V3 Smart Effector Defect

starworts

@dc42 So I can't for the life of me get this thing to read consistently and within any sort of reasonable tolerance. This is the output of ~25 G30 S-1 up 5mm as the commissioning tutorial runs just repeated with a one second delay between movements. Seems all over the place. The sensitivity settings are at the bottom to keep track of what tests I'm running.

5/21/2022, 3:35:51 PM	Stopped at height -0.370 mm
5/21/2022, 3:35:48 PM	Stopped at height -0.460 mm
5/21/2022, 3:35:45 PM	Stopped at height -0.599 mm
5/21/2022, 3:35:42 PM	Stopped at height -0.688 mm
5/21/2022, 3:35:38 PM	Stopped at height -0.451 mm
5/21/2022, 3:35:36 PM	Stopped at height -0.641 mm
5/21/2022, 3:35:32 PM	Stopped at height -0.688 mm
5/21/2022, 3:35:29 PM	Stopped at height -0.634 mm
5/21/2022, 3:35:26 PM	Stopped at height -0.641 mm
5/21/2022, 3:35:23 PM	Stopped at height -0.671 mm
5/21/2022, 3:35:20 PM	Stopped at height -0.630 mm
5/21/2022, 3:35:17 PM	Stopped at height -0.474 mm
5/21/2022, 3:35:14 PM	Stopped at height -0.541 mm
5/21/2022, 3:35:11 PM	Stopped at height -0.623 mm
5/21/2022, 3:35:08 PM	Stopped at height -0.670 mm
5/21/2022, 3:35:05 PM	Stopped at height -0.715 mm
5/21/2022, 3:35:02 PM	Stopped at height -0.617 mm
5/21/2022, 3:34:58 PM	Stopped at height -0.677 mm
5/21/2022, 3:34:56 PM	Stopped at height -0.676 mm
5/21/2022, 3:34:52 PM	Stopped at height -0.632 mm
5/21/2022, 3:34:49 PM	Stopped at height -0.640 mm
5/21/2022, 3:34:46 PM	Stopped at height -0.814 mm
5/21/2022, 3:34:43 PM	Stopped at height -0.455 mm
5/21/2022, 3:34:40 PM	Stopped at height -0.374 mm
5/21/2022, 3:34:37 PM	M98 P"0:/macros/Z_Tuning/8_Repeatability"
Stopped at height -0.572 mm
5/21/2022, 3:34:31 PM	M672 S105:20:235
5/21/2022, 3:34:30 PM	M558 P8 R0.4 F1000 C"zprobe.in+zprobe.mod" H5 A3 S0.15 T10000;

dc42

@starworts I am using a Smart Effector on my delta. I did have to make some changes to my previous setup using a version 2 effector to get it working well. In particular, I increased the sensitivity. I can't remember the exact value I used, I think it was about 20.

When using high sensitivities (i.e. lower numbers), there are a couple of things you can do to help them work reliably:

Remove capacitor C8 from the Smart Effector. This reduces the internal noise.
Make sure that the hot end fan wires run parallel (so either use figure-of-8 cable, or twist the two wired around each other) and not close to the electronics on the effector. The fan itself should be far away from the electronics.

starworts

@dc42 during
Fan lines are all far away from the electronics. Also currently not running fans during this testing due to any possibility of EMF issues. I removed C8 and am still getting a pretty big spread. I ran sensitivities from 10 to 30 (30 seems to not always trigger and crash).

5/23/2022, 9:21:44 AM	Stopped at height 0.303 mm
5/23/2022, 9:21:41 AM	Stopped at height 0.336 mm
5/23/2022, 9:21:38 AM	Stopped at height 0.413 mm
5/23/2022, 9:21:35 AM	Stopped at height 0.289 mm
5/23/2022, 9:21:32 AM	Stopped at height 0.297 mm
5/23/2022, 9:21:29 AM	Stopped at height 0.443 mm
5/23/2022, 9:21:06 AM	Stopped at height 0.412 mm
5/23/2022, 9:21:03 AM	Stopped at height 0.330 mm
5/23/2022, 9:21:00 AM	Stopped at height 0.261 mm
5/23/2022, 9:20:57 AM	Stopped at height 0.447 mm
5/23/2022, 9:20:54 AM	Stopped at height 0.483 mm
5/23/2022, 9:20:51 AM	Stopped at height 0.205 mm
5/23/2022, 9:20:48 AM	Stopped at height 0.311 mm
5/23/2022, 9:20:45 AM	Stopped at height 0.425 mm
5/23/2022, 9:20:42 AM	Stopped at height 0.422 mm
5/23/2022, 9:20:39 AM	Stopped at height 0.209 mm
5/23/2022, 9:20:37 AM	Stopped at height 0.308 mm
5/23/2022, 9:20:34 AM	M98 P"0:/macros/Z_Tuning/8_Repeatability"
Stopped at height 0.449 mm
5/23/2022, 9:20:03 AM	M672 S105:20:235

dc42

@starworts what M558 command are you using?

starworts

@dc42

M558 P8 R0.4 F1200 C"zprobe.in+zprobe.mod" H5 A3 S0.15 T10000;

dc42

@starworts that looks OK to me.

Two suggestions:

Here is a macro do probe a point on the bed repeatedly and report the mean and standard deviation. Please try it and see what results you get. You should send M558 A1 first so that it doesn't do multiple probing. You can change the X and Y coordinates if you want. I usually get standard deviation of 0.005 or better, except that if filament has oozed form the nozzle then the first probe throws it out.
To check whether any fans or heaters are causing interference, you can run the macro with the hot end cold, and also with the heater and hot end cooling fan running. The print cooling fan should be off.

var x=0
var y=0
var numProbes=20
while iterations < var.numProbes-1
  G30 P{iterations} X{var.x} Y{var.y} Z-9999
G30 P{var.numProbes-1} X{var.x} Y{var.y} Z-9999 S-1

starworts

@dc42
Hot end cold

5/25/2022, 8:31:19 AM	Error: Probe was not triggered during probing move
Error: Probe was not triggered during probing move
G32 bed probe heights: 0.570 0.441 0.240 0.430 0.492 0.213 0.223 0.329 -1.567 -2.932 -2.856 -2.898 probing failed probing failed probing failed -2.830 -2.969 -2.959 -2.827 -2.841, mean -1.087, deviation from mean 1.530
5/25/2022, 8:31:14 AM	Error: Probe was not triggered during probing move
Error: Probe was not triggered during probing move
5/25/2022, 8:31:12 AM	Error: Probe was not triggered during probing move
5/25/2022, 8:31:11 AM	M98 P"0:/macros/dc42"
Error: Probe was not triggered during probing move
5/25/2022, 8:30:49 AM	M558 A1

Hot end at 205C and bed at 60C No print cooling fans

5/25/2022, 8:35:45 AM	M98 P"0:/macros/dc42"
G32 bed probe heights: 0.297 0.298 0.280 0.356 0.182 0.294 0.473 0.204 0.374 0.279 0.355 0.146 0.179 0.205 0.443 0.253 0.366 0.359 0.356 0.089, mean 0.289, deviation from mean 0.097
5/25/2022, 8:35:12 AM	M558 A1```

dc42

@starworts that looks very wrong. What sensitivity have you set? With the nozzle cold, if you tap the bottom of the nozzle upwards, does the green LED flash?

starworts

@dc42
Sensitivity

M672 S105:20:235

It does not flash when tapping the nozzle

dc42

@starworts please submit a warranty claim to warranty@duet3d.com, linking to this post in your email. I'm sorry it has taken a while to sort out.

starworts

@dc42
No need for apologies! I am very happy with responses and working with me to figure out what was going on.