Mount/ enclosure design challenge
-
One thing I can say is that I am not getting any sort of consistent calibration of the filament sensor with this setup, there is just too much movement of the unit relative to the extruder. Its worth investigating Doug's idea, as whilst my design was more fiddly and made loading/unloading filament way more difficult, the function of the filament sensor was very stable, as the unit was rigidly held in place.
-
Yep, revision coming up.
And added. For now I have simply added a clip that clamps onto the bolt of the Nimble. It should be tight enough to hold the sensor in place.
I wanted to do a bracket that would fit under the screw, but that makes mounting it next to impossible and would make it very difficult to print.
But I have thought of another way to do it and will try that out as well. I intend to not put the bracket on the PCB Box but on the first stack. That way it can be assembled in place.If this works, I will have to change the shape of the Nimble and of the breech block itself to allow more space for the bowden between the sensor and the Nimble. It will make a re-design necessary, but hey, it is what I do.
-
Hi All
Thanks to everyone who has contributed a design, I count 7 people who have responsed to this thread and uploaded a design. Sorry I have been so slow in getting the beta boards out to you. Please can you email info@duet3d.com with your address and a link to your thingiverse/other design so I can get a beta filament monitor kit on the way to you.
Cheers
Tony
-
A little update, I have moved my design to a V2. https://www.thingiverse.com/thing:2551302
I have also included a modified version of Lykle's Nimble mount for smart effector which incorporates a nut trap, to hold the large nut keeping the heatsink in place. This prevents it loosening/turning, and means you can work on your heater block without worrying the nut will come loose (as you can't get a tool in there to hold it and tighten it). No need for Bowden tube between the Nimble and the filament monitor, even with 1.75mm filament the mount is stiff enough to calibrate to +/-9%.
I have also taken the opportunity to place a 3rd screw from the base of the filament monitor mount, which sits just below the drive cable down to the Nimble mount. This addresses the issue with the drive cable only being held on one side, but it only works if the connection between the drive cable base (which has a small notch on it to retain it) and my filament monitor bracket is tight.
-
Received my filament monitor today thanks!
-
Received mine too thank you!!
I changed my printer from bowden to direct during last week
I think bowden setup will be easier , you have plenty of space where to mount the sensor and previous mount fit perfectly. In direct mode I am worried about lack of space between sensor - extruder and hotend collect . Specifically in my setup with titan , could be hard to push the lever for manually insert the filament.
I am thinking to test a "fly "mount of the sensor.
hotend- PTFE tube- pneumatic connector - filament sensor- pneumatic connector - PTFE tube- filament reel
with the sensor not secured to the printer body, maintenance will be easy .
do you think could work? -
goold to hear you have received yours
If you look at the pictures on the wiki you will see i have mounted mine just above the titan, there is enough room to use the filament load leaver.
If you do go for a flying setup you need to ensure that the filament cannot push the sensor back up at all on retract or else you will get an issue with retractions not registering properly.
-
I am still setting up the printer. I hope I'll be able to do some tests during the next week end.
My worries with a filament sensor connected near the titan is that the ptfe tube connecting the 2 part ( titan and sensor) would be too rigid to make the lever works correctly. Do you experience such problems with your setup?This is a photo of what I would like to test.
-
I made a setup for the direct corexy D-bot printer here https://www.thingiverse.com/thing:2624832
I am afraid that the load lever will be very hard to operate with the the tube so near. I need to test it . -
Better late than never
Here is my version for the rotating magnet sensor, I had a variety of goals but the most relevant here was (1) to get filament counter information off the extruder idler wheel, the part the extruder drive wheel presses against, and (2) to develop my own extruder design to iterate and improve without waiting for parts in the mail. You'd think that there are enough extruder designs on Thingiverse (and only just got this working so it is not up yet) , but could not find a satisfactory one with a rotating spindle to connect to the magnet for the sensor.
I had to bend the 3-prong connector to come out the top and have not yet received some Dupont connectors so have to tie the Molex on with the white wire. Perhaps not clear from the second photo, but with the filament in place the two m5 drive sleeves align at the hobbed grooves and this puts the bar magnet at the right distance from the sensor chip.
-
i did one for my self-using an old broken stepper motor,s insides, and bearings if you remove the magnets you have a very well designed smooth running mechanism that can have a senor in to watch for rotation. all i needed to do is work out how to feed the filament it took me a few goes but i managed it.
it now uses a sensor watching the white dots painted on the inside of the wheel and if they stop turning it will send out an alarm. i just now need it to send a pause print to the duet card.
-
it now uses a sensor watching the white dots painted on the inside of the wheel and if they stop turning it will send out an alarm.
Interesting. How does your sensor work? (as in source and/or technology employed)
thanks -
Rob, nice work. How are you finding the accuracy of the magnetic sensor?
DMK, if your sensor outputs a switch like signal when it detects an error then you can setup the Duet to treat that input on the same way as a simple switch filament out sensor would work:
https://duet3d.dozuki.com/Wiki/Connecting_and_configuring_filament-out_sensors -
Rob, nice work. How are you finding the accuracy of the magnetic sensor?
Thanks Tony. Sadly can't say too much yet on the monitor as David said there's issues with that code in 1.21RC1 so stopped my calibration work until he can have another go at it. The numbers that look right have been mostly consistent to 0.1mm/rev though. That's a 3mm x 5mm bar instead of the disc magnet in the kit, if that makes a difference.
I have been meaning to post a comment that I prefer some kind of mechanical contact mechanism like this over the laser sensor just because it seems like it will be less subject to filament characteristics. Great having the non-contact system if you can make it work of course, but my interests lie in being OCD about the exact amount of filament being moved rather than detecting filament out or extrusion problems sufficient to need intervention on a print.
Ironically for calculating maximum volumetric extrusion rates, my extruder design makes so much noise when a tooth on the drive hob slips that I can do it without the electronics now
rob.
-
Wow, that is a nice aproach to a filament sensor. I will design and share a mount for the Anet A8!
-
How are you finding the accuracy of the magnetic sensor?
With my 3mm x 5mm magnet and and around 25mm circumference on the sensor hob I'm getting +/- 25% for E3.0. E1.0 takes that to +/-30%, E15 gets me +/- 5%.
So clearly there's better accuracy if the magnet can spin more. Gearing it would be effective but that's even more complicated mounting.Did you experiment with different shape magnets? I found a 2mm x 9mm bar option but shipping is expensive; do you think that would improve the sensor accuracy?
