My system is based from a 4 foot, by 4 foot Techno/ISEL gantry router table. with the use of DIY methods, i am working to bring it to life!
2020-11-20 03:18) currently exploring changing the PC, and moving past the parallel port control to a Mesa I/O card for control. the PC has a i3 motherboard, but it also has two PCI-E slots, so i can upgrade to more modern hardware, or use the two PCI slots as an alternative, depending on what Mesa card i use.
also, i want to move the control hardware from prototype stage, to a more compact, and cleaner layout that is inside some sort of protective cabinet. probably even mounted on the machine, or rather, under it. it will need to be able to swing out of the way for machine servicing though.
i also want to mount the VFD on the machine, someplace. it can be mounted onto the outside of the control cabinet, but the control cabinet will need to be EMF/EMI shielded. a simple Faraday cage will work.
i also want to simplify the electrical connection for the machine. just simple breaker setup should work, with a single 240Vac cord to power everything, except the control PC, though provisions could be made for the PC as well. the whole idea here, is to reduce, or even eliminate the need for a full desk, where space for a monitor and keyboard/mouse are all that is needed. ideally, a single cabinet for all would work, if i can filter the incoming air to reduce dust contamination.
2020-11-29 05:19) regarding the control computer change, i am currently working through some software and hardware issues, getting the new LinuxCNC 2.8 Buster installed.
i have a Mesa 5i20 card now! so when i get the control computer sorted out, i can start working on the interface between my Mesa 5i20 and the machine.
an idea i have for the control cabinet is to use a currently available DIY enclosed server rack i built a while back, as the control cabinet to house everything. it will give me the option to make everything modular, and use a sliding shelf system to mount each set of components on. so the stepper drivers and power-supplies can be on a shelf, the interface system can be on a shelf, the VFD for the spindle can be on a shelf, and the control computer can be on a shelf. that way i can pull out each shelf as needed to perform maintenance or upgrades.
2020-12-01 02:38) progress is being made on the control computer upgrade. i have it loaded with Debian Buster (10), with LinuxCNC 2.9. i have also installed a used Mesa 5i20 I/O card. this upgrade should improve machine speed, and leave plenty of room for future upgrades/add-ons, with the 72 pins of I/O. it also is capable of on-board stepper signal generation through the FPGA, so it removes the stepper signal generation from the computer software, as in the earlier method of using the parallel port, and places the load on PCI card hardware. at least that is how i understand it.
2020-12-02 03:08) after finding out my used Mesa 5i20 is quite old, i have purchaced a Mesa 7i80HD-16. same basic port connections, but interfaces over ethernet. this means i can move the card out of the computer case, and will make for a much cleaner looking install. also, if i upgrade computers, there is no PCI slot requirement.
2020-12-13 14:16) big progress was made this weekend. i have LinuxCNC talking to the Mesa 7i80HD-16, and i also got it to talk to my LHB04 USB pendant. i still have to build up the protection circuits before i interface with the machine hardware. just a bunch of opto-isolators, and resistors. i will post schematics when i have software that draws them, again (some sort of screw-up with my Eagle license).
2020-12-17 05:03) i think this weekends priority will be to dismantle the current control system, to make way for the new system. i will have to make sure and label ALL wires, so there will be no issues when i connect the new system back up. when building the first system, i neglected to label some wires, and it causes some confusion.
second priority will be restoring the network out to the CNC shed.
i still have yet to build the opto-isolation circuits for the various connections. the stepper drivers have built in isolation, so i think they can be a direct connection. the relay boards, i think, have built in isolation, but i am not positive. this leaves the rest of the I/O that needs isolation. i have a bunch of FOD814 opto-isolaters, but i just need to solder then to boards with the support components. not very hard, just a PITA. i am on the fence, so to speak, on if i should add current limiting resistors between the Mesa card, and the input opto-isolators. the opto chips are open-collector, so an active signal will short the Mesa pin to ground. if for some reason the Mesa card is configured wrong, i don't want to short a misconfigured pin that is outputting. and if i do, what value would be best?
2021-01-09 16:52) today i made progress with the control system by first shielding the mounting board (3/4 plywood), and then mounting the Mesa 7i80HD-16, and breakout boards, as well as the 5v to 24V conversion boards. i shielded the ribbon cables between the Mesa board, and the breakout boards, and tied the shield to the mounting board shielding. i have not done any wiring yet, other than the ribbon cables.
The stepper drivers will be driven by 24Vdc, just like everything else, to reduce noise issues.
2021-02-23 04:21) just an update, and thinking out loud... more wiring for the power buses is needed, and i need to mount and wire in the stepper drivers, as well as the homing switches. I also have to tie in the other circuits, like for probing, Z-axis brake, and tool touch-off.
when the wiring is close to being done, then i need to finish the shielding around the control circuits.
i am still thinking of sepperating the power supplies from the main enclosure. the reasoning for this is to limit PSU interfearence inside the enclosure. it will likely create a large bundle of wires between the two enclosures, but it should be fine. i will also need to source out some sort of detachable connector, so the power supply and control enclosures can be moved independantly, when i move the machine from the camper to the yet-to-be-built CNC shed.
Note to self: i need to lube the machine again, before i make much movement with it, as it has been sitting for a while. it would be handy if i could make some lubrication ports on the gantry and Z-axis. something where i just need to remove a plug, add lube, and put the plug back in.
2021-03-05 16:58) Thinking out loud again... at the moment, my machine is a randon pile of wires. at some point i need to bundle the various cables together into a tidy group that i can just plug into the control box. here is a thought: other than the stepper motor cables, i could have a small box on the machine with a breakout board for possibly a DB-25 cable, and a matching breakout board in the control box. in theory, i could then run all the low-volt connections through that single cable, and be able to lessen the wire mess.
2021-03-20 06:37) i have the above mentioned DB25 cable installed, and the machine is in a working, but "not for production" state. in other words, the basic functions work, but it is not yet made SAFE.
I have ordered some DC circuit breakers for the 4 power busses (5v, 12v, 24v, 36v), as well as heavier wire to carry power from the power cabinet, to the control cabinet.
my plan for the power cabinet is to have a master shut-off switch, and some voltage/current monitors for each buss. inside, i will have the PSU for the steppers (36v), and the PSU for the 12v buss. the 5v and 24v busses are derived from the 12v buss, using some DC to DC converters. after the converters and other power supplies, just before power is routed to the control cabinet, i will place the circuit breakers. i may also have a volt/current meter for the incomming AC power, just for diagnostic use.
my hope for todays TODO list, i want to get started on the power cabinet. i also want to get the stepper cable connections mounted into the enclosure. should make things look better. i may also mount it to the wall, depending on what i get done.
2021-03-28 11:21) i am thinking that when i install the tool touch-off sensor, i will locate it at X0 Y0 on the machine. then when i attach a spoilboard, i will leave it up a bit to clear the sensor. then i will add my squaring fence. the idea is, i don't want to have to recalibrate the tool sensor proing NGC file whenever i replace or resurface the spoilboard.
2021-05-28 13:46) not much progress lately as the CNC shed temperature has been rather high. perhaps this weekend...
i have ordered a different cable for the DC power, as the 10AWG wire i had ordered was rather ridiculous in diameter, when you have 8 of them bundled together. the cable is just 4 pairs of 12AWG wire, and should be enough for my needs. if not, i should re-evaluate my needs...
so, i changed a bit of the configuration, in the HAL config. i didn't like using the coolant-mist and coolant-flood channels to trigger the vacuum to start, in case i use coolant in the future. so it is now setup, if i did it correctly, that if neither of the coolant channels are active, but the spindle IS active, then it starts the vacuum for dust collection. i need to finish the power enclosure to verify this function.
now, on the provision of coolant... i was thinking if i make a tray for the bottom of the table, with sides as tall as i can get by with, i can make it water-tight, with the exception of the drain sump, most likely in front of the machine for clean-out. the spindle could then have brushes around it to condense the spray some, and direct it downward. this is just an idea, and i am not yet settled on if i need it yet or not. mostly for brass/aluminum milling, possibly steel, if the machine will do it.