2nd Build (first) - CNC Router

[David Falkner]

It isn't that so much as what happens with Windows when the power goes out. The UPS will make your Windows installation much more reliable. Keeping the CNC running is not the reason for the UPS.

I'll probably pick one up pretty soon. I don't really like having a desktop computer not on a UPS. Thanks, Paul!
+++++++++++++++++++++++++++++++++++
I drained the water out of the system and am preparing for distilled water and the RV antifreeze but thought I'd mark the lines so the pump will know which way the water flows when I turn it back on...

[David Falkner]

When the machine came in it was missing a few links in the cable chain and those arrived today so I was able to complete the cable run on the gantry. I need to get the correct bolt and then decide how I'll secure the cables and water lines but at least now the cable chain is in place.

[David Falkner]

I set up the homing based on what I've read and seen online but I'm not certain I have this correct. Many machines I've seen move very quickly toward home and when they approach the sensors they slow down. The way I have it set it's just slow across the board. It's ok if I jog it to a couple of inches away from home but if it has a long way to travel I might as well pack a lunch. So... I wonder what I have not yet set or if I'm just mistaken on how it is suppose to act.

Homing Speed is set to 20% but it was 5% so it is much faster now but not sure it's correct.


Here's a short video of the homing -

[David Falkner]

I set up to do a little calibrating and found the machine was out more than I thought it would be. The first test on the Y axis was 0.014" out in one inch. In the full travel of the Y axis, 50", that is almost 3/4" off (0.700")! So I used a dial indicator and did a little tweaking to the motor configuration and the Counts per Unit in Mach4. Now when I tell it to move an inch it moves an inch - poifect!! The only drawback is that the DRO on the screen in Mach4 is off by 0.0002" which isn't very much at all but I just don't see why it isn't 0.0000".

Here's a video to better show what I'm seeing -

[Bruce Page]

Can you set up Mach4 to display just 3 places? A machine shop Jig Borer will run "tenths" all day. I don't see the need for a 4 decimal place readout on a CNC router.

[David Falkner]

Can you set up Mach4 to display just 3 places? A machine shop Jig Borer will run "tenths" all day. I don't see the need for a 4 decimal place readout on a CNC router.

I hadn't thought about that, Bruce. Before I calibrated it, when it was off by 0.014", it went to all zeros every time. But now that it's actually going a true inch the DRO is off and I don't get it. I doubt I'll ever see it like this unless I calibrate again. Running jobs won't likely ever show 0.0000 on the display.

[David Falkner]

If you watched the previous video you probably noticed that I wasn't satisfied with the slight error in accuracy. So, I changed my methodology and process and now it is dead on and I'm good with dead on. I wrote a spreadsheet to use the current motor settings, take the error in accuracy, and then calculate new motor settings. Each time I ran it the errors became smaller and the accuracy got better so after running it about 5 times it is now dead on.

Here's the new process and video -

[Brian Lamb]

Hi David,

I don't want to bust your bubble, but indicators rarely travel exactly 1" when they say 1". You have all sorts of errors that can show up, if you don't have the plunger perfectly parallel in two axis to the direction of movement you get a sine error. If your indicator actually swings .0997" per .1000" revolution, by the time you spin it 10 times for an inch, you can be quite a bit off.

You should set travel distances using an accurate gage block, for example your Z axis, put the indicator into or on the spindle, set a 123 block on the table in the thin direction, 1" thick. Stand another 123 block on top of it so it is 3" tall. Zero your indicator in the top of the 3" block, slide it out of the way and using MDI command a Z-3.0 move and see if you get zero on the indicator again. You can do the same thing in the X and Y axis, just have to clamp one block solid so it doesn't move. I like to use 123 blocks as they are usually accurate on size within probably less than .0002", but I also have gage blocks called "jo blocks" up to 6" long in a single piece and like to use those too.

[Jerome Stanek]

Are you sure there is no back lash in the gear box

[David Falkner]

Hi David,

I don't want to bust your bubble, but indicators rarely travel exactly 1" when they say 1". You have all sorts of errors that can show up, if you don't have the plunger perfectly parallel in two axis to the direction of movement you get a sine error. If your indicator actually swings .0997" per .1000" revolution, by the time you spin it 10 times for an inch, you can be quite a bit off.

You should set travel distances using an accurate gage block, for example your Z axis, put the indicator into or on the spindle, set a 123 block on the table in the thin direction, 1" thick. Stand another 123 block on top of it so it is 3" tall. Zero your indicator in the top of the 3" block, slide it out of the way and using MDI command a Z-3.0 move and see if you get zero on the indicator again. You can do the same thing in the X and Y axis, just have to clamp one block solid so it doesn't move. I like to use 123 blocks as they are usually accurate on size within probably less than .0002", but I also have gage blocks called "jo blocks" up to 6" long in a single piece and like to use those too.

Yes, Brian, you are correct. And I considered that but my eyeballing is pretty close for what I was doing. I mainly wanted the DRO to read correctly even though I showed the dial indicator. If that was off 0.003" then that just gives me a good visual because it isn't going to be out that much just because it isn't parallel or perpendicular to the traveling piece. And alas, I don't have any 123 or Jo blocks so I worked with what I had. I'm keeping this info you've provided for future reference, though. It's been many, many years since I've been involved in machine shop and tool & die work but I remember using those all the time.

Are you sure there is no back lash in the gear box

Hey Jerome - I doubt there's much in the way of backlash, at least not at this point since the machine is brand new. See below -

++++++++++++++++++++++++++++++++++++

Pursuant to all this calibration -

After all the calibrations for accuracy I put the machine to a field test for about 10 minutes. I used simple G-code to feed at 300 ipm and then to various positions all over the machine. About 90% of the time it was spot on the number I put in the parameters and the other 10% it ended up on xx.9999" and I don't think I'll get it any better than that. I also picked up some reusable tie straps from HF so I could clean up the cabling yet still have the ability to redo something if necessary.

Machine calibration testing -

[Jerome Stanek]

with the machine on and your gauge in place try puling and pushing and see if there is any movement.

[David Falkner]

with the machine on and your gauge in place try puling and pushing and see if there is any movement.

LOL! You want there to be backlash, dontcha!!

I'm finished with this step, put the dial indicator and base up, and am moving on. Before I got the technique down on calibrating, when it was always going to 0.0002 and -0.0002 there was never a point where I suspected backlash was coming into play.

[David Falkner]

After calibrating I thought it would be good to run some very simple patterns in the X and Y directions, no cutting so Z isn't involved. There's not much to this but it does show me that the machine is smooth and does what I asked it to do, at least cutting air anyway. This is a short video, about 42 seconds, so you won't be bored very long.

[Bruce Page]

David, you should look into making or buying one of these marking tools to play with. It is really useful (and fun) to have on hand.

Cosmos does nice work. The fit & finish is excellent.

http://cosmos-industrial.com

[David Falkner]

David, you should look into making or buying one of these marking tools to play with. It is really useful (and fun) to have on hand.

Cosmos does nice work. The fit & finish is excellent.
http://cosmos-industrial.com/index.p...=index&cPath=3

Thanks, Bruce! I had told my wife about these but haven't started looking where I could get one. I know at some point I'll get or buy a drag knife, as well. Which pen(s) do you have?