Re: Yes, 3D printers can go on fire.
January 29, 2014 07:32PM
tjb1,

I'm not trying to make a one size fits all solution. The solution I proposed is for ABS printing with a J-Head. I never pretended that the thermal fuse was a solution to EVERY possible combination of printer design. That would be just silly.

Lastly, I'm not trying to prevent a fan from dying, nor am I trying to prevent damage to my printer. I'm trying to prevent my house from burning down. You can bet that a properly sized thermal fuse, if used within it's capabilities, will be better than not using one. In my situation, I have measured temperatures on my hotend that allow meto KNOW FOR CERTAIN that I can use a specific type of thermal fuse which will only allow the hotend to possibly melt down and then the fuse opens shutting down the printer entirely. I can't help you with your 500C E3D. You'll have to continue to poo-poo other's ideas on how to prevent their houses from burning down while the rest of us look for a solution.

Have a good one! smiling smiley

Edited 1 time(s). Last edit at 01/29/2014 07:34PM by umdpru.
Re: Yes, 3D printers can go on fire.
January 30, 2014 08:49AM
This has been an interesting conversation and at a most appropriate time! I have just finished building my RepRap printer and completed the first test print and was thinking about building a sound-deadening enclosure for it but hadn't considered the possibility of fire.

Based on this discussion, I have decided to incorporate a smoke detector with a relay output inside the enclosure and use the relay output as part of a latching power switch - I need a switch for mains power anyway - so the power drops out if smoke is detected. I will also put the filament spool outside the enclosure and feed the filament in through a metal tube (to stop flame propigation). If the enclosure is tight and relatively flameproof, anything nasty happening inside should be quickly starved for oxygen and go out with power removed.

Overnight printing would be a definite advantage and a noisy smoke detector is a good idea. One that shuts down the power is even better!
Re: Yes, 3D printers can go on fire.
January 30, 2014 01:41PM
I might be dumb, but I don't see the interest in shutting down the power of the printer once it has already catch fire ?


Most of my technical comments should be correct, but is THIS one ?
Anyway, as a rule of thumb, always double check what people write.
Re: Yes, 3D printers can go on fire.
January 30, 2014 02:49PM
Quote
DeuxVis wrote:
I might be dumb, but I don't see the interest in shutting down the power of the printer once it has already catch fire ?

Did you read my previous post? I would have thought if you are Admin you had to read everything.

To repeat myself, early intervention is the key. Smoke is given off before flames appear and a fire becomes self sustaining.

I can think of maybe three ways a fire could start:

1. An electrical fire caused by a short circuit or failure of an electrical component - burning wires would give off smoke. Cutting the power early would stop it.

2. Failure of the hotend temperature control. I had this problem and it charred the plastic. Smoke filled my living room. I switched it off at the plug and it stopped immediately.

3. The only scenario I can think of where there are flames first is a solvent fire. If you keep acetone next to the printer you get what you deserve.
Re: Yes, 3D printers can go on fire.
January 30, 2014 03:13PM
Quote
DeuxVis
I might be dumb, but I don't see the interest in shutting down the power of the printer once it has already catch fire ?

There is a very real need to shut off power. If/when your PSU catches fire, wires will be exposed and can create a short circuit. This can cause sparks and another fire at your wall plug. A vicious cycle that can be prevented by just flipping a switch.
Re: Yes, 3D printers can go on fire.
January 30, 2014 03:36PM
Thanks KingRahl. I couldn't have put it better myself. hot smiley
Re: Yes, 3D printers can go on fire.
January 30, 2014 07:17PM
For those who have fume hoods, I've been thinking a smoke detector could be placed inside it, in the airflow. I'm building a second fume hood, and don't want an situation where all the smoke from a failing machine is sucked out of the house while the fire grows!

I'm also fiddling with a thermal fuse that tiggers at 184┬░C. I noticed it can handle 10A, so I am wondering if I could place it under the PCB style heated bed. It isn't capable of handling the 11A or so of the heated bed... but I could run the 5A side of the RAMPS 1.4 through it. Not sure what that would accomplish, as it would not shut off the heated bed directly, and the USB connection if plugged in would supply voltage to the Arduino but not enough to power the motors or the hot end!

Can two of these things be effectively wired in parallel to raise the amerage capacity above 11A? Sound iffy, but I should look into that...


Yvan

Singularity Machine
Re: Yes, 3D printers can go on fire.
January 30, 2014 08:12PM
I just realized, most of my machines have a switch function on the power supplies(Xbox, PC, etc.). I can run the thermal fuse to the switch wiring, and shut off the power supplies themselves. Obviously. smiling smiley

No more power supplies that have switches on the mains side for me.


Yvan

Singularity Machine
Re: Yes, 3D printers can go on fire.
January 31, 2014 05:56AM
Having the printer inside an enclosure fitted with a smoke alarm that triggers an automatic CO2 extinguisher system would be ideal.

However I cannot find an electrically operated CO2 system that doesn't cost an arm and a leg. Systems used for race cars and boats are also usually powder.
Re: Yes, 3D printers can go on fire.
January 31, 2014 07:59AM
thumbs down
Quote
umdpru
tjb1,

I'm not trying to make a one size fits all solution. The solution I proposed is for ABS printing with a J-Head. I never pretended that the thermal fuse was a solution to EVERY possible combination of printer design. That would be just silly.

Lastly, I'm not trying to prevent a fan from dying, nor am I trying to prevent damage to my printer. I'm trying to prevent my house from burning down. You can bet that a properly sized thermal fuse, if used within it's capabilities, will be better than not using one. In my situation, I have measured temperatures on my hotend that allow meto KNOW FOR CERTAIN that I can use a specific type of thermal fuse which will only allow the hotend to possibly melt down and then the fuse opens shutting down the printer entirely. I can't help you with your 500C E3D. You'll have to continue to poo-poo other's ideas on how to prevent their houses from burning down while the rest of us look for a solution.

Have a good one! smiling smiley

Poo-poo your idea? Go ahead and try it. You are going to have a lot more trouble than you think trying to keep the fuse at its thermal use temp and then tripping it reliably, but go ahead and burn your house down with false security. smiling smiley Have you ever seen a j-head fail at high temp? The heater assembly sometimes falls out of the top end so how will your fuse work then when its not near the heater and the heater is still running wild? Your idea is poo-poo thumbs down, difficult to implement at the correct location and gives false security. The only way I can see it being reliable is you take the 230 running/280 trip fuse and place it in another hole near the cartridge/resistor and keep your temps below 230 or try your luck and run it up to whatever below 280. You have too many variables placing the fuse away from the heater and if you want to then bet your house on that and sleep at night, more power to you.
Re: Yes, 3D printers can go on fire.
January 31, 2014 11:59AM
Quote
dave3d
Having the printer inside an enclosure fitted with a smoke alarm that triggers an automatic CO2 extinguisher system would be ideal.

However I cannot find an electrically operated CO2 system that doesn't cost an arm and a leg. Systems used for race cars and boats are also usually powder.

My favorite is to have the printer enclosed in a sealed compartment, with an oxygen free atmosphere. winking smiley All it would take is a big nitrogen tank from the local welding supply...

Do the CO2 systems operate on a one use principal? In other words, if triggered(even from a false alarm) do you have to send it in to the shop to have it re-filled? I had a kid flick the trigger on one of our powder extinguishers here, and it leaked down to nothing after that. Took ages for me to get it refilled.

If the printer is in an enclosure, what about a shot of nitrogen? It would not be as effective as CO2, but that would not matter in an enclosure. Below 15% oxygen, combustion stops I think. Way cheaper to just have a motor or solenoid open and close a nitrogen tank valve than have to deal with a certified fire extinquisher.


Yvan

Singularity Machine
Re: Yes, 3D printers can go on fire.
January 31, 2014 12:45PM
perhaps something like that : [www.safelincs.co.uk]

and not money heavy with the 1kg version... residue free

you put it over your printer... clean... efficient... sleep in peace..

edit : even better [www.amazon.com]

30$ of automatic security... but use powder...

Edited 2 time(s). Last edit at 01/31/2014 01:09PM by othar.
Re: Yes, 3D printers can go on fire.
January 31, 2014 01:02PM
Yvan, my background is the chemical industry. I have never actually seen an automatic CO2 extinguisher system but I have seen dry powder systems. They are single use set off by a glass bulb that breaks at a preset temperature and releases the powder under pressure. The entire contents of the cylinder are discharged.

It would be possible I guess to design a system to use a nitrogen bottle particularly if the printer was sited in the garage. It would be cheap. Just a solenoid valve and maybe a timing circuit to shut it off after say a 30 second blast.
Re: Yes, 3D printers can go on fire.
January 31, 2014 01:06PM
Quote
dave3d
Yvan, my background is the chemical industry. I have never actually seen an automatic CO2 extinguisher system but I have seen dry powder systems. They are single use set off by a glass bulb that breaks at a preset temperature and releases the powder under pressure. The entire contents of the cylinder are discharged.

It would be possible I guess to design a system to use a nitrogen bottle particularly if the printer was sited in the garage. It would be cheap. Just a solenoid valve and maybe a timing circuit to shut it off after say a 30 second blast.


look at the link i give just above...
Re: Yes, 3D printers can go on fire.
January 31, 2014 01:12PM
othar, the halon extinguisher from safelinks looks a very interesting possibility. The Amazon one is dry powder.

Edited 1 time(s). Last edit at 01/31/2014 01:13PM by dave3d.
Re: Yes, 3D printers can go on fire.
January 31, 2014 06:03PM
Othar, I wonder how the heat sensitive valve on those would work in a fume hood. I think it would work well for other applications, but the way mine are set up, a lot of heat would go out the ventilation system, that would really delay the responce time.

Maybe a smoke detector in the fume hood could be hooked up to a relay controlling the exhaust fan. When the smoke detector triggers, it shuts off air flow. That might let the extinguisher trigger much sooner from the heat build up. It would also give you a warning and a chance to responce before the more expensive extinquisher goes off.


Yvan

Singularity Machine
A2
Re: Yes, 3D printers can go on fire.
February 01, 2014 03:47AM
Quote
Yvan
Maybe a smoke detector in the fume hood could be hooked up to a relay controlling the exhaust fan.

If the fan was off, or marginally operational, the smoke detector wouldn't be effective located inside the ducting.
A2
Re: Yes, 3D printers can go on fire.
February 01, 2014 05:40AM
There is also the possibility that belts could instigate a fire.

Static Conductivity: Rubber or urethane belts can generate an electrical charge.

Static can ignite combustible dust, vapors, solvents, and other materials causing fire and or/explosion.
It can also cause unprotected electronic instrumentation to perform erratically.



\

All copper anti-static tinsel (Garland) is used to eliminate static electricity.
Functions as long as it is grounded, (below 150 feet per minute).
Tinsel has a minimum static threshold normally considered about 2000 volts.
Below this threshold tinsel typically becomes ineffective.
However, most static problems related to static charges do not show up at the threshold level.

Carbon filled Nylon filament brushes is another option.

Garland
[en.wikipedia.org]

TAKK
[www.statictinsel.com]

Edited 1 time(s). Last edit at 02/01/2014 05:41AM by A2.
Attachments:
open | download - ScreenHunter_379 Feb. 01 05.18.jpg (53.2 KB)
open | download - ScreenHunter_380 Feb. 01 05.21.jpg (102.2 KB)
Re: Yes, 3D printers can go on fire.
February 01, 2014 11:39AM
Quote
A2
Quote
Yvan
Maybe a smoke detector in the fume hood could be hooked up to a relay controlling the exhaust fan.

If the fan was off, or marginally operational, the smoke detector wouldn't be effective located inside the ducting.

Yes, that is correct. Locating a detector in the ducting would be awkward. For example, the ducts I have are anywhere from 3 to 6 inches, not very big. The smoke detector would be hard to access, and could obstruct flow. It would also be prone to dust accumulation.

I had been thinking the best spot might be very near the intake in the fume hood. In other words, air would flow past the detector before going through the ventilation system.


Yvan

Singularity Machine
Re: Yes, 3D printers can go on fire.
February 02, 2014 10:22AM
I have now wired my smoke alarm via a second latching relay and it cuts the mains power and stays off.

Hope now it does not need to be used.
Re: Yes, 3D printers can go on fire.
October 18, 2014 05:26PM
Just had a idea, will try it soon and see how it works out.

What if you just wrap solder wire (without flux core) with the appropriate melting temperature with one or 2 turns around your extruder head. Insulating the wires from the extruder with Kapton before of course. Trough that solder wire you power your heating element. If now the extruder overheats the solder wire will melt and reliably cut the power to the heater.
I don't think that all the sophisticated electronic approaches are robust enough to act as a real safety device. A safety device should be as simple as possible, so that there is no way that the safety device could fail itself too.

Just a thought smiling smiley
Re: Yes, 3D printers can go on fire.
October 18, 2014 06:17PM
Quote
CraigRK
Correct. Perhaps we should think about it like this:
  • For each printer the hot end temp should increase in some ratio according to the amount of time max power is applied
  • Could we track the power applied (time power was on) say over the last 5 minutes
  • Track whether temp is decreasing significantly in the same time
  • Correlate the 2. If we are applying power, then temp should be going up, not down

Sorry, not articulated very well. But rather than needing to set ranges or anything, the firmware should know if it is expecting the temp to rise, and if the temp actually drops, then warning bells should be ringing.

Then again, this is to cover the situation where the thermistor is pulled out of the hot end. I believe you can set Marlin to have 2 thermistors on the one hotend and stop based on a differential. This might be a quick win for people (although it presumes you have the hardware to make this a reality).

I second this idea. The pid should recognise if it isnt working (changed X, expected y, got z) and tell the printer to power down (after logging something to the console)
Re: Yes, 3D printers can go on fire.
October 18, 2014 06:46PM
Quote
Gregor R
What if you just wrap solder wire (without flux core) with the appropriate melting temperature with one or 2 turns around your extruder head. Insulating the wires from the extruder with Kapton before of course. Trough that solder wire you power your heating element. If now the extruder overheats the solder wire will melt and reliably cut the power to the heater.

What you are proposing is a well established principle and known technically as a fusible link.

There are many different types and grades of solder. For this application the solder needs form a eutectic I think. In other words a sharp melting point, otherwise it just goes gradually soft and mushy. The problem is finding a suitable composition.

I had a quick look at Wikipedia and a solder known as Pb-5Sn (95% lead + 5% tin) has a eutectic of 310 deg C.

Another possibilities are Zn95-Sn5 (95% zinc, 5% tin) this has a higher eutectic of 382 deg C, and even pure lead at 327 deg C.

Pure lead wire may be suitable and would be cheap.

Edited 1 time(s). Last edit at 10/18/2014 06:50PM by dave3d.
Re: Yes, 3D printers can go on fire.
October 19, 2014 06:46AM
Push notifications? i think this would count as a Fatal error and would send a message to your phone.

you could probably set up a smoke sensor to send a message to your phone as well, because, you know, its a fatal error.

or could set up a sensor to detect fire or smoke, and it could disconnect the power and make a ringing noise..

all circuits through safety feature,
then the safety features kick in, then it all turns off...

the link for example codes can be found here
[linksprite.com]

you could also put a tilt sensor, so if you machine falls over, it turns off main power.

Edited 4 time(s). Last edit at 10/19/2014 07:05AM by pushthatbolder.
Re: Yes, 3D printers can go on fire.
October 20, 2014 01:30AM
I like the Pb-5Sn eutectic as thermal fuse idea

Until you get that tested, don't forget that there are bimetallic strips (with NC contacts rated for mains power, designed to be bang-bang thermostats) that can operate at higher temperatures than polymer thermal fuses. 250C seems readily available on eBay for <$5 and would work well to open a latched relay; I have a couple on the way and will test to see if one can hold reliably at 240C when bolted to a hotend.

A bimetallic strip shouldn't be damaged in the long term by operation just under the trip temperature, unlike a thermal fuse. And it's readily testable!
Re: Yes, 3D printers can go on fire.
October 20, 2014 04:53AM
Quote
polyglot
I like the Pb-5Sn eutectic as thermal fuse idea

Until you get that tested, don't forget that there are bimetallic strips (with NC contacts rated for mains power, designed to be bang-bang thermostats) that can operate at higher temperatures than polymer thermal fuses. 250C seems readily available on eBay for <$5 and would work well to open a latched relay; I have a couple on the way and will test to see if one can hold reliably at 240C when bolted to a hotend.

A bimetallic strip shouldn't be damaged in the long term by operation just under the trip temperature, unlike a thermal fuse. And it's readily testable!

Could you link one please? I don't know what I'm looking for.
Re: Yes, 3D printers can go on fire.
October 20, 2014 06:39AM
Quote
UkIan
Could you link one please? I don't know what I'm looking for.

I bought item #151373405129. I also bought a 150C version to use as a safety trip on my heated bed, and I use the 60C version for thermal protection of my DIY audio amplifiers.

I expect some process variation between devices on switching temp but it may well mean that with careful selection, you should be able to get one that will reliably hold at 250C.
Re: Yes, 3D printers can go on fire.
November 13, 2014 12:16PM
catch on fire. that's scary.
Before if you are running the printer over night, guys & girls.


BotFeeder - Professional Manufactured 3D Filament
Re: Yes, 3D printers can go on fire.
February 20, 2015 11:44AM
This whole subject has me seriously reconsidering my existence in the 3D printing community. :'(
Re: Yes, 3D printers can go on fire.
February 24, 2015 05:03PM
Although it wouldnt protect a wiring fault, I put a thermal fuse in series with my hotend heater and mounted it to the hotend. That way, if the heater gets stuck on for some catastrophic reason, the thermal fuse would disconnect it.
Sorry, only registered users may post in this forum.

Click here to login