Hello, I have blackbox and i install sfu 1204 on my workbee 1010 I don't know what is the new settings for my firmware So I dont understand what is "Thread start" so someone can explain how can i found steps/mm for sfu1204 or how i found thread start for this ? For me each thread is spaced by 4mm but after... thank you
Technically you don't need to know anything about the threads in order to calibrate the system, but it does help as a starting point . With power off you can just turn a leadscrew 1 turn with your fingers and measure how far the carriage moved. Fill that in here GRBL settings calculator under 'Z threaded rod pitch (mm):' The motor is most likely 200 steps per rev so leave that alone. Select '8 microsteps per step', (the standard on a Blackbox.) now in the result section at the bottom for Z you get a steps per mm value. In OpenbuildsCONTROL you can fill that number in for all relevant axes in the 'GRBL Settings tab', save and reset. Now use the 'wizards and tools|Calibrate' menu items to calibrate each axis to a final value. I have proven to myself that one can in fact calibrate an unknown system by just 'starting somewhere'. I took an old printer mechanism with an unknown motor and unknown diameter wire drive. I set the steps per mm to '1' and went through calibration 4 times to get to a final value. A bit tedious but it does work (-: (and one does have to be careful with test distances lest it run into an endstop) One might start with 1600 as that will give 1 turn of the motor on a stock Blackbox setup, (200 step per rev motor at 8x = 1600 steps per rev). Then giving G0 X1600 (assuming X) should turn the motor 1 turn. If it only turns 1/2 turn then either the motor is 400step per rev, or the microstepping is 16x. Either way you can now measure the movement and use the formula to correct the steps per mm value, do it several times to refine the value.
Yep, what David said. I'd just enter 400 (1600 step/rev, 4mm pitch) in the grbl settings and then tweak from there. It'll get you close enough to avoid crashing as you fine tune the value.
how you found 1600 ? normally it's 200 4 circuits x 2mm = 8 mm pitch so 200*8/8 = 200 step/mm I think sfu is 4circuits x 4mm 16mm pitch so 200*8/16 = 100 step/mm???
1204 = 12mm diameter, 4mm pitch. its a one-start screw. 1204 and 1605 is about the most popular sizes for hobbyists and are both single start. Multistart ballscrews are rare, and expensive. Consider the added complexity just manufacturing a recirculating ballnut with 2 or 4 channels (;
I was giving an example of an unknown system, and assuming a 200 step motor and 8x microstepping, 200x8 gives 1600 steps for 1 revolution of the motor. This gives us a starting point, we can drive the motor 1 revolution and measure the carriage movement, and calibrate from there.
Yep. Screw channels /= nut channels. You can put as many ball channels as you like around a single thread screw, it's just gonna change the length (and rigidity) of the nut.
Peter i read in datasheet than there is 4 chanels but i think like you there is only one ( chanel is how many enter you have in screw for one rotation ?) In openbuilds black box settings the units are step/mm so i have to put 400 step/mm (200x8 / 4 ) ? I understand that calibration make the job but i would like to have approximatively the good result and no damage with wrong manuel first calibration
Think of that as the "number of threads" that will engage with the screw In normal bolt-and-nut terms: Thats the difference between: (I can see about "2 channels") Here I can see maybe "5-6 channels" and And this one would have a tonne of "channels" as they call it But its just one Helix/Start - more contact area = larger load handling. No impact on the "pitch" This would have "5 channels" This is probably closer to the 1204 nut: 4 "channels" but its one loop Send a 1mm move first (use Incremental Jogging to test), even if you are out by a factor of 500 it shouldn't crash (; Yes, our 8mm pitch screws are +-200, so 400 should be good for 4mm pitch. The +/- comes in because screws are rolled, not ground (unless $$$) so there is some tolerance, theoretical is really close, but do check in Calibration