Optical endstops



  • So I have these endstops:
    http://www.ebay.co.uk/itm/191017996202?_trksid=p2057872.m2749.l2649&ssPageName=STRK%3AMEBIDX%3AIT

    They work on 3.3v and the duetwifi can read them open/triggered reliably.

    I'm new to using optical endstops, so for anyone else who is - make sure the flags for triggering them are black, and they block the entire slot in the opto-sensor.

    I was getting a lot less voltage range with the slightly smaller, red flags I originally had.



  • I have those too, and they don't work quite right for me. When one of them triggers, sometimes one or both of the others do too, when they shouldn't. I worked around it by duplicating the move commands in my homedelta.g file. I should probably have fixed them by using other resistors like dc42 suggested (in thread below), but my workaround has worked flawlessly, so I didn't bother. Maybe I should try making my triggers a bit larger, too.

    http://forums.reprap.org/read.php?416,673777



  • Reading your exchange on reprap, this is what mine were doing when I set them up this afternoon. I measured the voltage between signal and vcc it was 0.4v open and 1.8v closed. The led dimmed but not much.

    If I put something solid black blocking the entire slot in the endstop I get a solid 3.3v on signal wire. The led on the duet goes out totally. It seems plenty to get a clear read open or triggered.

    I had no idea these were analogue, I presumed they did some sort of filtering and outputted 0v or vcc.

    Definitely use big black flags.



  • I use aluminium tape to block them. I had never had a problem triggering an other one they are about 350 mm apart on my delta.

    http://www.thingiverse.com/thing:1790722



  • Thanks for that end stop flag item.
    I am building a larger delta and bought a set of the optical end stops before seeing all the messages recently about variable light blocking depending on the material being used to interrupt the sensor.



  • Anyone care to share details on getting these working because all i get is a red light that turns on and off when triggered but DuetWifi reports nothing changed



  • So you get a red light on the endstop? If wired correctly and when triggered you mean fully blocked by something dense and black, the endstop led on the duetwifi should change from off to on or on to off.



  • I wasnt getting a led on the duet, but the sensor was lit until blocked. Even the voltage on each connection was correct. That was until i realized it was only working when the connector was pushed upward by the tester. All figured out now thanks



  • Cool. I really like them for a delta I can't think of a more accurate endstop, although I might try piezo discs since I have some and an endstop controller board. But if Russ's sensor showdown is anything to go by then a good old microswitch might be plenty accurate enough.


  • administrators

    You don't need accurate endstops on a delta if you have a good Z probe and you auto calibrate at the start of each print. 3-factor auto calibration is sufficient if you just want to compensate for endstop switch jitter. I've even run my delta without endstop switches to prove the point.

    I agreed with Russ, microswitches are plenty accurate enough even without doing auto calibration if used correctly. The only counter example I know of is the Fisher delta printer, because in that design the carriages run past the endstop microswitches instead of running into them.



  • So I am pretty new to this whole gcode thing. I played around on some of the printers at the local makerspace and they all home on the build plate which drives me crazy (I like to be able to pluck any oozing before the print starts).

    I like how the stock davinci homed off to the side (also a requirement to home the z with the endstop if you dont want to smash the nozzle into the build plate)

    I have the endstops and regular homing working off to the side of the build plate (Z probe not installed yet)

    Now abviously I am working to have the IR z probe so the Z endstop will become redundant but I would still like to figure out out to calibrate with the z probe but have x and y be homed off to the side but still have the printing area defined so a print cannot go off the plate.

    What would be even slicker is if its possible to do this and have a macro the can change the print area for different build plates before z probe calibration (depending on material i like to switch it up)



  • Hi Whitewolf. Smashing the nozzle into the build plate is bad, sure but how about something that gently feels for the print bed?
    That's what the contact based piezo sensor can do.

    If you want to change print area, or swap a different build plate in, you can. It will probe another build plate quite easily and then print.



  • @Whitewolf:

    So I am pretty new to this whole gcode thing. I played around on some of the printers at the local makerspace and they all home on the build plate which drives me crazy (I like to be able to pluck any oozing before the print starts).

    I like how the stock davinci homed off to the side (also a requirement to home the z with the endstop if you dont want to smash the nozzle into the build plate)

    I have the endstops and regular homing working off to the side of the build plate (Z probe not installed yet)

    Now abviously I am working to have the IR z probe so the Z endstop will become redundant but I would still like to figure out out to calibrate with the z probe but have x and y be homed off to the side but still have the printing area defined so a print cannot go off the plate.

    What would be even slicker is if its possible to do this and have a macro the can change the print area for different build plates before z probe calibration (depending on material i like to switch it up)

    None of this is any problem. You can mix and match and have different end stops on different axes. It's normal to home X and Y at one extreme of the axis movement so X min and Y min but you an home pretty well anywhere as long as you tell the machine what the offset is. For example, you could have the switch 20mm off the left side of the bed and then put an offset in the homing file so the machine "knows" that when the switch triggers, X=0 is 20mm to the right of that point. Using a probe, it's also usual to home Z with the head in the centre of the bed so all you do is move the print head to that position (in X and Y) before you probe. If you want, and if your machine permits, you can also use the "Z probe" to home other axes. To do that, you need to arrange a tab of some sort that is a bit higher than the bed but off to one side. Endless possibilities…....all of which are catered for.

    For different build plates, the easiest way would be to have X0 and Y0 at the same point, then have the axes maxima (M208) defined in a macro. Come to think of it, I can't see any reason why you couldn't you couldn't have the axis minima defined as well. HTH
    Ian



  • So if I define x=0 y= on corner of build plate but home all sends it to X-20 and y-20

    then only that area is printable nothing in the negative area and if i change build plates i can have a macro that sets x-30 and y-30 as well as shrinking the area with min max.

    About how would this look in gcode?


  • administrators

    Yes X0 Y0 should be the corner of your usual build plate. The M208 command sets not only the travel limits but also the positions set when homing switches are triggered.

    When you switch temporarily to a smaller build plate I suggest you don't change the axis length definition. Just use a different M557 command to restrict the probing area.


Locked
 

Looks like your connection to Duet3D was lost, please wait while we try to reconnect.