so i got my gecko, but didnt have anything else ready so i started to hack together a control system.
first i tried to use it on my workstation... mach didnt run on xp64.
installed xp32... realized.. my workstation has no paralell port.
bought a pci parallel card... wouldnt boot - dead card.
store was now closed so i found an old pentium 4 system (mobo cpu ram) and placed it on a table, hooked up a hard drive, cdrom, and power supply and went to instal windows....
no go.. .my apple monitors are digital only, video on the motherboard was analogue.
soooo, i pulled out an lcd glass that was for a special project, along with its controller board. i taped the glass to a box so it would not fall over and hooked it to the pc.
voila! worked. finished installing windows and mach.
ran the mach driver test and at all speeds up to 100k got a "system excellent" rating. i used the g540 xml file from gecko to make a preset.
next issue was the gecko psu. i didnt have one yet, noone in town had a 36/48v psu in stock. so, i took an old notebook 19v 5A psu - the lower limit on the gecko - plugged it into the gecko and got the red fault light. so far so good.
i was going step by step in the instruction, and along the way i disconected the gecko from the psu. when i reconnected it no light up. after a short while i realized i crossed the wires and blew the fuse. didnt have a spare fuse so i just shorted the fuse socket.
wheee! green light!
hooked up the x axis motor on the mill and whadya know... it moved!
so, after playing around in mach i managed to get the axes going on 19v at 85 ipm with no lost steps and good reasonable acceleration. this is not bad at all. it should go over 100ipm with the correct psu easily as the motors are rated for twice the rpm they are moving right now. this will be a very decent rapid traverse speed. acceleration on the machine is ok but not spectacular. i assume the new psu wont help here, as its purely a function of low end motor torque fighting the table friction. more powerful motors are probably a better choice - but i have somethign else planned instead for later.
next step is to make up propper cabling to the mill and get the spindle VFD running. i also need to put the pc and gecko and psu into an old atx chassis.
i did notice one thing in my axis testing, but im not sure what it means yet. the machine is repeatable to better than .0005" which i tested with good digital calipers taped to the table. but there is a lead discrepency between what mach says and the caliper says. its something in the range of .001" to .01" at various points. i dont know if this is lead error, or if my screw isnt precisely 4mm per turn or something else. i do know it isnt slop or backlash, its a 100% repeatable offset.
anyone know if this is normal for screws to be off the spec'd pitch? do i just need to use mach screw mapping to compensate?
in any case, im having fun and am surprised how relatively easy it is to set up a gecko g540 in mach. this should all be working by the weekend ready to cut some test pieces of wood n stuff.  |