Z Probe

Z-Probe Calibration

---

NOTE: The Z-probe is currently optional. We are using the effector camera for calibration. IMPORTANT: If you skip this calibration (recommended), you must ground your probe pin. See [pins.h](https://github.com/firepick1/FireStep/blob/master/FireStep/pins.h) for your probe pin or use **syspb** to determine the probe pin. --------

We calibrate the Z-probe to verify repeatable accuracy. Your Z-probe should be accurate to one microstep, which is about 10 microns.

PREREQUISITES:

• Stepper Calibration
• [Latch Backoff Calibration](Latch Backoff)
• Z-probe operational and configured on pin specified by "syspb"

The Z-Probe Calibration process will Z-probe 9 times. Ideally, the z-probe values should all be within +/-10 microns of each other. The test will perform probes in the following order:

1. probe center
2. probe center
3. probe center
4. move A1 axis, probe center
5. move A1 axis, probe center
6. move A2 axis, probe center
7. move A2 axis, probe center
8. move A3 axis, probe center
9. move A3 axis, probe center

Notice that the first three probes involve symmetric arm motion, while the remaining six probe perform large movements at speed on individual axes between z probes. These large movements are mechanically stressful on the effector arms and may be useful in revealing step loss on individual axes.

Perform the test

The Z-probe test is a single FireStep command:

{"pgmx":"fpd-axis-probe"}

The test will take about a minute. Good data will be uniform as in this actual data taken from three invocations of {"pgmx":"fpd-axis-probe"}:

{"s":0,"r":{"prbd":[-42.524,-42.524,-42.524,-42.524,-42.524,-42.524,-42.524,-42.524,-42.524]},"t":45.793} 
{"s":0,"r":{"prbd":[-42.524,-42.524,-42.514,-42.524,-42.524,-42.524,-42.524,-42.524,-42.524]},"t":45.790} 
{"s":0,"r":{"prbd":[-42.524,-42.524,-42.524,-42.524,-42.524,-42.524,-42.524,-42.524,-42.524]},"t":45.793} 

Analyze data

If your machine provides repeatable data as shown above, your Z-probe functions properly and you should proceed to the next step in Delta Calibration.

Troubleshooting

If your data varies a lot, consider the following:

1. Z-probe assembly the LooseCanon Z-Probe must be free to pivot around its hinge but it should also be gently constrained from side-to-side movement. The choice of hinge wire is also important and a stiff 1mm wire with 5mm ends gently bent down at each end works best. The gentle bends ensure centering of the hinged interruptor. The weight of the thick stiff ends bent down will ensure consistent wire angle position. The Z-probe should feel loose but gravity should always bring it down to the same place.
2. Joint looseness Examine all FPD joints for excess play. If they wobble, that's bad. Tighten screws and/or springs.
3. Joint tightness Examine all FPD joints for excess tightness. Joints that are too tight will resist motion and stick, leading to inaccuracies.
4. Belt looseness FPD is inherently a low-backlash design (the effector always pulls the belt down) but a loose belt may slip on the pulley, since there are no teeth to grab.
5. Harsh acceleration FireStep controls acceleration using systv, which is the time to reach maximum velocity. A good value is 0.7 seconds, which is gentle and accurate. You can decrease this if it doesn't affect your Z-probes.
6. Maximum velocity FireStep can use a high pulse rate of 20KHz or more. This is faster than most other stepper drivers such as Marlin. If you're losing steps at sysmv:20000, try a lower sysmv:16000 or even sysmv:12000. If the lower speed fixes your Z-probe you can stay with the lower speed or repeat the Stepper Calibration.
7. Probe chatter The probe movement should be smooth and slow. If your probe chatters, try changing {"dimsd":400} or {"dimsd":300}. That's the pulse search delay. You can also experiment with the search pulse grouping, which defaults to {"syshp":3}. Listen to the video to hear a normal-sounding probe.
8. Stepper driver High current stepper drivers such as DRV8825 may not (?) run as smoothly as the lower current A4988. FireStep sips power where 3D printers guzzle it.