Help with pulse-generating Filament Sensor

dc42

Your M591 command is OK except that I find the L parameter suspiciously low. Do you really get 1 pulse per 0.037mm of filament motion through the sensor?

What does M591 D0 return?

Please note, you probably need to be printing from SD card for data to be collected.

kordris

This is what im getting when put the M591 D0 in G-code command prompt. I can tell the endstop is getting a signal because its flashing on and off. I have also tried to pause and resume the print. But I still pretty much get the same thing except the minimum will change and maximum and the over MM will change. But the current position will never update. This is where I think its not updating.
When i was watching the red light I could count about 3 to 4 secs between the pulses.
Also I print from the internal SD card is that what you mean printing from the SD card. Or printing from the SD card from the touch screen?

Pulse-type filament monitor on endstop input 4, disabled, sensitivity 0.500mm/pulse, allowed movement 60% to 140%, check every 5.0mm, current position 0.0, measured sensitivity 6.650mm/pulse, measured minimum 7%, maximum 8% over 2334.0mm

T3P3Tony

@kordris as dc42 said, you L value is way too low:

measured sensitivity 6.650mm/pulse

which may be correct or may be long between pulses. depends on the sensor. what does the sensor manual say about pulses per mm?

anyway try setting the L parameter in M591 to 6.65

kordris

This is the pdf on the filament sensor, it says use 7mm on marlin for the sensor. But I dont think thats for the pulse. It doenst give much info on what other info for setting for duet stuff. Im having to figure it out for my self.
PDF File Github Link

T3P3Tony

certainly looks like 7 or 6.65 would be a good start point for the L parameter.

kordris

Even if I use that, i still get no postion data. Which then leads to the printer stopping because it says theres to little movement. I dont think it's getting postion data correctly. Is there anything else should I enable in my config.g for the endstop on the extruder side. Since I'm using the extruder endstop E1. When I tried to get calibration data, it's the se way. Theres no postion point. So as soon as I go to enable the filament sensor it's going to error tell me there little or no movement.

T3P3Tony

@kordris run the monitor disabled by using the M591 S0 switch for now so we can gather some data.

looking at our previous output 6.650mm/pulse, measured minimum 7%, maximum 8% over 2334.0mm that shows that is was sensing something as it measured some minimum and maximum. What does M591 D0 show now?

I have not tried it myself but I am not sure what position data is displayed for a pulse type sensor. it may just be movement as the pulse sensor does not send position.

dc42

A pulse generating sensor doesn't distinguish between forward and reverse motion, so position cannot be inferred. I'll check why the firmware prints a position.

dc42

I've checked the code, and I confirm that the "current position" has no meaning for a pulsed filament monitor. I'll remove it from the report in the next firmware build.

kordris

Ok, so when I try to run it with just those lines. I get little movement error. This is why I think that line postion has something to do with the error. Because it expecting to get the position data but its not. This is the last calibration data that I got. I guess until you fix it. I wont be able to use it.

Pulse-type filament monitor on endstop input 4, enabled, sensitivity 6.650mm/pulse, allowed movement 60% to 140%, check every 3.0mm, current position 0.0, measured sensitivity 12.004mm/pulse, measured minimum 0%, maximum 111% over 24.0mm

kordris

When I change the E to 7mm distance I get a lot longer print time before it errors out before it wants to give me the problem with no little movement. So I guessing I might be close to the right settings maybe?

M591 D0 C4 P7 R60:140 A1 L6.65 E7 S1      ; pulse-generating sensor drive 0 Connected to C4 input, 40% to 130% tolerance, 7mm detection length
M591 D0                                   ; display filament sensor parameters for extruder drive 0

This is the last one M591 D0

Pulse-type filament monitor on endstop input 4, enabled, sensitivity 6.650mm/pulse, allowed movement 60% to 140%, check every 7.0mm, current position 0.0, measured sensitivity 6.886mm/pulse, measured minimum 47%, maximum 104% over 185.9mm

kordris

@T3P3Tony said in Help with pulse-generating Filament Sensor:

M591 S0

To get the Calibration data better do I just run M591 S0 ? I do I run the full M591 command with the S0 on the end like I had and then just check it with M591 D0 command.

dc42

@kordris said in Help with pulse-generating Filament Sensor:

@T3P3Tony said in Help with pulse-generating Filament Sensor:

M591 S0

To get the Calibration data better do I just run M591 S0 ? I do I run the full M591 command with the S0 on the end like I had and then just check it with M591 D0 command.

See https://duet3d.dozuki.com/Wiki/Duet3dFilamentMonitor_LaserVersion#Section_Calibration. It's written for the laser filament monitor, but it's the same for your filament monitor.

As it appears that your sensor only generates 1 pulse per approx. 6mm of filament, you will need to set the measurement distance (E parameter) several times greater then 6mm in order to get consistent results.

kordris

@dc42 Is the position data going to cause any problems or is that just going to just unrelated till you fix in the next firmware fix? I try to set the pulse to higher number. But I kept getting a error about every 5 mins telling little or no movement when I tried to enable the sensor with the S1. So Im guessing I still don't have something just right.

dc42

No, the position data being reported as zero shouldn't cause any problems. However, I can't guarantee that the code is entirely correct, because it was adapted from the code for Duet3D laser and magnetic filament monitors, so it isn't well-suited to low-resolution sensors. With a window of 60 to 140%, my guess is that you will need at least 20mm measurement length, perhaps even 25mm. A wider window will allow a shorter length, but even then the measurement length should be more than twice the extrusion length per pulse.

lifeform

@kordris would you like to share with me how you plug the cable of the bigtreetech smart filament sensor with duet?

keith89

@lifeform not sure if you figured out the wiring but if you trace the wires to the sensor, there is a 4 pin connector with only 3 wires. the single wire on the end with no wire next to it is Vin. the middle wire is signal and the wire right next to that is ground.

Back to the original topic, has anyone figured out the correct mm/pulse for this sensor?

omni

Hi guys,

Has anyone managed to properly set the Bigtreetech smart filament sensor to Duet meanwhile?

I am just finishing my build with Duet 2, and I have connected the sensor - but still haven't done any configuring.

Maybe the settings used in Marlin 2.0 can help?

So if anyone of you managed to set it correctly - please help.

Thanks!

dc42

Start with https://duet3d.dozuki.com/Wiki/Gcode#Section_M591_Configure_filament_sensing.

omni

@dc42

Thanks for the answer.
I've tried it yesterday - but it seems that Duet is not recording any data from the sensor.

I configured the sensor as following, just to log data when printing :

M591 D0 P7 C"^e0stop" L0.07853981634 R90:110 E3 S0

but it did not receive any data.

Any ideas on what I can try is highly appreciated.

Thanks again