hey guys,

i'm trying to calibrate my delta printer.. (geeetech g2s pro, with updated hotends)
and am slowly getting there i think.  but i'm not sure on one measurement.. namely the floow safety margin.

i understand what it is for but what i don't understand is what to measure exactly.

my understanding was.. from a) slider joint center point to b) top hotbed, when @) height 0.00.
but this gives me a measurement in the 200 plus.

which is compared to the other examples found online is extreme large. (between 4-10x as large)

hope some one can explain how to measure this size exactly.

would really appreciate that.

also is there o good tutorial online somewhere on how to calibrate for the repetier firmware?
although the website config tool is great. it isn't always very clear for the uninformed.

Kind regards



  • I thought it was so the tower won't go lower than that number. I think it's to protect the arm from going lower than horizontal.
  • @Roko if you set printer radius right it will never go below horizontal. You should even never go near horzontal as that motor has to move so fast in that reagion that you get speed problems.

    The safety margin was an addition from a delta user where something breaks if slider goes too log and hits the bed. So it just prevents that from happening. As I understand it is is ditance bed level to arm joint that you need at least to not crush slider into bed. 30-40mm should be safe for all printers I know.
  • edited November 2016
    @ ROKO and REPETIER: both thank you for the reply!

    it is still rather unclear how to measure it though.

    Repetier said:
    if you set printer radius right it will never go below horizontal. You should even never go near horizontal as that motor has to move so fast in that reagion that you get speed problems

    --> what would be an appropriate minimal angle with the hotbed to make sure it doesn't come close to horizontal?
    my hotbed measures 215 mm diameter. the inner edge of the screws used to mount the hotbed protrude 10mm. leaving 195mm. so i set the: Max. horizontal radius: DELTA_MAX_RADIUS to be. 90mm. at that position the angle of the arm to the bed is approx. 23 degrees. would this be oke?

    The safety margin was an addition from a delta user where something breaks if slider goes too log and hits the bed. So it just prevents that from happening.

    --> by the time the slider would hit the bed.. it would mean that the arm is indeed at least horizontal or even lower then.

    As I understand it is is ditance bed level to arm joint that you need at least to not crush slider into bed. 30-40mm should be safe for all printers I know.

    --> which arm joint do you mean? if you mean the lower one (the one mounting the spider) and with the nozzle's position at zero. it would be 57-58mm.

    do you think this would be the correct measurement? this would be very imprecise though.  better to go for 57 or 58 (looks like it would be more like 57.5) ??

  • two additional questions:

    1)  Z_probe_bed_distance: [mm] = Max. distance to go back for next test.

    i really don't understand what is requested here... specially if you take in account also question 2:
    the max distance to go back is back al the way to home. not necessairy but possible.
    the min distance to go back would be the answer in question 2, but it asks for max distance.

    so say home is at 184.3 and the trigger switches at 3.00. would this be 181.3? seems rather odd.

    2) Z_probe_switching distance: [mm] = Max. distance to go back for next test.

    i kinda understand this one. but the explanation and the name rather contradict themselves.
    say the probe triggers at 3.00 mm and untriggers at 4.5mm would this be 1.5 or rather 4.5 (relative or absolute distance)

    i guess i could take 4.5 but this gives an extra (unnecessary) move if only 1.5 is needed.

    kind regards


  • @spikey I just disabled steppers and moved the carriage down then I extended the nozzle assy  away from one tower to be directly opposite with arm horizontal or almost so (if you can go that far)and measured height of carriage bottom to the frame and I added 5mm.If it physically stops prior to that point I would add 5 to the measured height.  This should prevent the carriage from going lower than that number. If your config and hardware is such that your arm is extending to the horizontal position during normal operation you are pushing the envelope IMHO.
  • 1) That is height probing starts. So it homes, goes down to that height and starts probing. This value should include some saftey as it depends on current z length and if that is too high you will crash into bed.

    2) 1.5mm in your case. If is for repeated mesurement to have minimum movement to reduce measuring time. 
  • @ roko: i can move the arm till max 23 degree (0 degrees being flat) if i take the in account the 90mm radius. 
    so this is definitely not flat, but maybe it is already to close. 

    if i go for 0 degrees it really goes near the edge.. so like 105mm out. 
    at that moment the bottom of the slider is @ 55 mm from the frame (which is lower than the hotbed of course) so 
    for that height i should take like 60 mm? 

    i will set it like that for now, unless you let me know not to.. 

    @ Repetier: 
    i set Z_probe_switching distance: to 1.6mm. just to have that little bit extra buffer. 

    the Z_probe_bed_distance: i set to 20mm 

    but when i try the start the bed height mapping few things go wrong. 

    1) at the start (home) the carriage crashes into the end stops even thought the drop after the homing is set on 10 mm, plus there is the extra drop on the x and y pillar because of the selection of 1e extruder.  guessing approx. 5mm meaning it crashes into the end stop even though the 15mm height difference. 

    2) the first measurement seems to go perfect (goes up and down 3 times??? ) but afterwards it goes almost all the way back up. not just the 20mm

    3) 2e measurement goes oke too, but after the first touch on the 3rd measurement it give fatal failure, while nothing seems to go wrong physically.  

    kind regards Matthieu

  • "start bed height map" is misleading as I know 3 different occasions. Especially if you talk about host be dheight map you need to go to Z=10 or 20mm first and then start a sit will set xy first and that is not possible at top without crashing into endstops skewing everything.

    Only autoleveling/distortion map will go down after homing I guess. Always forget which printer type does what in which version.
  • @spikey1973  I think the main thing is that you don't want it  to be able  to go too low and stress/break anything but you would still want to be able to use your printable area without that setting stopping that function. I've got mine at 40 but I use printed carriages with bearings (R/K Delta model - home built). I set it once and don't worry about it. My carriages never come near to that number when printing. My arms don't have to "reach" to get to the limits of my bed, so I never get that low.
  • @ repetier: 

    i tried this indeed, did make the slider not crashing in the fatal error at the end occurs just the same. 

    hoe do i start the autolevelling / distortion map? 

    @ roko: thank you, i set it to 60 and i'll just leave it so then. thanks for the help, slowly getting there. 

    Kind regards 

  • I always start leveling commands at Z=10.

    At the end the z probe is disabled meaning extruder is activated. Depending on printer and position at end this might not be possible. What is the fatal error you get?
  • indeed feel like starting at z=10 would be a good idea. unfortunately it doesn't solve the basic problem.

    i've made some screen shots and will post the links. but i also filmed the action of the printer and put it on youtube. 
    both the printer as repetier host give following error: g30 probing failed! 

    hope this gives some additional info... 


    kind regards 


  • Yes, you would in deed reduce the probing time immensely if you start lower:-)

    I see you have set do x repetitions and the failure was after first move up. At that point it is expected that z probe is low which was not the case I think and then it aborted probing.

    Could have different reasons:
    1. Probe is not very reliable and got stuck.
    2. You need to go higher to turn signal off.
    3. False signal from crosstalk with motors/heaters. Are the probe cable and others nearby twisted to prevent cross talk?

    When it happens send M999 and M119 to see if probe is still high. If not chances for crosstalk increase.
  • oke.. the probing time really is not much of an issue for me.. i just want it to work.
    but personally i thought that "Z_PROBE_BED_DISTANCE" ([mm] = Max. distance to go back for next test)
    would make sure the spider wouldn't go all the way back up.. but if starting lower is the solution for that.. well easy enough.

    as far as i'm aware i don't have any repetitions set.. if so where would i need to find that? is there a benefit is repetitions?
    at the moment it measures three times. than goes all the way back up, comes down, measures 3 times goes back up and starts the 1 measure and crashes.

    as it gets stuck after 6 smooth measures, everytime at the same point.. i believe i can discard option one. also you can hear the end stop switch click, so also option 2. unfortunately it crashes at the 7 test..without ever going back-up. so testing afterwards has no use. leaving all but option 3.

    although i personally have my doubts about this.. but since i must admitt that the wires were, well at least less than neat arranged. i'll go for that untill disproven. i will fully rearrange the wires, make sure the endstop wires are twisted from the beginning (also i'm gonna upgrade the poor quality switches the printer came with) and get them as far away from the (extruder) motor wires as physically possible. pillar motors are at the bottom.. so not much interference there but also i will keep them away from each other as much as possible. also i will decrease the excess wire length at the bottom to get the neatest wire layout possible.

    unfortunately i will need to crimp the jst plugs used. for this i ordered the correct crimps in china.. so this will take a few weeks before i get them. so i will be quite on this question because of that. doesn't mean it is solved (jet), just on temporary hold.

    i didn't test m999 and m119.. as i already started to pull the wires when i remembered i still needed to do that, but i will do so if the issue remains after rewiring.

    kind regards

  • oke.. 

    Today i got the new endstop switches in and started on the rewiring. 
    for this i twisted each of the switch wires (x,y,z pilar and the probe switch) 
    i also braided the twisted x,y and z wires together as one bundle. 
    i plan to guide them seperately upwards to not get near the motor wires. 

    are there other wires that need to be twisted up? was thinking on the themistor wires and guide them to the spider together with the probe wire together. also separately from the heater wires. is this last part usefull / necessairy? 
    have i made any misassumptions int he above? 

    kind regards Matthieu. 
  • Twisting helps averaging influences to nearly zero. So thermocouples would also benefit and give better precision, but since we average readings over several reads it is not that important here.

    Since it always happens at same place I would shrink the testing square a bit to make sure it is not the position.

    As video shows repetition you will have them set like this:
    // Repetitions for motorized bed leveling

Sign In or Register to comment.