1

Topic: Melt down!

Has anybody modified their printer with fail safe measures for thermal runaway of the hot end and/or bed? Or of the controller if your active cooling goes to heck?

Let's face it, how many of us watch our printer during a four hour run? One hour? Half hour? Not me. Well, I did at first, but now it's so dialed in I almost went to the grocery store the other day during a four hour print run...and then I slapped myself across the face with a banana peel.

Aren't we just a bad mosfet or thermistor away from disaster? Now that I've got active cooling on my barrel (which, btw, is fantastic), what happens when my three dollar cooling fan goes to hell?

I don't know about your printer, but if my controller fan gives up the farm during the print it will start the wonderful "yskip" dance in about 5 minutes flat. So why not measure this too, and all our fans? I have never know more frustration in my life than loosing a print after running for three hours. Ok, I have, but that's for another forum. If a fan gives up the farm, maybe we send the printer into pause, and shut down the hot end temp? 

Does anybody have some thoughts on this topic, and can suggest some really simple mods that can add one level or more of safety to our machines? I'm really not worried about burning my house down, but a melted hot end or x carriage is not on my dance card either. Hey, it just might save your own hot end tomorrow! Oh, and if you got one of those fancy new all metal hot ends, you're not safe either, it might just melt your x-carriage or extruder mount instead, which might actually be worse than a nozzle melting out of the bottom of the PEEK barrel.

Let's think outside our little box. Web cams for our smart phones (is there an app for that?)? SMS alerts from monitors on printer? Adding another thermistor? Adding thermal or current draw measuring to the mosfets? Sensors? Pause commands?

I don't have the answers, just the questions at this point. I do think we need something. Thanks.

2

Re: Melt down!

A failed fan isn't a guarantee of a failed print, so it would be a shame to have a long print aborted unnecessarily.  After melting my J-Head PEEK for no apparent reason, I drilled a hole into the side of the new one near the bottom and installed a thermocouple so I can plug it in to my multilmeter every now and then to keep an eye on it.  It might be possible to add a thermistor there, and on one of the driver chips and have an emergency shutdown based on over temperature in those spots.

3

Re: Melt down!

I agree Ian, I think we have to be careful about shutting things down too quickly, but what you suggest sounds more than feasible. Seems like hot ends could easily have two thermistors, one for regulating temp (first line of defense), the other for shutting off the power should it reach a certain predetermined temperature (second line of defense).

On a related front, how many hot ends get fried because the filament breaks? Or, stated in another way, if your filament does break, and you're not around to notice, what are the chances your hot end will not produce a clog? About zero. So why not rig up a little seeing eye that monitors the passage of filament. After so many minutes of being at a certain temp, with no filament flow, it shuts down the hot end?

I know theses sorts of safety measures introduce complexity to the system, and may in fact hurt reliability, so we have to find the right balance of what's most important. Filament breakage is not that high on my list. I would suspect if it's happened to you it could be high on your list.

As for the fans, on my SD3 the extruder and controller fans (both upgraded to qu-bd 8cfm 40mm fans) are crucial. My 25mm hot end side fan and 40mm x-axis hugger fan, not so much. I agree Ian, it's possible critical temps wont be reached so there is no reason to abort the print. Seems like monitoring the temp is the way to go. As for the extruder fan, perhaps the thermistor on the upper portion of the barrel can serve double duty, by being the fail safe for the extruder and the hot end? So maybe the themistor gets mounted on the top of the hot end?  As for the controller, this could get its own thermistor.

That said, I've read that people who have upgraded to the Azteeg X1 and X3 don't even use active cooling. So maybe that's the path to take. And you get fan control for the front fan on your hot end.

Food for thought.

4

Re: Melt down!

With a ramps or X3 you get a thermistor input for a second hot end.  You could use that to monitor an additional location and set the second extruder max temp to what you think is appropriate.

5

Re: Melt down!

That is a great idea, totally going with the X3. I'll let you know when I get it installed, may need some help on the firmware side of things. Thanks Ian.

6

Re: Melt down!

I know this is slightly off topic, but I use a box fan with some creative bits of cardboard placed to cool my electronics and motors. Works like a charm and is a little more reliable (about not just crapping out I mean) I know its not an elegant solution but it does work.

I edit my posts a lot.

7 (edited by 2n2r5 2013-05-31 21:40:35)

Re: Melt down!

I have a spare Sanguinololu. I am sure I can find a way to use a thermistor to monitor the temp in the PEEK and kill the DC that feeds the other Sanguinololu board. Maybe with the mosfet of the hotend from my spare which can act as an inline power controller. My problem when mine went into melt down was that my mosfet was stuck open. Even when the max temp was tripped and the heater was told to turn off, there was still 12VDC feeding hotend. The only way to stop it was to kill the power supply feeding the board.

This is starting to sound like a frankenstein but I think it will work. What are the traces on the board rated for? Can I feed the entire Sanguinololu board through a mosfet? The mosfet is rated for 30A or something sick like that.

SD3 w/ mods:
Glass bed with QU-BD heat pad upgrade, threadless ballscrew w/ 8mm smooth rod, spectra line belt replacement, lawsy MK5 extruder, Lawsy replacement carriage, E3D hotend, Ramps 1.4 w/ reprap discount controller, DRV8825 drivers, 12v 30A PS, Acrylic case, Overkill Y-idlers, Filament alarm, Extruder fan + more.

8

Re: Melt down!

2n2r5, the stuck MOSFET is probably at the top of my list for problems needing a solution. Next would be getting a backup thermistor in play, mounted higher up the peek barrel.

Perhaps the techno wizards on this forum could come up with a solution to trip the dc to the board, as you suggest. I think it's a great way to handle the problem.

9

Re: Melt down!

Just use that spare sanguinololu and MOSFET to trip a 30A relay...

10

Re: Melt down!

The Sanguinololu has extra analog pins that you might be able to use for an additional thermistor.  I don't know how you would incorporate that into firmware however.

11

Re: Melt down!

IanJohnson wrote:

The Sanguinololu has extra analog pins that you might be able to use for an additional thermistor.  I don't know how you would incorporate that into firmware however.

Even better!

It'd just be a few lines. You'd probably want a resistor and a capacitor - the Sanguinololo uses a 1-pole RC filter to smooth things out.

Firmware is pretty easy - look for the code that holds the overtemperature fault trip, and add a line to check your new thermistor.

12

Re: Melt down!

The problem comes with making it trip a relay to shut itself off. I guess with would have to be a normally open circuit that gets held closed while conditions are a-okay and goes open when thermistor screams or loses power/fails to read.

SD3 w/ mods:
Glass bed with QU-BD heat pad upgrade, threadless ballscrew w/ 8mm smooth rod, spectra line belt replacement, lawsy MK5 extruder, Lawsy replacement carriage, E3D hotend, Ramps 1.4 w/ reprap discount controller, DRV8825 drivers, 12v 30A PS, Acrylic case, Overkill Y-idlers, Filament alarm, Extruder fan + more.

13

Re: Melt down!

I have a few spare hotends too. Maybe I will test it out with my ceramic one so I don't have to worry about melting it.

SD3 w/ mods:
Glass bed with QU-BD heat pad upgrade, threadless ballscrew w/ 8mm smooth rod, spectra line belt replacement, lawsy MK5 extruder, Lawsy replacement carriage, E3D hotend, Ramps 1.4 w/ reprap discount controller, DRV8825 drivers, 12v 30A PS, Acrylic case, Overkill Y-idlers, Filament alarm, Extruder fan + more.

14 (edited by adrian 2013-06-01 05:58:57)

Re: Melt down!

Simpler electromechanical solution, which thus removes the need to ensure that a watchdog is inplace on the Sang to ensure the firmware hasn't locked up causing the issue....

http://www.ebay.com/itm/1-pcs-Temperatu … 0613907818

Stick it against the PEEK barrel a bit further up than the nozzle end, inline with the heater cartridge.

http://www.ebay.com.au/itm/KSD301-120-C … 0915910431

Stick that one inline on your Bed.

15

Re: Melt down!

adrian wrote:

Simpler electromechanical solution, which thus removes the need to ensure that a watchdog is inplace on the Sang to ensure the firmware hasn't locked up causing the issue....

http://www.ebay.com/itm/1-pcs-Temperatu … 0613907818

Stick it against the PEEK barrel a bit further up than the nozzle end, inline with the heater cartridge.

http://www.ebay.com.au/itm/KSD301-120-C … 0915910431

Stick that one inline on your Bed.

Those are convenient temperatures! smile Great idea!

I like this one for the hotend, I would put it right on the brass barrel if there is room.

http://www.ebay.com/itm/2PCS-KSD301-Tem … 1c28219e18

SD3 w/ mods:
Glass bed with QU-BD heat pad upgrade, threadless ballscrew w/ 8mm smooth rod, spectra line belt replacement, lawsy MK5 extruder, Lawsy replacement carriage, E3D hotend, Ramps 1.4 w/ reprap discount controller, DRV8825 drivers, 12v 30A PS, Acrylic case, Overkill Y-idlers, Filament alarm, Extruder fan + more.

16

Re: Melt down!

adrian wrote:

Simpler electromechanical solution, which thus removes the need to ensure that a watchdog is inplace on the Sang to ensure the firmware hasn't locked up causing the issue....

http://www.ebay.com/itm/1-pcs-Temperatu … 0613907818

Stick it against the PEEK barrel a bit further up than the nozzle end, inline with the heater cartridge.

http://www.ebay.com.au/itm/KSD301-120-C … 0915910431

Stick that one inline on your Bed.

Those are pretty big. I'm not sure if they'd get to 200C internally, even with the outside of the PEEK at 250C+. What you really care about is the temperature at the threads of the PEEK, and the only way to get that is drill the PEEK, insert thermistor.

17

Re: Melt down!

2n2r5 wrote:

The problem comes with making it trip a relay to shut itself off. I guess with would have to be a normally open circuit that gets held closed while conditions are a-okay and goes open when thermistor screams or loses power/fails to read.

then you have a relay that is being held closed for hours and could burn out at any time, thus shutting down a perfectly good print. probably better to just monitor your printer occasional, maybe a simple audible alarm to alert you of a condition in need of your attention.

lets face it, this technology is far from mature enough to hit print and forget like a laserjet. lol. i check on mine every 20-30 minutes, and i dont start a print if i know i have to go to sleep/out soon. i have had a few filament breaks so far, so i am designing and will be incorporating some sort of micro roller switch alarm bolt on for the mk4 extruder, complete with a coil spring filament strain relief for dealing with those pesky brittle filaments. could work an overtemp alarm for the PEEK into it with relative ease. but that project is superseeded by several others at the moment, like getting my rostock up and printing as a backup should anything happen to the SD3. haha.

18

Re: Melt down!

As long as the right relay is picked with the right specs you should not have a problem with the relay burning out... it's when designers try to push the envelope on parts and push them to 100 percent or more of their ratings that you start to have problems... would be akin to the fuel pump relay in pretty much every automobile made in the U.S. since the late 80's early 90's they all have electrical fuel pumps with a relay on as long as the motor is running.

19

Re: Melt down!

What ronsii said... Or use a 2 transistor latching circuit as per some estops.