ShopDroid [custom sized]

[zool]

Louie states: I should add that a buddy and I had a 15-year run when we would build computers with the solepurpose of beating each other's score in the latest version of 3DMark. Got too expensive!

SIGH!! kids these days!

[zool]

After three weeks of frustrations and lost production, and a burned out controller board, after doing more research on controller hardware and software, plus a suggestion from Louie Atienza, am focusing on this:

CNC Control Software | Flashcut CNC

Have spoken to Ron and Anne at FlashCut and have downloaded the demo version of the software and am going through setting it up on my work station.

I like the integrated hardware/software. The manual is well written and has a user's point of view.

Yes, it does cost more than a MACH and Rockcliff or MACH and Gecko set up, but how much does lost production cost, being from a CNC not setup or functioning, and not being able to work on g-code to cut dovetails and box joints, which was one of the main focuses of this build.

Senior management has raised the usual issues, but in the end, especially if there is a dollar sign associated with an activity, I suffer no matter what I do, so might as well have fun!

[louieatienza]

After three weeks of frustrations and lost production, and a burned out controller board, after doing more research on controller hardware and software, plus a suggestion from Louie Atienza, am focusing on this:

CNC Control Software | Flashcut CNC

Have spoken to Ron and Anne at FlashCut and have downloaded the demo version of the software and am going through setting it up on my work station.

I like the integrated hardware/software. The manual is well written and has a user's point of view.

Yes, it does cost more than a MACH and Rockcliff or MACH and Gecko set up, but how much does lost production cost, being from a CNC not setup or functioning, and not being able to work on g-code to cut dovetails and box joints, which was one of the main focuses of this build.

Senior management has raised the usual issues, but in the end, especially if there is a dollar sign associated with an activity, I suffer no matter what I do, so might as well have fun!

Buy now, suffer later... NICE!

[jcoats]

The biggest issue for me with EMC2 was learning to zero a job where I wanted it consistently. I never felt like I had it figured out in spite of all the reading I did. Other than that it worked fine. I did finally find the dual motor on the X axis info and got it working successfully. It isn't very obvious how to do it.

In Mach3 Loader I copied a previous setup and created another setup based on it, and it didn't work correctly. Major racking with the cloned setup. Now I always start a new setup each and every time I need to make a new one.

CarveOne

Please take a few minutes and update the EMC2 Wiki (or somewhere!) whatever you found that is confusing or obfuscated! ... I'm following not to far behind!

[zool]

Louie Atenzia observed: Buy now, suffer later... NICE!


Based on my statement: "Senior management has raised the usual issues, but in the end, especially if there is a dollar sign associated with an activity, I suffer no matter what I do, so might as well have fun!"

Louie, just trying to forewarn you about the hazard that lies before you.

[CarveOne]

Please take a few minutes and update the EMC2 Wiki (or somewhere!) whatever you found that is confusing or obfuscated! ... I'm following not to far behind!

Setting the XYZ zero point in EMC2 and using offsets was not as easy to do (for me) as in Mach3. Other people don't seem to have the problem so I assumed that it is just me.

I have EMC2/Ubuntu 8.04 installation on my Linux CNC computer. The later releases may be a bit different. I used the XYZ stepper configuration in StepConf Wizard, not the XYZA stepper configuration.

Here are the two pertinent Stepconf Wizard screens that matter. You can assign the pport pin pairs to any axis you wish to, but what matters is that you set up the axes as shown here. There really are two X axis motors. The second X axis motor Step and Dir pins are connected to the A axis (slave) motor. Confused yet?

CarveOne

[louieatienza]

Louie Atenzia observed: Buy now, suffer later... NICE!


Based on my statement: "Senior management has raised the usual issues, but in the end, especially if there is a dollar sign associated with an activity, I suffer no matter what I do, so might as well have fun!"

Louie, just trying to forewarn you about the hazard that lies before you.

That's why I'm building all my toys NOW... before havoc arises...

[CarveOne]

Please take a few minutes and update the EMC2 Wiki (or somewhere!) whatever you found that is confusing or obfuscated! ... I'm following not to far behind!

Maybe I'll try to do that Wiki update after I have my Steel Channel Rebuild machine running. At that time I'll have the latest version of EMC2/Ubuntu installed on the computer as dual boot with WinXP Pro and Mach3.

If you are close to running your machine, download the latest EMC2 release from linuxcnc.org and see if there are any changes to Stepconf Wizard that make slaving easier. If not, try these settings.

One thing that was not obvious is that starting a new XYZ configuration will show pins for four axes, but after setting it up for three axes and saving the configuration, it will only show pins for three axes when you look at it again. So, don't expect to modify an existing XYZ configuration to add slaved motors. Start with a new configuration.

Meanwhile, you have something to work with. Looks like zool is taking another path now instead of EMC2.

CarveOne

[zool]

Looks like zool is taking another path now instead of EMC2.

CarveOne

C1:

I am taking another path with this machine as I have to get production going and have exceeded my frustration quota with the build and software/electronics setup.

And, as you will note the "this machine", I am already planning the next machine that will need eight axes of control which means EMC2. [R&P drive with two Z-axes, one with B/C head, and a rotary axis.] This is down the road a ways though .. one does need dreams, or will this be a nightmare.

And yes I am a glutton for punishment as well as a recidivist CNC builder.

[CarveOne]

C1:

I am taking another path with this machine as I have to get production going and have exceeded my frustration quota with the build and software/electronics setup.

And, as you will note the "this machine", I am already planning the next machine that will need eight axes of control which means EMC2. [R&P drive with two Z-axes, one with B/C head, and a rotary axis.] This is down the road a ways though .. one does need dreams, or will this be a nightmare.

And yes I am a glutton for punishment as well as a recidivist CNC builder.

Hmmm, didn't know EMC2 could do more than six axes. I have similar glutton and recidivist traits as well. Apparently it's not deeply embedded enough for eight axes though. My mind can't rotate and move in that many directions at once without massive lost steps.

CarveOne

[louieatienza]

C1:

I am taking another path with this machine as I have to get production going and have exceeded my frustration quota with the build and software/electronics setup.

And, as you will note the "this machine", I am already planning the next machine that will need eight axes of control which means EMC2. [R&P drive with two Z-axes, one with B/C head, and a rotary axis.] This is down the road a ways though .. one does need dreams, or will this be a nightmare.

And yes I am a glutton for punishment as well as a recidivist CNC builder.

Galil controller/software can handle 8 axes. Be prepared to pay...

[NIC 77]

NIC 77 stated:


I'm about to start a 4 x 8 cutting area shopdroids build myself. It's going to be for Routing and Plasma. The one weakness, in my opinion, of using the shopdroids design for larger builds is the lack of stiffness. An aluminum T-slot extrusion 2 inches by 2 inches could have alot of flex over a 4 foot span.

NIC 77, I am builting a Shop Droid for routing that is 83-inches by 53-inches. I use 3030 t-slot from 8020. http://www.cnczone.com/forums/cnc_wo..._sized%5D.html

I did moment of inertia calculations on the 83-inch length and with the use of a leg in the middle got the deflection to 3/10,000-inch assuming a load of 500-pounds evenly distributed, and support only at both ends and in the middle, and this did not take into consideration any of the cross beams which reduces deflection even more.

Though I do not have the statistics for the material you are going to use, my suggestion to reduce deflection to next to nothing would be to put a leg in the middle of both the sides. The deflection over two-feet will be almost nil. The deflection over the 8-foot side will be almost nil as well.

Zool,

I've posted my response in your build thread because it's probably more appropriate than posting it in Metalhead's thread.

I've been following your build. I've actually made a post on your build thread but it was with regards to your gantry orientation. If I look at your pic on post #27 of your thread I can see your design quite clearly.

It appears your calculations are for the 3030 T-slot with a support in the center. Hence if you were to evenly distribute 500 lbs over your table area you wouldn't have any noticeable deflection as you have indicated. I don't disagree with you at all on this point.

Unfortunately, this has nothing at all to do with my comment. My comment has to to with the shopdroids gantry system which uses the Pacific Bearing IVTAAG Rail which has a dimension of roughly 2 inches by 2 inches.

In your design this 83 inch length is unsupported. Using the 8020 deflection calculator (free program from 8020) and a 2020 extrusion (2 inch by 2 inch) to approximate the IVTAAG rail we get the following results:

Assuming the forces from your gantry act as a point on each top rail, and assuming a gantry weight of 50 lbs (25lbs on each rail) and that the rail is supported on both sides (this is worse than being bolted to something on both sides as yours is but it's what we have to work with in the estimation software) we get:

a deflection of 0.066 inches or 1.7 mm when the gantry is in the center of the rail. That's huge! It has nothing to do with the stiffness of your table. It has to do with the stiffness of the rails that are supporting the carriages.

That's assuming the biggest problems will come from deflection due to gantry weight. I suspect the biggest problems will come from vibration. You are building a router not a plasma machine. To counteract vibration problems you need stiffness.

Is there a light at the end of the tunnel?

Yes, I think so. If I were you I would add some additional steel supports along the length of the table attaching to both the T-slot from the 8020 and the T-slot from the IVTAAG rail, that would stiffen it up a whole lot. Additionally, you could bolt some steel stock to the IVTAAG rail used for the gantry to stiffen that as well.

What I'm saying is there are probably a few cost effective ways to stiffen things up without redesigning the whole thing.

Do you see my point?

[zool]

Zool,

I've posted my response in your build thread because it's probably more appropriate than posting it in Metalhead's thread.

I've been following your build. I've actually made a post on your build thread but it was with regards to your gantry orientation. If I look at your pic on post #27 of your thread I can see your design quite clearly.

It appears your calculations are for the 3030 T-slot with a support in the center. Hence if you were to evenly distribute 500 lbs over your table area you wouldn't have any noticeable deflection as you have indicated. I don't disagree with you at all on this point.

Unfortunately, this has nothing at all to do with my comment. My comment has to to with the shopdroids gantry system which uses the Pacific Bearing IVTAAG Rail which has a dimension of roughly 2 inches by 2 inches.

I disavow any response made after 12 midnight!

In your design this 83 inch length is unsupported. Using the 8020 deflection calculator (free program from 8020) and a 2020 extrusion (2 inch by 2 inch) to approximate the IVTAAG rail we get the following results:

8020 has a calculator??? I crunched all those numbers with paper, pencil and my calculator!!!

Assuming the forces from your gantry act as a point on each top rail, and assuming a gantry weight of 50 lbs (25lbs on each rail) and that the rail is supported on both sides (this is worse than being bolted to something on both sides as yours is but it's what we have to work with in the estimation software) we get:

a deflection of 0.066 inches or 1.7 mm when the gantry is in the center of the rail. That's huge! It has nothing to do with the stiffness of your table. It has to do with the stiffness of the rails that are supporting the carriages.

That's assuming the biggest problems will come from deflection due to gantry weight. I suspect the biggest problems will come from vibration. You are building a router not a plasma machine. To counteract vibration problems you need stiffness.

Is there a light at the end of the tunnel?

Yes, I think so. If I were you I would add some additional steel supports along the length of the table attaching to both the T-slot from the 8020 and the T-slot from the IVTAAG rail, that would stiffen it up a whole lot. Additionally, you could bolt some steel stock to the IVTAAG rail used for the gantry to stiffen that as well.

What I'm saying is there are probably a few cost effective ways to stiffen things up without redesigning the whole thing.

Do you see my point?

Yes, I see your point and you are correct. And the same thought is on my mind as well and I have the same solution in mind. Probably will ask John to make them for me as he has all the measurements and the plasma cutter.

I had not even thought of the gantry rail .. good thing it does not bite!

At the moment these are not on top of the list as trying to get the software and electronics to play together, which as you can see, I have been less than successful with.

On Monday am sending both controller boards back to Rockcliff to have them checked out.

This leaves me with only finishing a report, cleaning up my studio and addressing the Honey Do list!!!!

Life can be so trying.

Thanks for your concern and insights.

[zool]

Hmmm, didn't know EMC2 could do more than six axes. I have similar glutton and recidivist traits as well. Apparently it's not deeply embedded enough for eight axes though. My mind can't rotate and move in that many directions at once without massive lost steps.

CarveOne

C1: Ah, don't worry about them lost steps, they were not important anyway.

LinuxCNC.org - Basic Configuration

2.2.9 [AXIS_] Section

The [AXIS_0], [AXIS_1], etc. sections contains general parameters for the individual components in the axis control module. The axis section names begin numbering at 0, and run through the number of axes specified in the [TRAJ] AXES entry minus 1.

AXIS_0 = X
AXIS_1 = Y
AXIS_2 = Z
AXIS_3 = A
AXIS_4 = B
AXIS_5 = C
AXIS_6 = U
AXIS_7 = V
AXIS_8 = W

[zool]

Galil controller/software can handle 8 axes. Be prepared to pay...

Galil DMC-1886 PCI 8-axis Accelera $3195

GalilTools
Version 1.5.0

GalilTools is Galil's newest software suite and is recommended for all supported Galil controllers.

The free version contains a Program Editor Tool, a Terminal Tool and a Watch Tool.
The FULL version ($195) also includes a real-time scope and servo tuning capabilities.


What's a few grand amount friends!

[ger21]

I don't believe that GalilTools can control your machine. You'll need to either write your own software, or spend several thousand dollars more for something like Camsoft, which starts at about $3K for a 3 axis system. I can only imagine what they charge for 8 axis.

You can always use TurboCNC for $60.

[zool]

I don't believe that GalilTools can control your machine. You'll need to either write your own software, or spend several thousand dollars more for something like Camsoft, which starts at about $3K for a 3 axis system. I can only imagine what they charge for 8 axis.

Well, OK. All I can say is that there is a limit to my friendship!

Writing software is not something I even want to do, I DESIGN stuff!!

As for paying Camsoft for 8-axis software ..HA!! It was traumatic enough to convince senior management that the FlashCut CNC controller hardware/software was necessary [and its my own money not her's!], but to try and convince her that I should pay that much for something she could not touch...FERGIT IT!!!

[louieatienza]

Well, OK. All I can say is that there is a limit to my friendship!

Writing software is not something I even want to do, I DESIGN stuff!!

As for paying Camsoft for 8-axis software ..HA!! It was traumatic enough to convince senior management that the FlashCut CNC controller hardware/software was necessary [and its my own money not her's!], but to try and convince her that I should pay that much for something she could not touch...FERGIT IT!!!

She can touch the CD ROM....

[zool]

She can touch the CD ROM....

Once again Loius, you've been a great help!

[robe_uk]

She can touch the CD ROM....

And make sure its a spare blank one, hide the real one lol