Misaligned Layers, Nozzle Nudges and Thermonuclear Meltdowns

Snapmaker 2.0
63

So, this is interesting.

My bed is hotter than it should be now

I used two different thermometers

I have it set to 70 and im seeing everywhere from 75-85 on it

I cannot explain this.

I thought maybe its the fact that its on glass and that reflection action is affecting it, but i was doing this yesterday and it was accurate. also, it does infact feel warmer to the touch

image
image763×817 114 KB

As of last night, I was noticing suddenly my petg prints were coming out differently, the skirt i usually keep .75mm away was been fusing into the parts and the stringing has increased significantly

i had decided that the filament must have been getting moisture in it over time so i changed spools to something else for awhile.

So, im not so sure i want to complain about the bed getting hotter, as i could simply dial it down - but i really dont know why this would be happening, the heated bed has such a hard time getting to that temperature.

Something may be going bad and its going to burn down or something :stuck_out_tongue:

Now, i have noticed that having my laptop plugged in, sometimes the machine acts a little goofy, whether or not i am connected on luban. perhaps it is just related to this for some reason

64

Have you looked at the bottom? It could be that the piece of kapton tape got a bit loose and the resistor isn’t touching the bed as good as it was before.
Also check the cable/solderpoints at the back of the heated bed. if it’s starting to fail it could theoretically also give a higher resistance (resulting in the bed thinking it’s colder than it really is). But usually that would be more “catastropic” and fail completely resulting in invalid values.

My first suspect would be the thermistor on the underside of the bed that has shifted place a bit.

65

Interesting.

No i havent taken the bed off or anything at this point.

I just made an observation, that one corner is much cooler.

Maybe something happened to the circuit over there so it isn’t heating up and therefore more power is going to the rest

I will have to try to take it apart to evaluate like you said I guess

edit: regarding the leveling grid above, the top left corner is low. Is this table upside down by chance? because the bottom left corner is the one that is not the same temperature.

edit 2: the front right corner is also lower, but it about more the temperature as called for while the other corner is about 50 degrees

66

One corner being cooler wouldn’t really matter. There is only one thermistor at the bottom in the center as far as I understand. So the temp-value of that one controls if the bed is heating or not.

1 Like
67

@MooseJuce

OctoPrint will run on any recent Rpi (not good on Zero) easiest way is to use the prebuilt OctoPi image.

OctoPrint.org - Download & Setup OctoPrint

Install the image on your SD card using BalenaEtcher

balenaEtcher - Flash OS images to SD cards & USB drives

Don’t bother with their fancy OS management stuff just use it to flash the SD card.

You then need to add the Bed Visualiser plugin.

The only thing you need in the GCODE Commands area is

M420 V ;get data

You need to change the Graph Z-Limits to cover the size of your Z numbers so in your case 7-10.

Octoprint is great, connect a rpi camera and you can see what’s going on, then OctoLaps plugin provides fantastic time laps video’s of builds, this is my first Benchy. Didn’t line the camera up quite right and the bridging is rubish, now fixed with the addition of an external fan.

1 Like
68

Shifting layers again. Defective module #4.

Of which, so far I have out of pocketed one due to not getting both of my spares i purchased back.

Out of an attempt to not drive them insane, I offered to buy one more if they send me a 2nd one so I can just have all new modules… because I know they are not going to be sending me two (even though they should be doing so).

69

Well…

I just had some catastrophic bullshit happen to me.

Running my job, my spool is getting low, trying to use it up so i let it.

Filament runout recovery kicks in.

I try to remove the filament and i can NOT get it out. Its stuck real hard.

Grab some needle nose and try to grab it from the inside to push up, WILL NOT BUDGE.

the half of the filament that was down in the heatbreak comes out and the upper part is stuck.

Try feeding needles, filament, allen wrenches, whatever i can to try to get that out, nothing works. fiddling with it for 45 minutes, it just will not come out. push in, pull out, heat gun, nothing. im even supporting the head and tapping a little metal rod with a brass hammer, nothing.

the top ring was off of there, easy access right in.

So, long story short, i get a drill bit and start chipping away at it.

Finally, i get through it, but its still stuck!

size up a bit and try again, again i get thru, this time it cleared out.

carefully dust out the machine with a paint brush, clean up all the remnents, feed some filament cleaning filament thru, load up the right filament, run like 6 load cycles just to get it good and primed…

press continue, filament runout detected… cannot get it to accept that theres filament in there.

So what is with this sensor? is it a limit switch that something broke off of and totally jammed up my 3dp module?

70

it also seems like the calibration routine has changed a hair, and now my glass clamps are interfering from the nozzle pushing down on it, so thats pretty cool too. at least i figured it out, at first i thought my sensor wasnt working

71

So a little update, per advice from Potter I disassembled the side plate to investigate the filament runout sensor, which as I expected is a microswitch

What happened was, then I presssed unload, the filament was fed behind the lever and jammed behind it, hence why I could not remove it and why it didnt come out when when drilled thru

image
image568×676 107 KB

While I have attempted to straighten the lever out on the switch, and am able to make the contact, I am unable to do so reliably. It requires some force in order to engage the switch.

I am not really able to meter it out, but you can hear it, its a standard snap acting switch.

With that said… This is a relatively easy switch to access and this may prove fruitful for modders who wish to add their own runout or other type of action that could depend on a runout scenario (forced filament change via a relay?)

I am trying to now see what will happen next, it is doubtful they will send me a switch, but maybe they will. It is a bit unusual in lever profile as far as i know, but I will check mcmaster carr to see if i can find a match.

72

I don’t know the exact model number and size, but that general class is called “bent lever” or “curved lever” microswitches. It looks like a submini size.

A style like this may have been a better fit for the application:

image
image750×752 62.7 KB

Honestly, I think the switch would be better applied turned around, where the filament pushes into the lever at a gradual angle rather than jamming onto the front. I don’t know if there’s enough room or wire slack, but you could see if you could flip the direction.

73

Yeah, its got a goofy kind of curve like a roller would be there to it. might just be better off with the roller lol (probably not)

im sure i can find something to stick in there, or just jump the contacts and not have the feature.

low key hoping i can talk them into sending me the updated 3dp module. ive gotten my share of warranty parts out of them though. they did finally end up agreeing to send me 2 more linear modules

74

I looked at it closer, most of the subminature switches i see are like almost .8" wide while this one is roughy 1/2"

so it must be some special thing that i am sure will not be easy to swap out due to space restraints.

75

I just looked up the first submini on mouser and it’s 1/2" wide:
image

image
Not the cheapest option, but they do exist, and I think it is a standard size.

76

hm ok, must not have them in mcmaster

oh a crouzet part, i got a crouzet guy. ill make em gime a sample

1 Like
77

Oh, there we go

Its not sub-miniature

its sub-subminiature

ok new term learned

image
image905×72 9.52 KB

1 Like
78

That’s new to me, awesome find. And Crouzet calls is a “simulated roller” lever, interesting.

79

@MooseJuice have you a defekt linearmodul? Think its the same as the endsops in these.

80

Oh? as a matter of fact i do.

ill check it out

81

crouzet has an odd way of doing things, so it could just be some bs they made up too.

i been dealin with them for 15 years now and they are soooo wacky

82

Naturally, the switch connector is not the same as the fan connector, so i guess ill be lengthening these wires lol