Hi All, I am building a plant phenotyping system for my laboratory (I do plant research). I have placed a link to a video of a system from Univeristy of Nebraska. Nebraska Plant Phenotyping: HTS Scanalyzer I do not need something this complex to begin with. This is in more in line with what I am looking at as I have a controlled room I am putting the system in. I have purchased an ACRO 1010 system with the BlackBox controller. I have it set up, calibrated, and ran the test logo program. Now I need to reprogram the system so I can take images of plants over time (7-10 days). The schedule will run on a 6 hour cycle. The cycle is (1) wake up, (2) move on X / Y axis till RGB camera over plant of interest, (3) capture plant image in RGB, (4) move until IR camera is over plant, (5) capture plant image in ITR, (6) move to 0, 0, 0, and (7) go to sleep until the next cycle. So here are several items I need some more information on. 1. Can I access the Blackbox board using the USB cable so that I can program the system as needed to perform the steps above? 2. Are there any things that I need to be aware of if I go poking into the system that might cause things to go poof (besides the usual stupid things I do to burnout boards)? 3. Are there any suggestions, comments, ideas on how to make this as painless as possible (give up is an answer, but not an option. I'm stubborn that way)? Thank you in advanced for all your help. Hector
You can yes, protocol is explained in detail here gnea/grbl Never move motors by hand. They turn into generators backfeeding the stepper drivers. Never do wiring work with power on. Check wiring twice, no shorts. Use good quality PSU. Google for "grbl" +coding question, blackbox runs Grbl, its widely used, you will find more results searching how to do what you want to do using grbl as search term
Measure/calculate/calibrate (if you have realtime feed) the IJ vector offset between the two cameras and use that as your XY coordinate system offset. If you use G54 for for VIS and G55 for IR, programming will be much easier because each plant sample will only require a single coordinate. G53 is your machine coordinates and is zeroed right at the limit switch actuation. You can move purely in machine coords but it's not considered a modal command- that is, if you say "G53 G0 X34.8 Y253.5", that will move to that location but the very next line you'll also have to start with G53 because the control will default back to G54 (or G55, or whatever modal coordinate system you were in). Going to G53 0,0,0 will alarm out the machine because it'll actuate a limit switch, of course. For consistently spaced rows and columns of incremental motion, you may want to look into G90/G91 absolute/relative coordinates. This can be dangerous, typically more so in a machining context, because normally almost all code is produced under the assumption that the control is in G90 absolute coordinate mode- "G0 X5" means "rapid to [G54] X5" NOT "move positive 5mm in X". As long as you're careful about where you're using it and how, it may make the loops you'll be programming to generate and send the G-code a little easier. All of this is discussed on the grbl wiki linked above and the LinuxCNC G-code table page (grbl uses the LinuxCNC G-code standard).
Hi @Peter Van Der Walt and @Rob Taylor, thanks for the response. I am taking a look at the link and suggestions provided above. Learned that the hard way. On previous builds. Only the best. Openbuilds ... Hmmm. Very nice observation. Yes, really will make it easier. In particular since position along one axis will be the same and I only have to do a translation move on the other axis. It would be easy to set up a matrix X[#VAR+n] for each offset. I will keep you posted on the progress. Hector
