[lcvette]

Just to show how flexible the ultra flex conduit from McMaster is I took this picture last night. This is a 10 foot piece of the 1" conduit I linked to before, with a little over a foot hanging over the edge of the table and one of the control panel pivots (small piece of steel, less than 1 lb) hanging on the end.

Attachment 336274

-Dan

Nice! That stuff looks like exactly what I'm looking for!
The panel looks great!! L looking forward to seeing more of your project come together. As for your 4th axis, have you considered a harmonic drive instead of using the belt drive and worm gear? Will probably be significantly more accurate in terms of eliminating backlash.

Chris

Sent from my QTAIR7 using Tapatalk

[engnerdan]

Nice! That stuff looks like exactly what I'm looking for!
The panel looks great!! L looking forward to seeing more of your project come together. As for your 4th axis, have you considered a harmonic drive instead of using the belt drive and worm gear? Will probably be significantly more accurate in terms of eliminating backlash.

Chris

Sent from my QTAIR7 using Tapatalk

I will have some more updates soon, I am plowing forward now with a hope to have all the machining done by this weekend and maybe start mounting the electronics in the enclosure this weekend.

As for the 4th axis I considered all the possible ways to build one, but I went with the 8" Rutland Super Spacer because I picked it up with a 8" chuck in very good condition for under $400. It has a 2.5" through bore and a high enough centerline that I don't need to make rider blocks. While the belt drive is not "ideal" I think the downfalls of it are minimal in this situation, the worm drive is a 90:1 so the most angular error I will see in the spindle of the 4th axis is going to be 1/90th of the error caused by the belt drive with a 1:1 ratio or 1/180th with a 2:1 drive ratio.

I will get more into the 4th axis in a separate thread.

-Dan

[CitizenOfDreams]

Many of the interfaces shown in the 2nd post are "good examples of bad examples". The Haas control panel particularly makes me cringe, with its 150 or so buttons arranged in perfect geometric formation by an OCD designer.


Attachment 311646Attachment 311648

[engnerdan]

A quick update

A couple weeks back I picked up some steel to make all the pivot mount parts. Since then I have tried to dive in an get the parts made.

Attachment 336404
Pivot pins machined, and the control panel side plates machined

Attachment 336406
Attachment 336408
Attachment 336410
Making holes in the enclosure for the control panel

Attachment 336412
Test fit of the upper pivot - I changed my mind at the last minute and moved the conduit entry to the top of the control panel instead of the bottom as shown in all the previous 3D models.

Attachment 336414
Attachment 336416
Attachment 336418
The pivot pins and the conduit end

Attachment 336420
Pivot pins welded to the plates. Now I just need to drill and tap a couple holes in the pivot pins for the lock levers.

-Dan

[engnerdan]

Another step closer, both of the pivots/hinges are done.

Attachment 336618

Now I just need to get the "spine" that holds both pivots to the machine made and I can hang the enclosure.

-Dan

[lcvette]

Another step closer, both of the pivots/hinges are done.

Attachment 336618

Now I just need to get the "spine" that holds both pivots to the machine made and I can hang the enclosure.

-Dan

Looking good!! I am taking a simplistic approach, I want to use the touch screen as much as possible but have critical buttons available for quick access. So E stop of course will be easily accessible, and I will use the pendant to double as the jogging keys and fit right on the panel face with a little recess to be flush with the main control face.. do you have any key wants for your buttons? Which ones to use etc?

Chris

Sent from my QTAIR7 using Tapatalk

[engnerdan]

Looking good!! I am taking a simplistic approach, I want to use the touch screen as much as possible but have critical buttons available for quick access. So E stop of course will be easily accessible, and I will use the pendant to double as the jogging keys and fit right on the panel face with a little recess to be flush with the main control face.. do you have any key wants for your buttons? Which ones to use etc?

Chris

Sent from my QTAIR7 using Tapatalk

As of right now the only switches I know I am going to have are
Cycle Start
Feed Hold
E-Stop
CPU Power
CNC Power
Program Stop
Program Rewind (go to the beginning).

Switches I would like
Feed Rate Override
Rapid Override
Spindle Override
Tool length set (record to the tool table the length of a new tool)
Go to machine home (move Z up and bring the table front and center for part loading/unloading)
All Coolant off
Air Blast on
MQL On

Basically I think I will end up with most of the functions that are on the Mazak control panel below the keyboard.

I have over 50 other buttons modeled but I am not sure how many will actually get added to the panel and what they will be used for. Or how I will accommodate them yet. You mentioned using the touch screen for as much as you can, and I am trying that now but the buttons in KMotionCNC are a little small to hit at times. But they did just release a new screen editor so I can resize the buttons to be easier to use. The thing to keep in mind is there is no tactile feedback to a touchscreen button and I personally don't like using them for anything that could result in disaster if pressed accidentally. Its not hard to catch the wrong job button or accidentally have a stray finger contact on a button next to the one you intent to push (especially when rushing).

-Dan

[engnerdan]

I made a big step forward this weekend. I finished out the mounting bracket, short of painting it, and got the enclosure hung on the mill for the first time.

Attachment 336750

Attachment 336752

Next up is cutting the front panel, which I am hoping to do by this coming weekend and order some lever handles to replace the lock screws on the hinges.

-Dan

[engnerdan]

A little more progress, the 24" X 24" steel panel has proven to be the most difficult part to machine so far. The problem being its size and the enclosures on the my mill and the mill at work. Typically panels like this I would have done a hand layout and plasma cut the large holes. But I wanted to make everything perfect on this one.

Attachment 336842

-Dan

[engnerdan]

Getting even closer.

Attachment 337074



-Dan

[engnerdan]

A little more forward progress.

When I machined the front panel I was not able to get to the very edges of the panel due to the enclosure on the mill. So the additional screws I wanted to add around the perimeter didn't happen and the top row of holes for mounting the monitor did not get drilled. In order to add the monitor mounting holes I took a scrap piece of sheet metal and made a drilling guide. I drilled all the monitor holes in it and then screwed it to the enclosure panel with the three holes I was able to drill. Then used the guide to drill the top row of holes. Unfortunately I forgot to take a picture of that process.

I also had time to make two switch panel blanks that I can add holes to at a later time, when I determine what I want for switches.



I am hoping to get the monitor mounts made today so I can get that last piece mounted and start wiring the control panel.

-Dan

[engnerdan]

I made a little more progress at the end of last week. I got the monitor mounting bars machined. But when I went to mount the monitor I got a surprise, it did not fit. The factory spec for the monitor said 10.875" tall, I measured it as 10.940, so I adjusted the bars 0.065" (0.0325" each). But when I tried to bolt it on to the front panel it installs like it is 10.875" tall. I had to flip the top bar around and still had to shim with washers. Unfortunately I did not get photos of that yet.

Here is the monitor with the bars installed.
Attachment 338110
Attachment 338112

CAD Model of the monitor mount
Attachment 338114
Attachment 338116

-Dan

[lcvette]

Looking great man!! Keep up the awesome work!!

Sent from my QTAIR7 using Tapatalk

[engnerdan]

Progress is slow but steady.

I got everything mounted to the front panel and the initial wiring done so I could test it all out this last weekend.
Attachment 339654

Then I got the bug and dove into finishing out the umbilical cord that connects the control panel to the mill controls. Just a few wires right now, I am using most of the 48 pins right now but I expect that to change in the future. Right now there is USB, Ethernet, Encoder, Modbus 485, power and discrete I/O.
Attachment 339656

I was able to finish the control panel side of the umbilical cord last night and could not wait so I started landing the wires into the terminal blocks. The next steps will be to finish the wiring at the control panel, this includes wiring in the jog pendant. Then I need to dig out the pallet jack and move the mill out from the wall so I can get into the electrical cabinet and finish wiring in the other end of the umbilical cord.
Attachment 339658
Attachment 339660

-Dan

[CNCMAN172]

Looks like you are doing a very nice job on your build. Looks like you much be running a KFLOP as I can see you have KmotionCNC running in one of the photos.

Russ

[engnerdan]

Looks like you are doing a very nice job on your build. Looks like you much be running a KFLOP as I can see you have KmotionCNC running in one of the photos.

Russ

Yeah I have a KFLOP and KANALOG boards running my mill. I have also added a Automation Direct PLC for added IO and some basic functions that the KFLOP does not do, like ethernet and Modbus 485. I am excited to get in and play with the new screen editor that Dynomotion just made available, which could eliminate some of the functions I am using the C-More HMI for. I wish Dynomotion would add rigid tapping and part probing into their software.

-Dan

[m_c]

Dan, what happened to your tests for getting rigid tapping working via a G-code?

What kind of probing are you thinking about?
Basic (i.e. locating edges/bores etc) could be done via a C program.
I know I still run Mach 3 on my digitiser, but I've often wondered what would be needed to create the framework to expand the probing capabilities to something similar to the Probe-It plugin for Mach3. If somebody could create the basic framework, I'd quite like the challenge of writing the required probing algorithms, however graphical applications are most definitely not my strong point!

PS. I've been following this thread for a while, and the panel is looking good. It's certainly giving me plenty inspiration for my upcoming panel build.

[engnerdan]

Dan, what happened to your tests for getting rigid tapping working via a G-code?

What kind of probing are you thinking about?
Basic (i.e. locating edges/bores etc) could be done via a C program.
I know I still run Mach 3 on my digitiser, but I've often wondered what would be needed to create the framework to expand the probing capabilities to something similar to the Probe-It plugin for Mach3. If somebody could create the basic framework, I'd quite like the challenge of writing the required probing algorithms, however graphical applications are most definitely not my strong point!

PS. I've been following this thread for a while, and the panel is looking good. It's certainly giving me plenty inspiration for my upcoming panel build.

The honest true about the rigid tapping is I have too many irons in the fire to really have time to work on it and when Tom suggested running it as a G-Code sub-routine type program I became disinterested. All my g-code is generated with CAM software, I get no enjoyment from hand writing any of it, but I understand enough to tweak things if I must, but I don't have any desire to post out code and them go edit in special calls for tapping just to make it work for me. It frustrates me that Tom has not just done it himself. They have synchronized tapping for lathe use from what I understand but not the mill? I know that the Dynomotion product was not originally designed just for CNC machine operation, its a universal motion controller that many companies build into OEM machines but there is a huge user base just using it for CNC.

As for probing, I want raw stock WCS setting , single plane, stock corner, bore, boss, middle face, ect like the pictures below and tool length probing would be nice too. I have spent more than a few hours contemplating switching to LinuxCNC/Path Pilot so I would not have to do it all myself. I have a 2 year old Haas VF-2 at work and the probing is so simple that it feels like using a primitive machine when I use my machine at home. On the Haas I can setup a job, loading a half dozen tools and setting the WCS of the stock in 5-10 minutes. On my machine at home it takes like 30 minutes. I have a copy of all of the probing routines/macros from two different Haas machines if you want to see them. I think they are all done in G-Code. Unfortunately the guy I knew who could make sense of them passed away a few months back so I am not sure how to make sense of most of the code. I think they contain all the algorithms used but I think they might reference special built in memory locations (like persist values in the KFLOP).

Haas Part Probing Screen
Attachment 339782

Renishaw VQC (visual quick code) Screen 1
Attachment 339784

Renishaw VQC (visual quick code) Screen 2
Attachment 339786

Haas tool setting
Attachment 339788

Renishaw VQC (visual quick code) Tool setting
Attachment 339790

-Dan

[m_c]

I know all about having too many things on the go!
I'm currently in the middle of retrofitting a mill, so my time is a bit limited just now, but I'm certainly willing to put some time into probing routines.

Having re-read the tapping thread, my understanding was Tom was wanting you to test it via a M-code, to make sure it would work as you expected with the necessary parameters, and then if it worked as planned he'd add it to the main interpreter as a G-code. I suspect Tom's reluctance to just add it, is he doesn't have any machinery to test it on, so is wanting somebody to physically test it first. I would test it, but I'm not sure if my spindle would be up to it. It's rated 1HP at 4000rpm (1.3lbft/1.7Nm), but it's actually a 4.5Nm/4000RPM DC servo motor, so it may be able to provide a bit extra power for testing with a small tap.


I'd be interested in seeing the Hass probing routines, but all those routines you've shown are what I'd class as basic probing. They're basically probing certain features for location purposes, and not like perimeter probing where you need to rough guess where the next touch is going to be, and how to react if it's not.
The actual routines for them are reasonably simple, it's just getting suitable graphics in KMotionCNC that might be a problem.
I think the graphics might be possible with the new screen editor, but I've only had a quick play with the new screen editor, as I'm trying to avoid distractions and concentrate on getting the mill retrofitted first.

Tool probing is probably one of the easier things to implement. I've already got a couple tool setting C programs for my lathe, which involve a taking test cut, entering the measured value, at which point the program calculates the offset and updates the tool table. It would be easy to alter them to make use of a tool probe (which has just reminded I need to add an input for on my retro!)

[engnerdan]

I know all about having too many things on the go!
I'm currently in the middle of retrofitting a mill, so my time is a bit limited just now, but I'm certainly willing to put some time into probing routines.

Having re-read the tapping thread, my understanding was Tom was wanting you to test it via a M-code, to make sure it would work as you expected with the necessary parameters, and then if it worked as planned he'd add it to the main interpreter as a G-code. I suspect Tom's reluctance to just add it, is he doesn't have any machinery to test it on, so is wanting somebody to physically test it first. I would test it, but I'm not sure if my spindle would be up to it. It's rated 1HP at 4000rpm (1.3lbft/1.7Nm), but it's actually a 4.5Nm/4000RPM DC servo motor, so it may be able to provide a bit extra power for testing with a small tap.


I'd be interested in seeing the Hass probing routines, but all those routines you've shown are what I'd class as basic probing. They're basically probing certain features for location purposes, and not like perimeter probing where you need to rough guess where the next touch is going to be, and how to react if it's not.
The actual routines for them are reasonably simple, it's just getting suitable graphics in KMotionCNC that might be a problem.
I think the graphics might be possible with the new screen editor, but I've only had a quick play with the new screen editor, as I'm trying to avoid distractions and concentrate on getting the mill retrofitted first.

Tool probing is probably one of the easier things to implement. I've already got a couple tool setting C programs for my lathe, which involve a taking test cut, entering the measured value, at which point the program calculates the offset and updates the tool table. It would be easy to alter them to make use of a tool probe (which has just reminded I need to add an input for on my retro!)

You are right that Tom wanted me to test the code out using an M code and passing some variables like a sub-routine. Part of the problem is laziness on my part and part is I don't wish to hand write code for every test part. I may get motivated soon to give it another try. Now that I have the control panel in place and mostly functional I am finding my mill much easier and more enjoyable to use.

Attached are the probing routines I have. There are from 2 different machines, the one titles Haas #5 has a spreadsheet in it with some variables listed I didn't make this so I am assuming it was a record of the variables on the machine the macros were copied from. I think these would be great to use, being they are g-code based, already written and easy to add to or modify instead of c based code that has to have a thread to run in. But to make them work KMotion would need to call the correct macro and them there would need to be a variables file in the background that these files could reference and write to. For instance when you do a calibration the probe tip diameter offset is stored in the variables.

-Dan