Hi together, I just set up the probe correctly, however when testing, the axes react too slow to the probe response. This is because I use Openbuilds for a refitted CNC tabletop metal milling machine where the axes (due to the high mass) need more time to accelerate. So I set up Z-axis acceleration to 1.000 mm/sec², for example. So somehow, I anticipated the problem. This would be no problem, if I could change the parameters of the probing process manually (speaking of reducing axis speed and possibly also deceleration, specifically). How could I do that? I cannot change the standard and homing speeds since it would take ages if they worked at the minimum speed I need for the probing process. Thank you in advance!
can you post a backup of your GRBL settings? slow acceleration should not be causing a probe failure, usually probing fails if the distance is reached before the probe makes contact, time is not measured.
It could cause a probe crash though 1mm/s^2 is extreme - Grbl does do a decelerate move after contact. CONTROL's macros uses stopping position (not PRB) for simplicity but a large overshoot on deceleration will make that quite inaccurate too Mainly, while I get that you have a heavy gantry, 1mm/s^2 just sounds like "i gave up and didn't even try tuning the real number" (or woefully underpowered for the machine)
Hi, Thank you for yours responsees. It's exactly how peter described, I already use 4A NEMA 23 with 5:1 planetary gear and I absolutely finetuned the acceleration, so it was no giving up. The head/z-axis alone is approx. 50kg in weight and although it has a large inbuilt transmission ratio, it just needs some time to accelerate and raise. As I know, much more than 4A would be difficult for the black box(?). The 1mm/s² is just for the Z-axis, the other axes are way lower in values. Also, lowering and decelerating/braking (as is the relevant case for the probing) wouldnt be as much of an issue, so if possible to set up a (lower) deceleration value different from the acceleration value in the CONTROL macro for the probing, this would be perfect. @David the swarfer: to prevent the anticipated crash, I put the probe on a foam for testing. The error didn't occur when touching the probe, but when the system pulled back but the probe signal persisted due to the foam expanding again. @Peter Van Der Walt: If the speed was slow enough, there wouldn't be a large overshoot nor crash. How can I edit the relevant macros in CONTROL for the automatic probing to be slower? Thank you in advance!
You'll have to use custom Macros on the Macros tab Constant force springs or counterweights might be something
ah, yes that is obvious, when it retracts it expects to see the probe connection BREAK during the retract move, but it did not move far enough to do that. this will be obvious from the GRBL settings and knowing that the probe is mounted resiliently in the first place. seems to me you need to move away from the blackbox toward seperate drivers that can run at ~48 volts. one of these .... GitHub - grblHAL/Controllers: List of controllers supported by grblHAL (I built a X32 clone from a ESP32 Devkit board that runs my external drivers jsut fine) Then you will need a suitable power supply and external drivers that can handle 48 volts and however many amps your motors are specified for. Alternatively you can use ClearPath servos. the higher drive voltage gives your best max speed and acceleration which your machine obviously needs. You may need to reduce the microstepping level since you have so much gear reduction. it is all a balancing act
