edge finder that beep when you touch the part, you could use it to send the signal.

Jeff

[VPT]

You cnc guys always amaze me!

[forhire]

I ended up buying a Wildhorse Inivations probe off Ebay for $80. The price just couldn't be passed up. I've been messing around with it for a few days. Right now I have it repeatable within a thousand or so but the overall reading is off by about 5 thousands consistently. I can correct that in CAD. All said, I'm still working out the kinks but I have it working well enough to digitize a plaster cast off a tool tote. The probe left little dimples everywhere it probed on the plaster so I had to adjust according. Being mechanical it's fairly stiff. I probed three times around at different heights and all three traces were consistent between them, even on the 5 degree draft. The cut profile fit perfectly. The only thing that would have made it better would be to profile the outside with a 5 degree endmill.

[forhire]

I'm still trying to get the probe to be accurate. I've learned that the probe "electrical radius" is less than the physical radius due to pre-travel deflection. That accounts for the consistent variation I was seeing. I used a test ring to calibrate it and I'm within a few tenths in all directions but one direction was always reading high about 3 thou from the others. I figured the body must not have been flat. So today I took off the wimpy 1/4" shank and discovered the thread wasn't straight. It had also raise a burr. So I decided to make an adjustable shank. I made a short 45 degree taper that screwed down to the probe body using the stock #10-24 thread. I then make a shank for a 3/4 CAT40 tool holder. I'm going to use 4 conical set screws to retain the probe and adjust the center. This will then allow me to use the planer adjustment screws to fine tune. I dry fit it with some #8-32 screws... I'll have to pick up setscrews in the morning. It feels a lot more rigid.

[forhire]

After spending a bunch of time looking for conical set screws... and not wanting to wait to order them... I opted to grind my own. I made a quick mandrel from chromed rod. I started by drilling and tapping for the #8-32 set screw. I then flipped the mandrel and drilled it from the other end to allow a hex key to reach a set screw used to lock the screw being ground. I then dialed in the mandrel on the whirligig and quickly ground the cone. Easy as pie. After that I assembled the probe on the new shaft and dialed the body in concentric. The four set screws allowed me to adjust it concentric, just like a four jaw chuck. I used a hex key on each side, loosen one and tighten one. I then used the screws on the probe base to level probe and dial it in concentric. I'm now repeatable within tenths and consistent between each axis direction within a half thousand. I calculated my probe electrical diameter to be 0.1165". Using this adjusted diameter I can now find hole centers and other measurements within a few tenths. The new shank solved my problem.

[forhire]

Here's a video of the touch probe in action. Actual digitizing time was about 20 minutes... sped up to spare you the pain.

[Canner]

That is awesome, what program do you use with the probe readings to compile the data points and put them into a useable format? Do you expect to use this for checking parts after they are machined , locating stock, orgetting dimensions for a job? I know it will work for all but where do you think it will be used the most? Thanks

[forhire]

That is awesome, what program do you use with the probe readings to compile the data points and put them into a useable format? Do you expect to use this for checking parts after they are machined , locating stock, orgetting dimensions for a job? I know it will work for all but where do you think it will be used the most? Thanks

I'm using the EZ-Digitize program that came with the machine. It's DOS based, hi-tech circa 1994. In addition, the machine jog menu can use the probe to calculate basic geometry like hole centers, mid-points, and intersections. I've only used it for XY 2D stuff so far. It can export in DXF or g-code. It can also do 3D mesh and some other stuff I haven't figured out. Unfortunately the manual for EZ-Digitize is lost to time so I'm flying blind. I'd love to get the "EZ-Digitize Reference Manual" if someone has a copy.

If you look at post #263 you can check out the bottle tray insert that goes into a carry tote. The profile was digitized from the plaster mold and then I added the square pockets.

I think I'll likely use it mostly to trace parts like the tray. I frequently have parts that I need to grab an odd shape or profile. I've wanted something like this for a long time and I don't have much invested in this probe. If I find I'm using it a lot I may upgrade to a more sensitive commercial probe. I'll need to get a little more time on the probe before I'm ready to trust it for measuring.

[forhire]

I've been thinking about adding some CAT40 tool storage to the TC-3 for a while. So yesterday I milled out a couple basic 9 tool holders from a scrap of 0.125" aluminum. I originally envisioned making it longer but my brake is only 26" inches wide so that's how wide it is. The upper flange is 2" and the lower lip is 1". I was concerned with the weight of 9 tools but it appears plenty rigid. I did over-bend the shelves a degree or two to compensate for deflection when loaded. Getting the angle of the cabinet was trivial. I simply took a 4' level and measured from the top to the vertical level and then solved for the triangle. In this case I measured 11 inches and it came out close to 78 degrees. I decided to put the tools on this sloped face because they would be out of the way when walking by... and I can't see through the old window anyway.

The holes are 1-15/16" and they are spaced 2-3/4" inches apart. Here's a few build pictures. Looking at the photo of the racks installed they don't appear level but they are. Odd photo angle I guess.

I made three but I'll have to do some shuffling before installing it.

[forhire]

Sometimes I spend more time fixing my junk than running it, or at least it feels that way. Fired up the mill and the computer wouldn't post. A quick check and it was clear the power supply was likely the cause. No problem, I'll swap it, or so I thought. For starters the Bridgeport engineers put a raceway in line with some of the screws. I had to use a 90 degree screw driver which took an eternity to remove. Once I had the power supply in hand it was as expected, standard AT power supply. Top of the line 1994.

The fan had frozen up so hard it no longer could be turned and as a result the power supply board was very clearly baked. Come to think of it... it has been kinda quite. So I hunted around and found a very old Zenith computer (Yup, they made computers, scary I know) in the storage room that I'd been using as a step stool. Pulled the power supply and thought I'd be back making chips in minutes. Nope. The cables were way too short and they exited the power supply on the wrong side to fit in the cabinet. Argh! Poked around the internet and I could easily order a power supply or even an ATX to AT adapter cable. Well not really wanting to wait for one to arrive in a few days I rolled up my sleeves and did a board swap. Nice that most AT power supplies have the same board dimensions. My Hakko desoldering gun made amazingly fast work of the massive amounts of solder leaving the holes very clean. I then soldered the longer cable assembly to the Zenith board. I did have to break out my large soldering iron as my small one just didn't make enough heat to wet out so many connections. No cold solder joints allowed. I then assembled it and tested it on the old Zenith pc... just in case I goofed. Worked perfectly. Put it back into the mill and it's back up and running.

Soldering counts as welding right?

[israelorama]

Thanks for the polishing advise. I need to order something and try it.

I finished hanging the EMT conduit and pulling the cat5 today. Went pretty smooth aside from the pesky RTL8139 chipset requires PNP to be enabled in the bios even though it is DOS. I suspect it has something to do with the fact it's a PCI chipset on an ISA SBC. All said the packet driver works and it didn't break anything.

I used mTCP http://www.brutman.com/mTCP/ which has a basic FTP client and some other tools. I also used the NTP client so I can synchronize the clock with my NTP server on bootup. Nifty. Here's a screen shot of the FTP working.

Hi forhire, your awsome, excellent job, please I would like to do this on my TC, could you please tell me what I have to do first like step by step to load programs by FTP ethernet?, I will appreciate so much your help.

[forhire]

Hi forhire, your awesome, excellent job, please I would like to do this on my TC, could you please tell me what I have to do first like step by step to load programs by FTP ethernet?, I will appreciate so much your help.

To use mTCP you'll need a packet driver for your NIC. Do you know what NIC you have installed?

Here are a few links that will get you going.
http://www.brutman.com/mTCP/mTCP.html
http://www.brutman.com/Dos_Networkin..._packet_driver

Normally you'll find the packet driver in the DOS folder of the NIC driver zip archive. This site as a bunch of the more common packet drivers in one place.
http://www.georgpotthast.de/sioux/packet.htm

Basically you load the packet driver on boot up. Run mTCP DHCP client to get an IP address. Then open the mTCP FTP client to transfer files to and from your FTP server/host.
http://www.brutman.com/mTCP/mTCP_DHCP.html
http://www.brutman.com/mTCP/mTCP_FTP.html

If your using a static IP address simply edit c:\mtcp\tcp.cfg

Alternatively, you can use MS LANMAN to connect to windows file shares (SMB) but because LANMAN is so old you'll have turn off some of the security features on your server. In the end I prefer using FTP because it has lower memory requirements. And I don't need to keep a connection open to the server all the time. I just connect and download my g-code.