By the way, I'm posting a lot of detail photos of the way I did things. I don't know that they are "best" or "perfect" or anything like that; I will say that I've solved multiple problems on multiple CNC machines and I've done the best to transfer some of the best practices that I've personally verified into this build.
Specifically, I wanted:
- "Star" ground of shields. This is grounding all shields to ONE central point. To be clear, this means the 'other' end of each shield, the one nearest the motor or switch or whatever, is NOT connected in any way.
- The "Star" ground IS connected to "EARTH" ground via the power cord wiring.
- The "Star" ground IS NOT connected to power supply ground (meaning zero volt point of the +_24V system).
- Every wire can be disconnected at the device end and the central end, thus allowing for maintenance, changes, upgrades, etc.
- Both reset and e-stop (which completely removes 24V power) are brought out front.
- TinyG is mounted in a protected location; LEDs are visible (thanks to some fiber optics)
- TinyG uses one USB port on the host computer.
- A powered hub for all cameras uses a separate USB port on the host computer.
A few other notes:
I did use a commercial anti-static mat and I did glue it to my table. This seems to be working just fine. Carefully following the spray glue instructions on how to "...achieve a permanent bond..." is crucial. This procedure is bit counter-intuitive in that you spray both surfaces and allow the glue to dry before pressing them together. If the glue is wet (at all) the bond will be temporary. I've seen other posts indicating that their mat wouldn't stay glued and I'm curious what process they used with the glue. Also, my table is "melamine", pretty much the perfect non-porous surface for this kind of glue.
I did substitute an "Anondstar" camera for the down look. I modified the bracket supplied in the kit by passing a 1/2 inch end-mill through the big hole. This makes the hole slightly too large for the Anondstar, so I also widened the tightening slot to 1/16 of an inch. These two mods allow the original bracket to hold the Anondstar just fine. I did discover that the moving USB wire to the Anondstar will move its internal mechanism, and therefore the camera view, no matter how firmly the exterior case is mounted. Therefore, I tied this cable down in a couple of places along the top part of the carriage assembly. A moving wire could easily be a difficult to find issue with this camera.
I am using the LitePlacer Store's drag-chain kit. The Anondstar camera's built-in USB cable happens to reach PERFECTLY through the first chain and is then plugged into a 1 meter extension that then enters the second chain. The point being that the connection is NOT in a chain and is NOT subject to motion. The extension is plugged into the powered camera hub on the bottom of the table.
I have not yet done the uplook camera; I am planning on using a C270 Camera in a printed housing.
I have done some playing with acceleration and so forth. So far the machine is quick and smooth. I'm running it a lot harder than the numbers in J's sample documents, and so far the camera seems to come back to the exact same spots. We'll see if this remains true when it begins placing. I took considerable care in squaring everything, and ensuring the two belts that drive the long cross-carriage drive it smoothly and squarely, and so forth.
More to come...
Danal
