Welcome! Log In Create A New Profile

Advanced

Yes, 3D printers can go on fire.

Posted by Antslake 
Re: Yes, 3D printers can go on fire.
January 17, 2014 05:32PM
brnrd: yes, you are correct. I didn't think of that. A direct short would be the same as overtemp.
Re: Yes, 3D printers can go on fire.
January 17, 2014 06:40PM
Quote
cnc dick
Quote
dave3d
umdpru: you have just given me an idea. Is it possible to pick up a signal from a smoke detector that can be used to shutdown a printer?
I think that's a great idea.This is what I mean people start throwing ideas around brainstorming is what we used to call it all of a sudden one pops out of thin air

There are inexpensive interconnectetable smoke detectors that you can buy for less the $20 from hardware stores like homedepot. They are hardwired to AC (black and white wires) with a 9V battery backup, and they have a red wire that can be interconnected to other units so that when one goes, the others also sounds the alarm. I read somewhere that a unit sends 9V to the red wire to when it detects smoke. This line can be connected to one of the analog ports of the Arduino and sensed in the event loop of the firmware. I can also be used to control a relay to shut off the power to the whole printer.

Edited 2 time(s). Last edit at 01/17/2014 06:41PM by brnrd.
Re: Yes, 3D printers can go on fire.
January 17, 2014 06:55PM
I am in the UK. We don't have a Homedepot.
I have been looking for one with a relay. I would prefer it to be totally separate to the printer and wire it so it cuts power to the mains plug. I have found a module from China:

[www.ebay.co.uk]

Put the whole thing in a vented box with a larger mains relay and a mains socket.
Re: Yes, 3D printers can go on fire.
January 17, 2014 07:44PM
One way that I can imagine implementing a software safety check is to make sure that the temperature responds to a probe voltage.

For instance, once at temperature, you enter something like a simple phase-locked loop by applying a small oscillation voltage to your heating element and look for the frequency components of the measured temperature which correspond to the oscillation probe voltage. If the thermistor is not reading the temperature of the hot end, it should be able to detect the absence of those frequency modes with an inability to phase lock.

This might be asking too much for a 16mhz atmega, but I'm sure some of the ARM chips could do it.

Edited 2 time(s). Last edit at 01/17/2014 07:47PM by iquizzle.
Re: Yes, 3D printers can go on fire.
January 17, 2014 07:59PM
I forgot to mention, in RAMPS-FD there is a hardware interlock on the stepper enables and heater outputs, which goes some of the way to a hardware safety solution. This could be connected to an EMS (Emergency Stop switch), or other device such as thermal switch. It's common in CNC boards to use a charge pump to provide a watchdog timeout, but I stopped short of that.

A hardware interlock could be connected directly to mains power relay which could be applied to any printer of course.


What is Open Source?
What is Open Source Hardware?
Open Source in a nutshell: the Four Freedoms
CC BY-NC is not an Open Source license
Re: Yes, 3D printers can go on fire.
January 17, 2014 10:51PM
Well my printer is back up running. The extruder stepper sounds little funny, but prints are fine. Going to replace that, just in case. I had to repair 8 burnt plugs, and the ramps board. For the time being, I placed aluminum tape along the ramps board to cover the wood and act as a fire barrier. I went to the hardware store to buy a smoke alarm, and they had 2 kinds, a general one, and a more sophisticated one that is trigger by more kinds of smoke, and more sensitive. It can "see" smoke too. It was $27, so I got that going. Next will be my fire retarded shelf, once it's built I will post pictures and specifics.

Printing out wire chains now. When that's done, I will post pictures too. This thread could be a good resource for those wanting to take as many precautions as possible. I love all the ideas.
Re: Yes, 3D printers can go on fire.
January 18, 2014 10:04AM
What chains are you making?
Re: Yes, 3D printers can go on fire.
January 18, 2014 12:23PM
Smoke chains?

grinning smiley
Re: Yes, 3D printers can go on fire.
January 18, 2014 09:02PM
I found a smoke alarm on clearance for $5 at Lowes, I'll see if I can get anywhere with it.
Re: Yes, 3D printers can go on fire.
January 19, 2014 04:12AM
I have got an Aico smoke alarm on the way with an optional base unit that has a 5 amp relay built in.
Re: Yes, 3D printers can go on fire.
January 24, 2014 09:21AM
Quote
Antslake
Well my printer is back up running. The extruder stepper sounds little funny, but prints are fine. Going to replace that, just in case. I had to repair 8 burnt plugs, and the ramps board. For the time being, I placed aluminum tape along the ramps board to cover the wood and act as a fire barrier. I went to the hardware store to buy a smoke alarm, and they had 2 kinds, a general one, and a more sophisticated one that is trigger by more kinds of smoke, and more sensitive. It can "see" smoke too. It was $27, so I got that going. Next will be my fire retarded shelf, once it's built I will post pictures and specifics.

Printing out wire chains now. When that's done, I will post pictures too. This thread could be a good resource for those wanting to take as many precautions as possible. I love all the ideas.

Did you ever get the wire chain mounted? I have a makerfarm i3 as well and would love to see how you mounted it. Thanks!
Re: Yes, 3D printers can go on fire.
January 24, 2014 07:30PM
How about something like this:

[dx.com]

Inexpensive, small and shouldn't take much to wire to RAMPS board.
Re: Yes, 3D printers can go on fire.
January 25, 2014 01:29PM
LarryJ: it is exactly the same as the Chinese one I posted a link to. I actually ordered it, but it will probably be 4 weeks delivery.

I have also bought an Aico Ei141 smoke alarm and base unit that has a 5A relay output. It has nc and no contacts, so it will make wiring easier. Will report back when I wire it up and test it. Smoke alarm is ionisation type.
I prefer keeping the smoke alarm wiring separate from the printer controller. I just want it to cut the mains power.
Re: Yes, 3D printers can go on fire.
January 25, 2014 06:05PM
Ok folks lets look at the obvious here. by the time the sensor senses smoke, what good does it do to turn something off? IT IS ALREADY ON FIRE OR COMBUSTING.
at this point you need human intervention, someone to put out the fire, cool down the wires, unplug the circuitry, open up the windows and vent out, some one to calm your wife down and let you still print inside the house. I don't think any 12 volt relay kit can ever do all that! the best bet is a thermal fuse that goes off before the temp reached combustible temperatures as a FAIL SAFE to mechanical and software failures. the ones used in coffee makers are less than 2$

fire alarm is to alert someone who should be at home if you are printing, that something is going on.

Edited 2 time(s). Last edit at 01/25/2014 06:09PM by jamesdanielv.
Re: Yes, 3D printers can go on fire.
January 25, 2014 07:38PM
On the occasion I had when the thermistor was pulled out of my hotend, the print and ball of plastic charred and filled the living room with smoke. I just turned it off at the mains and it stopped right away. It is the first thing to do when electrics are involved.
I caught it early on and there were no flames. I did not need to use a fire extinguisher. Since then I have kept a smoke alarm closeby.

A fire will be progressive and steadily get worse and may eventually set fire to surroundings. The secret is early intervention. It starts with smoke.

No system is 100% failsafe, but several measures taken together will reduce the incidence. Some are easy to impliment such as a smoke alarm with a relay and also changes to the firmware.

I am always in the house when printing but not always in the same room. Some people on here print overnight which I think is risky unless maybe you do it in the garage.
Re: Yes, 3D printers can go on fire.
January 27, 2014 12:01PM
Quote
jamesdanielv
Ok folks lets look at the obvious here. by the time the sensor senses smoke, what good does it do to turn something off? IT IS ALREADY ON FIRE OR COMBUSTING.
at this point you need human intervention, someone to put out the fire, cool down the wires, unplug the circuitry, open up the windows and vent out, some one to calm your wife down and let you still print inside the house. I don't think any 12 volt relay kit can ever do all that! the best bet is a thermal fuse that goes off before the temp reached combustible temperatures as a FAIL SAFE to mechanical and software failures. the ones used in coffee makers are less than 2$

fire alarm is to alert someone who should be at home if you are printing, that something is going on.

I've had wires catch fire and go out as soon as power was removed...

You have a better idea for these situations?
Re: Yes, 3D printers can go on fire.
January 27, 2014 04:26PM
be home to catch the issue. just because an off switch is included does not mean it will make it any safer.

the best fail safes are
thermal fuses before combustible temp
fuse linkages or fuses to the power supply that blow before shorted wires have a chance to heat up.

proper wiring,


last resort choices that work after a fire is started, and in my opinion as a fail safe this is too late: the 12$ relay fire alarm kit.

Edited 1 time(s). Last edit at 01/27/2014 04:34PM by jamesdanielv.
Re: Yes, 3D printers can go on fire.
January 27, 2014 06:41PM
Quote
jamesdanielv
be home to catch the issue. just because an off switch is included does not mean it will make it any safer.

the best fail safes are
thermal fuses before combustible temp
fuse linkages or fuses to the power supply that blow before shorted wires have a chance to heat up.

proper wiring,


last resort choices that work after a fire is started, and in my opinion as a fail safe this is too late: the 12$ relay fire alarm kit.

I can not find a thermal fuse rated to sustain temps over 230c. Most components on the printer are rated to make heat, the likelihood of a fuse catching it is slim. Any other things?
Re: Yes, 3D printers can go on fire.
January 27, 2014 08:21PM
place the thermal fuse at a spot on the hot end that should normally not get that hot. the hot end normally does not touch anything but the printed part correct? (under normal conditions)

look at fuses, or fusable linkages, to stop over current and wire shorts from causing fires.

If all else, what i am saying is the best method to keep a machine safe that generates heat, be it a laser, a cnc, or a 3d printer is to have someone always nearby when it is on and running.


Don't ever assume that any protection is good enough to leave the machine unattended.
Re: Yes, 3D printers can go on fire.
January 27, 2014 09:16PM
Quote
Happycamper
Quote
Antslake
Well my printer is back up running. The extruder stepper sounds little funny, but prints are fine. Going to replace that, just in case. I had to repair 8 burnt plugs, and the ramps board. For the time being, I placed aluminum tape along the ramps board to cover the wood and act as a fire barrier. I went to the hardware store to buy a smoke alarm, and they had 2 kinds, a general one, and a more sophisticated one that is trigger by more kinds of smoke, and more sensitive. It can "see" smoke too. It was $27, so I got that going. Next will be my fire retarded shelf, once it's built I will post pictures and specifics.

Printing out wire chains now. When that's done, I will post pictures too. This thread could be a good resource for those wanting to take as many precautions as possible. I love all the ideas.

Did you ever get the wire chain mounted? I have a makerfarm i3 as well and would love to see how you mounted it. Thanks!

Not yet, I printed out the chain, but having a brain fart in trying to get it mounted. There is no real place to do it, so that means I will have to design a piece or two. Just been busy lately. I own a plumbing hvac business and with the polar vortex all I do is unfreeze houses, and fix burst pipes. For the time being I have not been letting any prints go unattended, and I have a nice smoke alarm right above the printer.
Re: Yes, 3D printers can go on fire.
January 28, 2014 05:02AM
Got my smoke alarm and relay base and wired them up. It is an Aico Ei141 ionisation smoke alarm with an Ei128R relay base unit.

Tested it with a bit of burning paper. It cuts the mains power a few seconds after the alarm goes off. However, once the alarm silences and the unit resets, the power comes back on. This is not the action required. It needs to stay off.

Only way I can think of doing it is to add another relay to make a latching circuit. Back to the drawing board.

Edited 1 time(s). Last edit at 01/28/2014 05:10AM by dave3d.
Re: Yes, 3D printers can go on fire.
January 28, 2014 03:52PM
Quote
dave3d
Got my smoke alarm and relay base and wired them up. It is an Aico Ei141 ionisation smoke alarm with an Ei128R relay base unit.

Tested it with a bit of burning paper. It cuts the mains power a few seconds after the alarm goes off. However, once the alarm silences and the unit resets, the power comes back on. This is not the action required. It needs to stay off.

Only way I can think of doing it is to add another relay to make a latching circuit. Back to the drawing board.


You can get relays that latch, and stay in the 'on' or 'off' state even after the coil has energised or de-energised... such as [uk.farnell.com], datasheet here [www.farnell.com]

a double winding latch relay has two coils - one for set, and one for reset. So no matter how many times your smoke alarm goes off or resets, your printer will stay off smiling smiley
Re: Yes, 3D printers can go on fire.
January 28, 2014 05:00PM
Thanks. I've been printing my chain links out and have been trying to figure out how to mount em too.
Re: Yes, 3D printers can go on fire.
January 28, 2014 05:53PM
Quote
TOmmm wrote
You can get relays that latch, and stay in the 'on' or 'off' state even after the coil has energised or de-energised... such as [uk.farnell.com], datasheet here [www.farnell.com]

a double winding latch relay has two coils - one for set, and one for reset. So no matter how many times your smoke alarm goes off or resets, your printer will stay off

Thanks. However, I don't wish to change the relay in the smoke alarm. That would be more trouble than it is worth. If I just add another ordinary non latching relay in the circuit I can make it latching without using a reset button. Switching the smoke alarm off and on at the mains plug will reset it. I need to find a miniature 5A relay with a 240v coil.
Re: Yes, 3D printers can go on fire.
January 28, 2014 10:14PM
Hi all,

Here is a picture of the J-Head on my MakerFarm Prusa I3. I just measured the temperature at the junction of the heater block and the PEEK temperature break/support/guide part of the j-head. My j-head is currently printing at 230C and this junction temperature measured at 165C. It's food for thought. I recently purchased some 200C thermal fuses. I plan on using them, in combination with a solid state relay, to "fuse" the AC power to the main power supply via the hot end. I will be buying some thermal fuses with a slightly lower holding temperature and I also plan on performing some additional temperature testing. Unforunately ,this is as close as I can get with the thermistor I am measuring with. It would be ideal to get a measurement near the center of the PEEK tube between the heater block and the PEEK. That's just no possible without compromising the PEEK or heater block. The space where I stick the thermistor to measure was pretty tight and was covered pretty well by the self almagamating tape I use to wrap my hotend. I also used thermal grease to get a better temperature reading.

What do you all think?
Attachments:
open | download - 20140128_215759-1200.jpg (62.6 KB)
Re: Yes, 3D printers can go on fire.
January 29, 2014 04:51AM
Positioning of the thermal fuse will be critical I think.
I have had a jhead overheat enough to melt the peek at one end. I don't know what temperature it reached but peek softens at 248 deg C, so it must have been well in excess of that. It was however melted just at one end. The rest of the jhead was unaffected so obviously relatively cool. Because of the low thermal conductivity of peek there will be a steep temperature gradient along the jhead. It might be OK in one position but move the fuse a couple of mm further along the jhead it might never fuse.

Edited 1 time(s). Last edit at 01/29/2014 09:41AM by dave3d.
Re: Yes, 3D printers can go on fire.
January 29, 2014 09:32AM
That's why I measured right at the junction of the PEEK and the heater block. Under normal (my normal) operations, my printer prints at 230C registered on the thermistor. That temperature doesn't mean much when it comes to the thermal fusing. I measured the junction and it measures at approximately 165C. I will probably get a thermal fuse rated for 180C holding. My goal is to prevent a fire. I don't necessarily care about a J-head melting down so long as nothing around the printer is damaged. I'm looking for protection against catastrophe. So, on my machine, the point at which I measured the temperature is 165, if it rises another 15, then the thermal fuse should cut out the power in my design and save me from catastrophe.

Ultimately, I'd like to sacrifice a J-head for true peace of mind. In otherwords, do some practical testing and bring the temps up to melt to see at just what temperature, in my setup, they begin to melt to ensure the setup works but that will have to wait until I have a few hundred bucks burning a hole in my pocket.

In response to an earlier comments
Quote

I can not find a thermal fuse rated to sustain temps over 230c. Most components on the printer are rated to make heat, the likelihood of a fuse catching it is slim. Any other things?

You don't need a thermal fuse rated for the temperature of the hotend. You only need one rated for the location that it's in. In my case, the location that I will be placing it in very close to the hot end but still cool enough to use off the shelf components and prevent catastrophe (FIRE).

Edited 1 time(s). Last edit at 01/29/2014 09:36AM by umdpru.
Re: Yes, 3D printers can go on fire.
January 29, 2014 11:35AM
On the topic of cable chains.

I don't really think they are necessary if you are organizing your wire bundle correctly. I think the MF I3 needs to have either the X-idler or X-motor printed/redeisgned to have a proper cable chain mount made available.

Here is a picture of my printer.

I bundled the wires and routed them through braided sleeving that I got from cable organizer a long time ago. Use shrink wrap on the ends to make it neat.

I zip tied one end around the motor and the other end to the X-idler carriage. The result is that the wire bundle makes a wide and lazy arc to the x-carriage as it moves back and forth. There is plenty of strain relief and the wiring is well managed. There is a short loop of wiring from the x-carriage to the Ramps board.

Wire management on a device like this is critical. I just don't htink the small size of these warrants the use of bulky cable chains.
Attachments:
open | download - 20140118_162448-1200.jpg (95.6 KB)
Re: Yes, 3D printers can go on fire.
January 29, 2014 03:16PM
Quote
umdpru
That's why I measured right at the junction of the PEEK and the heater block. Under normal (my normal) operations, my printer prints at 230C registered on the thermistor. That temperature doesn't mean much when it comes to the thermal fusing. I measured the junction and it measures at approximately 165C. I will probably get a thermal fuse rated for 180C holding. My goal is to prevent a fire. I don't necessarily care about a J-head melting down so long as nothing around the printer is damaged. I'm looking for protection against catastrophe. So, on my machine, the point at which I measured the temperature is 165, if it rises another 15, then the thermal fuse should cut out the power in my design and save me from catastrophe.

Ultimately, I'd like to sacrifice a J-head for true peace of mind. In otherwords, do some practical testing and bring the temps up to melt to see at just what temperature, in my setup, they begin to melt to ensure the setup works but that will have to wait until I have a few hundred bucks burning a hole in my pocket.

In response to an earlier comments
Quote

I can not find a thermal fuse rated to sustain temps over 230c. Most components on the printer are rated to make heat, the likelihood of a fuse catching it is slim. Any other things?

You don't need a thermal fuse rated for the temperature of the hotend. You only need one rated for the location that it's in. In my case, the location that I will be placing it in very close to the hot end but still cool enough to use off the shelf components and prevent catastrophe (FIRE).

You now face another issue, the trip temp is usually 20-30C above the holding temp...the Digikey fuses that TRIP at 180C are only rated to sustain 160C so you will need to take that into consideration. I still don't see a thermal fuse as the way to go, my E3D ran up to around 400-500C I am guessing (from the crispy kapton) and the PLA fan shroud that was 2-3mm away from the heater block on the heat sink was not affected at all. There are many things to catch fire below the fan shroud like thermistor (which I no longer use) lead covers, cartridge/resistor, kapton, and possibly the filament itself.

Too many variables I see with the thermal fuse, most hotends overshoot if you go from PLA to say PC which is nearly 100C different temps, even PLA to ABS can be steep because I can print some PLA at 170 and I usually run ABS at 240-250 so that is even a 70-80C swing at temps a J-Head can run. You can't autotune for everything, and the overshoot may be enough to trip the fuse. Being $1 a piece, not something I would want to replace all the time.

What if a fan dies, doesn't necessarily mean the hotend is going to melt but the temp may run up temporarily until the PID can catch up and you just tossed a print because the fuse cut out.

Like another said, the thermal gradients are too steep to see any benefits from a fuse I think. You may not even hit the trip temp at the fuse before the peek melts or catches fire, move it close and now you can suffer from all issues I just covered. I'm pretty sure I can hold the heat sink in my hand with the hotend at 400-500c on my E3D, so where do I put the fuse at then?

Edited 1 time(s). Last edit at 01/29/2014 03:17PM by tjb1.
Re: Yes, 3D printers can go on fire.
January 29, 2014 05:30PM
Another solution to fire detection I would like to explore is the use of a thermal imaging camera. They are widely used in fixed commercial installations to detect hot spots in warehouses etc. The emergency services also use portable units. It is a well proven method but the fixed systems I have seen have been expensive.

I do not know if the cost can be brought down for amateur use. Ideally using an ordinary webcam and free software.

Anyone care to contribute?
Sorry, only registered users may post in this forum.

Click here to login