Magnetic Filament Monitor Odd Sensitivity
-
I put on a different MFM this one a V1 also with a 1.7b PCB and it working as expected.
Duet3D rotating magnet filament monitor v3 on pin 121.io1.in, disabled, sensitivity 26.00mm/rev, allow 70% to 130%, check printing moves every 3.0mm, version 3, mag 129 agc 82, measured sensitivity 25.76mm/rev, min 98% max 103% over 30.5mmWhat are the chances the PCB is bad? I guess I can swap them out too... I can try redoing the case as well but the values for mag and agc are within spec. [EDIT] No they aren't mag is a little too high in both cases... okay see what I can do to fix that.
-
Okay using a different 1.7b pcb and v1 mechanical parts I was able to show that its the mechanical side of the sensor that's causing the error. Probably the housing... so Not much anyone can do here to help I'll just have to make a better housing. Thanks for letting me ramble
-
@nurgelrot I've been keeping an eye on the thread, but hadn't had anything much to add. I take it you used the housing designed for each type of rotating assembly when swapping? If so, just out of interest, how did you print and what layer height? You might be interested in the thread over here, it discusses some of the issues you can run into when FDM printing the case. I've never had crazy figures like that out of the MFM, I have to admit!
-
@lo-fi I typcally do the housings on my resin printer. But the V2 housing I printed the 2.5mm holes even on the smaller hole version where too large to give the threads any bite. So I did second one on an Voron V0 in ASA with .12 layer height and it looked okay but clearly has issues. I'm going to shrink the holes slightly in the model and reprint in resin when I get time. Or possibly test your redesign
Id have to change how I mount it but that's no big deal. -
Thanks for the info, that's interesting. You may find that the 1.2 layer height had an impact on the housing. One of the critical distances is between the bearing flanges, with the next being the PCB standoffs. For V1 and V2, the flange to flange distance is 6 and 5mm respectively. 1.2 divides in 6 neatly, so you'll get 5 full layers on the V1. 1.2 into 5 is rather more awkward at 4.16 recurring, so depending on the slicer and printer, you may actually get 4.8mm which is definitely too small.
I can't remember the stock housing PCB standoff heights - I have 5.9/6.9mm from the top bearing flange (respectively), so 1.2 may be an awkward division there too. The official design is suggested 0.2mm layer height, though I found that a little course in my experimentation with my own design and opted for designing in 0.1mm increments.
Anyway, after all that waffling on
: Give it a go at 0.1mm layer height, you may find it all fits a little better I'm getting closer to releasing my updated design for anyone that wants to try too.I'm sure I remember some documentation saying that it's possible for the magnet to be too close to the chip, though in typical fashion I can't find it now!
-
@lo-fi Resin printed a new housing and I thought it had fixed it but the sensitivity mm/rev started out at a great 27mm but just kept creeping up as the print went on making 5-10mm jumps every 20-30 second until it was well over 150 and climbing. So something is still off. Not very pleased with the V2 magnet at this point. I have 4 of the V1 and all of them work great. Lucky I have a spare V1 To use for now,
When you get the kinks ironed out send me a link to your housing and I'll print it out either 1mm layer or in resin and give it a go. -
@nurgelrot That's a puzzle for sure... I've got the V2 magnet to try myself, so I'll keep you posted on how I get on and send some links your way. Maybe I can replicate it!
I'd be interested to see some M591 D0 outputs throughout that print if you still have them handy.
I'd really like to know why the change V1 to V2 rotating assembly. I'd assumed the larger magnet means the distance between it and the chip isn't quite so critical, but that's pure conjecture on my part.
-
@lo-fi ha .12 not 1.2 I cant type
-
Okay Solved. Housing had the magnet too close. I shimmed it with 5 layers of 20# paper. that got the agc up to 112 and the mag down to 128 and its giving reliable readings. Going to modify SLT that I already fixed the screw hole size on to make ~.5mm more space for the magnet when using resin
-
@nurgelrot lol, I should have realised!
That's a great result and valuable info. Bet the previous one got worse as the housing relaxed a little and the mag got closer than it should be.
-
@lo-fi For the record if anyone is reading this. .5 was too much but I printed different sizes in a batch in the resin printer. The one with .3 added to the inner surface of the lid of the MMF v2 housing got me to the magic agc 110 mag 128 that seems to be working... All magnets vary of course so maybe mine was just a little bit strong.
But if you have a sensor that's reading way high mm/rev sensitivity back your magnet off a tiny amount. the tolerances in the thing are super tight.
