Answered Inoperative C2 filament run-out switch

Original setup for this:
1. print jobs have been running fine as long as there's filament
2. I have not been receiving the filament run-out warnings in OctoPrint
3. the filament sensor plugin appears to be running in OctoPrint
4. at least one job has run out of filament in one way or another and yet the printer blissfully "air prints" without knowing that it's out of filament
5. I attempted to run the filament through the other hole in the back of the printer with no difference whatsoever

Taking the printer apart now and getting at the sensor, I see what the underlying problem is. The switch design is painfully stupid, in my humble opinion. The mechanism is a proprietary blade switch which hopes to touch the filament, only the blade is so small that it will only touch the filament if the orientation is so that it's curled in a particular way. It's effectively an inoperative part if you're lucky.

There are times in which if you actually manage to close the switch it will remain stuck in place regardless of whether or not you remove the filament.

At its best, it's a filament guide. At its worst, it will prevent your print job from running if it thinks there's no filament. At the moment, I have no work-around to suggest.

A proper switch design would push something out of the way if 1.75mm filament is inserted into the channel. This "something" would connect a switch.



Something shaped like this would work for the actuator, only the size of the particular black box is all wrong for this. You'd need to redesign the box to accommodate it.



This crash sensor for the Arduino may be closer to what we're looking for but it's a three-pin connector rather than two. Ultimately, this switch may be the one to have.

 

Every reference to the C2/R2 sensor I've seen so far was it being optical. Maybe you have a very early C2 and it changed soon after?

 

Really? (you guys have optical out-of-box) Weird. They must have realized that the blade switch was fail.

Mine has a red/black pair that goes to pins 19/20 on the Raspi and would be a N/C when the filament's installed. That would be the pair of GPIO pins on the far opposite side from power.

If you're using an optical version then in theory yours provides power for the LED. Would be good to know if you have more than two wires to yours.

I'll likely write-up the fix and will eventually visit Fry's to see what they have. The car's in the shop at the moment (radiator replace) so I'm stuck at a Starbucks.

 

Haven't taken mine apart but it sems it's 2-wire, so could be switch only. But it works.

 

Back from Fry's. Picked up four of the micro-switches and two of the opti-style switches, noting that the latter have the expected four conductors for these.

Although I prefer the optical coupling on this and am surprised that mine has a simple switch, I do want to later work with the clear filament and worry that the optical switch will be confused by this. My plan then is to use two switches, wire them in series and then simply park a bit of filament in the other channel until I later upgrade this to two extruders.

And I also managed to snag the speaker for the sound event upgrade (plus the Amazon Alexis upgrade as well).

 

My filament switch was open from day one. I got no continuity at the black Raspberry 40-pin connector between the red and black wires. I pulled apart my C2 back presuming the switch was bad. I found out the switch was fine (not optical), but the red wire into the black Raspberry connector was open. I pulled apart the black 40-pin connector and found the blade was spread apart so there was no connection between the red wire to the connector blade. I replaced the connector with a new one I had, and it now works great.

 

I wonder which is the newer design - switch or optical? While my unit was purchased new about 2 weeks ago, the newest file on the SC card (before updating) was 14-Jan-2017 (config.txt). I expect my C2 was made in January to March timeframe, but hard to tell when they updated the factory SC card image.

The switch makes a soft mechanical click you can hear when you insert/remove the filament from the outside hole. There is no switch on the inside hole.

 

I don't have the photos or STL uploaded yet but here's the project I was working on today. It's now soldered and the switches installed in the part, awaiting the swap-out.

 

Nice project! You might add a paragraph on how to remove the back. Something like -

First remove the two screws holding the white top on, and snap the top off. Pull out the black front panel (pull out starting from the top - it is only a snap in fit). Lay the front panel down, you don't need to disconnect it. Disconnect all the connections at the bottom rear (power, etc.). Remove the black filament holder. Disconnect the tube and the flat cable from the extruder head. The flat cable connector has two levers that you push out to release the connector.

To remove the rear white cover, four screws must be removed. The top rear screws are fairly easy to get to, about 4" from the top on both sides. The two lower screws are below this by 7" or so and are tricky to get to. A magnetic phillips screwdriver with a 12" shaft makes this easier. With the screws removed, the back should lift up and out.

 

Thanks, I usually include photos when I get to that point (see this one, for example). To be honest, the switches were giving me problems yesterday. I spent a fair amount of time adjusting them, soldering them, hot-gluing them in... only to find that in the pre-test it doesn't seem to close at all. I'm starting over to take advantage of the new BuildTak...

...and I'm managing to get enough part adhesion to forego the raft. So the pretty side of this (the back) should look a lot nicer anyway.