[Elkhunt]

I am interested in automating a jig we are currently using to build a 10' steel assemblies. The piece has a different height on all 4 corners and each piece is different based on info from the field. I want to be able to use a 4 axis motor controller, and essentially have 4 "Z" axis where Zero would be level with the table and I need to be able to move 5" above or 5" Below the table for any given piece. All of our dimensions are in Excel. Each mark number is on it's on line so Line one is P1, line 2 is P2 and so on. On each line there are 4 cells that contain the info for each corner. So Cell A1,B1,C1,D1 All pertain to a corner on Pc mark #1 Cell A2,B2,C2,D2 ALL Pertain to a corner on Pc Mark#2 and so on. We have a screen that shows all the info in the shop and each corner has an adjustable plate (scissor jack) that we manually turn and measure up or down from "Zero". I'm wanting to automate the jig by pulling the info out of excel and sending it thru something like a Geko G540 Controller and out to each of the 4 corners of the jig with a Servo and sliding stage set up that can adjust my vertical dim on all 4 corners. I have the hardware all figured out but I need help with getting the info out of excel and out to the servo motors so it will put the bearing plate where we need it to be. Any Suggestions or can you point me in the right direction for info? What would be awesome is if we could use the existing excel display an when we hit the arrow button to go to the next part the jig would set the four corners shown on the screen we could load it and go. Current pics of one of the corners attached. In my mind this doesn't seem difficult but i'm having a hell of a time finding any helpful info searching google......

[Jim Dawson]

The Gecko 540 is fine, as is Excel. The problem is that the Gecko only understands electrical pulses and a direction signal, and Excel only outputs math. Somewhere in between you need some software that converts that math into pulses. So let's say you want to move P1 from 0 to + 1 inch and the resolution is 1000 pulses / inch then you would need to send 1000 pulses to the P1 motor driver.

Mach3 software would generate the pulses and direction signal when connected to the Gecko, but it would have no idea what to do with the math from Excel. Mach3 only understands G code. Now Excel could convert the math into G code and output that to a file using the macro function, but you would still have to manually select the file to load into Mach3. Very close to what you want, but not fully automated. To automate the last piece would require some custom software.

It might be possible to write an Excel macro that accesses the printer port directly and could be made to turn the lines on and off. Not sure if it's possible to control the printer port using VBA.

[kawazuki]

Interesting puzzle ! If i'm understanding correctly the Excel file is holding height dimensions for each screw jack - 4 of - #1, #2, #3 & #4. If these were renumbered #x, #y, #z & #a, each screw could be commanded via a stepper in normal G code couldn't they? ie. G0 x100 y105 z105 a 100? Obviously you'd have to establish the accuracy of the jack screws - I can't imagine accuracy across their range is high on their manufacturing tolerances or you could look into closed loop control and add a linear scale to each corner point to assess commanded height against datum - drive motors would need encoders obvs so steppers / servo's equally applicable?

But to my mind this seems a very overly complicated way to achieve something that takes mere seconds done manually - how many times is it adjusted every day and what level of precision are you working to? If say, 0.5mm, surely a manually adjusted height setter is just as quick especially if you had to check an automated machine that could suffer knocks and bangs during fabrication work?

[Elkhunt]

So just to clarify a few things, We are not going to automate the jacks, We are using Nema 23 Stepper motors with a linear guide rail that has a 12" travel. The Motor will be turning a C7 1605 Ballscrew. It takes 200 Steps to complete 1 revolution and there is 5mm of travel for each revolution and I only need accuracy to 1/16" so this set up should be good for that. I understand the point about manually measuring taking mere seconds but this job we have coming up in march has 6000 pcs that need to be fabricated. Lets say each Corner takes roughly 5 seconds to set up and walk to the next one and repeat. Thats 33 hours spent just setting the heights on the jig for this one project. I figure since we already have the data at our fingertips it would be nice to automate as much as possible. I'm just stuck on the software side of things. I don't have tool paths per se like in a normal CNC part. I just need to Click the Part Number, have the 4 jigs set the heights, load the jig, weld, remove from Jig, Repeat. It would be nice as i'm switching our monitor in the shop to the next part number if the jig would switch at the same time and we could go straight to loading material. Right now the monitor is linked to an excel data base, you it the arrow up or down key and it moves to the next part number and displays the dimn. required at all 4 corners. I don't mind moving the data from Excel to another program, but I don't want to have to load a new file for each piece. I'd like to be able to arrow up or down and go to the next or previous setup with whatever software I use. I Have reached out to a few automation companies but I guess this is such a small project it doesn't seem like they are very interested in looking at it. I don't think what I'm trying to do is impossible but maybe just not feasible.

[Jim Dawson]

........... I don't mind moving the data from Excel to another program, but I don't want to have to load a new file for each piece. I'd like to be able to arrow up or down and go to the next or previous setup with whatever software I use. I Have reached out to a few automation companies but I guess this is such a small project it doesn't seem like they are very interested in looking at it. I don't think what I'm trying to do is impossible but maybe just not feasible.

You could export the raw Excel file as a .csv (comma delimited) or .xml file, and read that with a custom piece of software that would drive the steppers. The functionality that you want is not that difficult, but is a bit time consuming to write. I'm retired now but, when I was doing industrial software development my shop rate was $125/hr and I'm sure the vendors that you have contacted are in the same range. I'm guessing about 40 hours to develop what you want.

As I stated above, it is possible to write a macro in Excel that would output G code directly which could then be imported into Mach3. This is something you could do yourself, or at least in-house if you have someone who is good with Excel.

The output format would look something like:
(Part #)
G01 Xnn.nnn Ynn.nnn Znn.nnn Ann.nnn
M01

The M01 would cause the program to halt after setting the positions, and require the operator to press start again to move to the next setup. I should note here that Mach3 is not my prefered platform, but rather something that you could do yourself without a long learning curve.

It really depends on your budget for this project as to what way you go.

[kawazuki]

Hmmmm - you may have a production run of 6000 pieces but that is potentially 6000 individual parts so set-up and fabrication should ? be costed on a per part basis and would include that potential 33.5 hrs 'wasted' labour? As said a potential solution with code could cost you 40hrs @ £125/hr = £5000 + the machinery / infrastructure to implement it? I understand the desire for automation as a 'one button' solution but its the time and money that's not seen beforehand to get to that stage that is a 'make or break' solution.