Hello!

I've got an old MKS Sbase that used to run smoothieware. I forgot I can't just dump everything off the SD card, lost my config file. Instead of setting it up again, I decided to try and switch to Marlin

I can't for the life of me figure out what's going on with the probe on this thing. From what I see it should be working, but it doesn't. I am pretty sure some software check is preventing the probe cycle from ever running.

If I use M43 E1 I can see my probe trigger like it should when I hit it. I know MKS Sbase is kinda junk, but in this case I do not think the hardware is the issue. I removed pullups and greatly reduced the series resistors. My opto endstops no longer float.

If I run the probe with debugging I eventually get an error "SLOW Probe fail! - No trigger. Triggered early." I do not understand the text of this error, seems contradictory. Regardless, the probe does not trigger as indicated by M43 or the light on the smart effector. I have tried inverting the logic of the probe input, doesn't change anything.

If I am at max Z right after homing and run G30 nothing happens, no movements. If I jog down close to the bed it will move up and then error out. I've never seen it run down, only up. It will only move when within 10mm of where it thinks Z0 is.

I've tried a bunch of different probe offset values. I used to use Z-0.2 on smoothieware, one thing I do remember, I'd get perfect first layers regardless of layer heights, no messing with a first layer offset or anything. I take that as the probe taking +0.2mm of offset?

I am at a loss. What's it take to get the damn thing just probe and return the travel distance!?

Thank you in advance for any pointers!

What probing speed are you using? The Smart Effector likes quite a high speed, such as 20mm/sec.

Have you configured Marlin to delay between moving to the correct XY coordinate and starting the probing move? Otherwise the XY movement may trigger the probe. This may be why you had a "Triggered early" message. A delay of 0.2 seconds is usually sufficient.

---

Large delta printer [miscsolutions.wordpress.com], E3D tool changer, Robotdigg SCARA printer, Crane Quad and Ormerod

Disclosure: I design Duet electronics and work on RepRapFirmware, [duet3d.com].

If I am reading the config right, 10mm/min for a fast speed, and 5mm/min for slow speed. These are probably slow, but the probe won't even move in the proper direction so tuning speeds isn't a concern at this point.

The probe is not triggering when it should not you can see this happening, the code in the effector will latch and hold so you cannot miss a trigger on the led. Delays or false triggers are not an issue at this point. A command to probe can be issued and it will fault out before moving. Inverting the input pin does not change the behavior. The probe will trigger the pin change as shown in the M43 E1 endstop watch test. It works fine on smoothieware. All this together makes me positive it is not the hardware.

I must be doing something else wrong.

Edited 1 time(s). Last edit at 05/03/2021 04:36AM by deezums.

@deezums, there was an error in the latest releases of Marlin (both bugfix and 2.0.8) that caused probe errors. It was fixed yesterday (May 2). Try downloading a fresh copy of Marlin again.
The error was in the motion.cpp file.

That looks like it'll fix it. At least I'm not going crazy! Not sure how I missed that issue on github...

Thank you!!

That fixed it, the probe works now! Can even use the sensitivity M codes I never was able to use with smoothie!

Edited 1 time(s). Last edit at 05/03/2021 02:28PM by deezums.