Fire Hazard

Those stove Range top Fire suppressors would probably work as a last line of defense, but the an electrical cutoff still needs to happen. A redundent thermistor and maybe a current sensor could be employed in determining failures to cut the power.

I saw someone suggested it... but there probably should be some kind of a sub forum to address safety... I could see this turning into several threads as people figure out different ways to address these problems. It would be better to have them in one section. IMO

Quote
Hazer
And all of this does not cover the electrical fire potential. I have seen some people place extra thermal cutoffs around the electronics, but they would not trip out until the fire has already ignited. So what does cutting power do when the flames are already there?

I guess the point (as I see it) is that, even though a fire has started, removing the power removes any further heat build-up from that source. A fire *may* extinguish itself at that point - maybe not. But it doesn't hurt.

That's why I am going further with the extinguishing aspect of my plan.

Edited 1 time(s). Last edit at 07/03/2014 09:21PM by andy.wpg.

---

"I have noticed that even those who assert that everything is predestined and that
we can change nothing about it still look both ways before they cross the street"

-Stephen Hawking

Quote
Hazer
The point I have been trying to make is that you need more than just temperature-cutoff devices in order to eliminate the potential for fire. As mentioned, if you do not have the thermal cut-off directly on your heat block, then you run the risk of your hotend dropping out of its plastic fixture before the thermal cut-off does its job. And even then, there is no guarantee that the thermal cutoff will trip BEFORE the ignition occurs (and if your already on fire, who cares about disbaling the power to yuor hotend). You can only solve thermal runaway if your trip device is designed for the media you are heating in the first place.

In the case of thermal cut-offs in appliances, you forget that the high voltage heater element is attached through metal fixtures that wick away the heat by design and that is what the thermal cut-off is attached to. It is directly coupled to the secondary heated material. Now you could say by attaching a thermal cutoff to the top of your all-metal hotend would work, except 99% of those designs have a fan blowing on that portion of the hotend, and thermal runaway would still occur.

And all of this does not cover the electrical fire potential. I have seen some people place extra thermal cutoffs around the electronics, but they would not trip out until the fire has already ignited. So what does cutting power do when the flames are already there?

I've had wiring catch fire (not on a printer) that went out when the power was removed. The wiring was providing heat required to burn, without it, it went out.

I had wire explode one time..
Hah, that's a short that I quickly and over effectively removed and prevented. Apparently the wire can take enough current to explode before my power supply trips and shuts off [and it has done so in other curcumstances, so I know it's there, Seasonic would not miss something like that]. Heat didn't go much furthur than the area at the start of the wire and return of the other pole. Where it actually shorted was hardly warm. So even a thermal fuse at the point of the short would not prevent heat.

---

Realizer- One who realizes dreams by making them a reality either by possibility or by completion. Also creating or renewing hopes of dreams.
"keep in mind, even the best printer can not print with the best filament if the user is the problem." -Ohmarinus

Quote

As mentioned, if you do not have the thermal cut-off directly on your heat block, then you run the risk of your hotend dropping out of its plastic fixture before the thermal cut-off does its job.

You're assuming a thermal cut-off would be improperly chosen such that it stayed intact and let hte hotend fall out.

I think you either didn't read what I posted or just dind't understand it.

the point in what I posted and the main design goal is to stop the melting part.... :-/

Quote

And all of this does not cover the electrical fire potential. I have seen some people place extra thermal cutoffs around the electronics, but they would not trip out until the fire has already ignited. So what does cutting power do when the flames are already there?

I guess you missed the part about the automatic fire bottle above your work bench? I have MUCH greater peace of mind since I installed one along with fire alarms right above my printing bench.

Quote

you forget that the high voltage heater element is attached through metal fixtures that wick away the heat by design and that is what the thermal cut-off is attached to.

THAT is entirely inaccurate! You're telling me that my stove's metal heating element is connected to structure on the stove that is DESIGNED to wick heat away from the element into the stove? That's an absolutely assinine assumption and absolutely wrong. The efficiency of heating applicances would suffer catastrophically if they were required to heat the appliance AND the medium needing heating (my food). Sorry to be so blunt but you are way out in left field on that one.

the bottom line here is that if you shut off power to your hot end it will begin to cool. They aren't nuclear reactors with latent heat buildup issues after power shutdown. If you catch the runaway at the proper time, you can prevent a catastrophic accident - period. Beyond that, use an automatic fire suppression system.

I suggest no one go into this blindly and some testing should be done on the most vulnerable hot-ends.

Lastly, no amount of safety mechanisms are going to stop everything. If you implement everything then your costs will go beyond your reach.

Edited 2 time(s). Last edit at 07/03/2014 04:55PM by umdpru.

The simple solution is to use a 20W heater instead of a 40W heater. If it is 100% on the hot end gets to about 300C and smokes a bit. The insulator might be damaged. If it has PTFE it might kill your pet fish or give you flue like symptoms. Other than that it is fail safe.

40W heaters are dangerous. If they are permanently on they get to about 600C, melt aluminium and set fire to plastic.

Heated beds that use a PCB are fail safe because the resistance of the copper increases with temperature and self limit. Constant resistance types should incorporate a thermal fuse.

Other than that all electrical devices have a small chance of starting a fire. Think wire wool on a 3V lithium cell. So you should always have smoke alarms and a fire extinguisher in your house to protect you if it catches fire when you are in. If you want to be absolutely sure it doesn't burn down when you are out turn everything off, or take a chance.

---

[www.hydraraptor.blogspot.com]

I like that solution, Nophead. A 40W cartridge is certainly overkill. My J-heads come to 235C in a a few minutes at most. I can wait 10 or 15 for them to heat up. the heatbed takes that long anyways.

No amount of safety prevention will prevent all disasters with these machines. they aren't for the ignorant to use.

Quote
nophead
The simple solution is to use a 20W heater instead of a 40W heater. If it is 100% on the hot end gets to about 300C and smokes a bit. The insulator might be damaged. If it has PTFE it might kill your pet fish or give you flue like symptoms. Other than that it is fail safe.

I object. No one has scientific experimentation to support this. There are a bunch of factors that determine how hot a hot end can get. Air flow, insulation, ambient temperature, etc. I'm going to have to kill a few hot ends tommorrow, for the sake of science!

Quote
umdpru

I guess you missed the part about the automatic fire bottle above your work bench? I have MUCH greater peace of mind since I installed one along with fire alarms right above my printing bench.

No, I did not. You forget I mentioned an automatic fire extinguisher first. I am glad you have one, as the point I am trying to make is that no matter what people come up with, the only way you should leave these printers running unattended (as in you leave the house or are asleep) should require that extinguisher.

Quote
umdpru

Quote

you forget that the high voltage heater element is attached through metal fixtures that wick away the heat by design and that is what the thermal cut-off is attached to.

THAT is entirely inaccurate! You're telling me that my stove's metal heating element is connected to structure on the stove that is DESIGNED to wick heat away from the element into the stove? That's an absolutely assinine assumption and absolutely wrong. The efficiency of heating appliances would suffer catastrophically if they were required to heat the appliance AND the medium needing heating (my food). Sorry to be so blunt but you are way out in left field on that one.

Ovens are all metal with insulation. I am talking about smaller appliances that have thermal fuses inside them. Like toasters. I have ripped quite a few of these apart. The have plastic components, but most of the plastic used in appliances are non-ignitable. So, the heating elements in those appliances are generally attached to the sheet metal parts and the thermal fuse is located near the mounting. There is no ceramic insulator keeping the heat from passing from the element to the structural body. Thus, if the sheet metal heats up beyond the design of the appliance, it trips before the plastic gets effected. This is common practice to pass UL testing. Most 3D printers (all DIY ones) are made from ignitable plastic. Most actually use this plastic to hold the hotend. Its a fuel, and self-sustaining once ignited. Take a lighter to the end of your filament and see.

---

"Never argue with stupid people, they will drag you down to their level and then beat you with experience."

It's late and I'm almost out, but before I forget;
What is this? What do suppression capacitors do? It was in my subwoofer that I'm repairing and I looked up the numbers. For safety apparently. "Active and passive flame retardant."

---

Realizer- One who realizes dreams by making them a reality either by possibility or by completion. Also creating or renewing hopes of dreams.
"keep in mind, even the best printer can not print with the best filament if the user is the problem." -Ohmarinus

Searched for Suppression capacitor, pulled this up...

Safety and EMI/RFI suppression film capacitors
[en.wikipedia.org]


EMI suppression capacitors are used to suppress any noise from an electronic device by reducing the input impedance of the device.

These capacitors are divided into 2 classifications, X and Y.

X class capacitors are capacitors that are connected line to line and in the event of failure of the capacitor the potential for electrical shock is not present.

Y capacitors on the other hand are connected from line to ground. They are typically a very low capacitance value. In the event a Y capacitor fails the potential for electrical shock is present.

[www.illinoiscapacitor.com]

Quote
umdpru
I like that solution, Nophead. A 40W cartridge is certainly overkill. My J-heads come to 235C in a a few minutes at most. I can wait 10 or 15 for them to heat up. the heatbed takes that long anyways.

The OP has a printer with thee hotends. Slowing down the warm-up time by the figures you suggested is a functionality killer because it would take much longer to complete a modest sized multi-color print. That would be bad for the kit vendor's marketing I suppose...

I wonder if the OP has contacted the kit vendor to get their input on what safety mechanisms they would recommend?

Andy

What I'm reading into all of this dialogue is the following...

1) As much equipment as possible above the hot bed level should be made from metal or at least something non-flammable. I plan to redesign my hot end holder to be made of aluminum and move the filament drive assemblies to the side if possible, with the filament spools. As you can see from my photos of the fire, the X cartridge and the filament drive assemblies were completely involved in the meltdown. The spool holders were off to the back but had already begun to ignite when I found it.

2) I plan to experiment with various thermal fuses and mounting locations to find a combination that will work for my design to trigger a power cutoff before ignition.

3) I will be looking into the use of 20 watt heaters instead of the 40 watt currently used. The 40 watt units heat so fast that I'm always waiting for the hot bed to reach temp.

4) I had been using thermistors that plug into the heater block from one side. This generates a high potential for it pulling out and causing a thermal runaway. I'm switching to the radial type with a wire sticking out of each side of the block and using PTFE casings on both leads. I believe I can then safely secure the wires to prevent dislodging.

5) Depending upon my success at item 1 above, I may or may not need to install thermal shielding above the nozzle.

6) In spite of all of the above precautions,
- I know I can never leave the printer running unattended, and..
- I need to understand that it will still have the potential to catch on fire so I will try to design in some fire suppression features like automatic fire extinguishing. Enclosing it in a ventilated, smoke detector monitored, metal box is also an option.

I'll let you know how things go. I want to thank all who provided feedback and suggestions. You have been a great help. My wish is that others will also benefit from your ideas.

Gonetotx

has anyone considered enclosing the machine in an airtight container then replacing the air inside said container with an inert gas, say CO2 or N2? If there is no oxygen and your hotend does go in to runaway the CO2 should starve it of oxygen and thus a fire should not be able to form. the enclosure should be doable with sheets of glass, polycarbonate, acrylic, hell any sheet of material you can get a good airtight seal out of, and silicon epoxy or caulk, maybe some metal or plastic braces to help it hold its shape. When it comes time to replace the air inside with CO2 a check valve on an outlet port would allow it to only flow out, same deal on your inlet for CO2, check valve just faces the other direction. A big plus is that CO2 is not expensive, and should be available anywhere they refill CO2 tanks for paintball/airsoft guns. The only big problem i can see is how do you know when you have replaced all of the air inside with your inert gas of choice. A candle could be lit inside of the chamber but if your inert gas is heavier than air then it would have to be as close to the top of the chamber as possible, also if the atmosphere replacement process causes drafts inside the chamber then you might see the candle go out when you havent finished filling the chamber. As far as any sort of "heated build enclosure" or "draft free build environment" this isnt meant to be either of those, this is purely a safety mechanism, it should be fair game as far as patents but patent law is not my specialty so who knows. This is just my thought on the matter, by no means am I an expert in matters of fire safety or prevention.

Edit: I also realize this does nothing for shutting off the power to the hotend but it could definitely be combined with a method for that thermal fuses, circuit breakers, etc.

Edited 1 time(s). Last edit at 07/09/2014 10:43PM by JMoses.

Removing each print would require a refill on the volume with the gas. That's a high cost/safety ratio in my mind.

---

Realizer- One who realizes dreams by making them a reality either by possibility or by completion. Also creating or renewing hopes of dreams.
"keep in mind, even the best printer can not print with the best filament if the user is the problem." -Ohmarinus

Seems to me that some kind of secure mechanism to lock the heating element and thermistor in place and a redundant thermistor should be things that are looked into along with upgrading firmware and implement watchdog/thermal runaway fail safes. An appropriately sized fuse that would fail if the hotend would draw max amps for too long might help. Then Fire / Smoke Alarms and one of those Range Top Fire extinguishers.

Has anybody tested one of the all metal hot ends (which presumably conduct away a bit more heat than a PEEK hotend) with a 20W heater as @nophead suggests? If we can still get to 315C then this just seams like a no brainer. I would like to be able to print at 310C to print PC but I would very much like not to be able to print at 600C! Intrinsically safe just seems like a good idea.

Quote
ElectroWomble
Has anybody tested one of the all metal hot ends (which presumably conduct away a bit more heat than a PEEK hotend) with a 20W heater as @nophead suggests? If we can still get to 315C then this just seams like a no brainer. I would like to be able to print at 310C to print PC but I would very much like not to be able to print at 600C! Intrinsically safe just seems like a good idea.

this is on my list of things to test, along with power supplies and other components which could be dangerous

i have just done a 40watt heater cartridge on a 12v supply [see here]

---

-=( blog )=- -=( thingiverse )=- -=( 3Dindustries )=- -=( Aluhotend - mostly metal hotend)=--=( Facebook )=-

I think that SeeMeCNC's hotend heater block is interesting:

http://seemecnc.com/collections/parts-accesories/products/seemecnc-bowden-hotend

Does using two heater resistors speed up the warmup process without using a 40W cartridge heater? Would it allow for 300C+ prints?????

the resistors they sell are 6.8ohms so each one is a 21watt heater on it's own, with two you've got 40watts , ultimately i wouldn't expect much different heating behaviour

---

-=( blog )=- -=( thingiverse )=- -=( 3Dindustries )=- -=( Aluhotend - mostly metal hotend)=--=( Facebook )=-

Quote
thejollygrimreaper
the resistors they sell are 6.8ohms so each one is a 21watt heater on it's own, with two you've got 40watts , ultimately i wouldn't expect much different heating behaviour

Would there be any safety benefit to modified electronics and firmware that have separate circuits for each resistor, and only use one of the two for warming up? Ie: shutting the second resistor down once the hotend is in the PID loop?????

Quote
vreihen

Quote
thejollygrimreaper
the resistors they sell are 6.8ohms so each one is a 21watt heater on it's own, with two you've got 40watts , ultimately i wouldn't expect much different heating behaviour

Would there be any safety benefit to modified electronics and firmware that have separate circuits for each resistor, and only use one of the two for warming up? Ie: shutting the second resistor down once the hotend is in the PID loop?????

it would be dependant on a lot of factors, there are so many different ways hotends get setup , it is possible to heat a hotend to 500c even with just a 21watt heater, it just depends on how it has been setup,

I'm looking into a fail-safe switch which in the event of a thermal runaway will shut down the power before it gets to a dangerous temperature where it can cause a fire

---

-=( blog )=- -=( thingiverse )=- -=( 3Dindustries )=- -=( Aluhotend - mostly metal hotend)=--=( Facebook )=-

I have been thinking about this since this thread started.

What about putting a second thermister in the heater block (just drill another hole) run it to an Arduino that is powered seperatly (wall wart) and then set it up to turn on a relay for the mains power to the machine as long as the temperature is below a set point (say 300c). This would assure that if something happens to the heat or to the Arduino that the power would be cut to the printer.

I checked with several manufacturers of thermal fuses but none have high enough temperatures for our use.

So for the cost of an Arduino, thermister, relay and wall wart (say $50 or less) you could have a way to power off the machine in case of thermal runaway.

I am not a programmer but I would think the program for the Arduino would be dead simple.

Quote
tmorris9
I have been thinking about this since this thread started.

What about putting a second thermister in the heater block (just drill another hole) run it to an Arduino that is powered seperatly (wall wart) and then set it up to turn on a relay for the mains power to the machine as long as the temperature is below a set point (say 300c). This would assure that if something happens to the heat or to the Arduino that the power would be cut to the printer.

I checked with several manufacturers of thermal fuses but none have high enough temperatures for our use.

So for the cost of an Arduino, thermister, relay and wall wart (say $50 or less) you could have a way to power off the machine in case of thermal runaway.

I am not a programmer but I would think the program for the Arduino would be dead simple.

the program would be very easy and quick to do, i would do it on a small pic chip or a small avr chip as opposed to a full blown arduino

i've been looking at putting very specifically chosen fuses in line with the hotend so that the large current draw blows the fuse when the mosfet fails but not during normal operation, what is interesting though is that most of the automotive fuses are only rated for a service life of 100 hours at 100% of their rating but for 45c a piece it's a cheap disposable option

Edited 1 time(s). Last edit at 07/28/2014 06:35PM by thejollygrimreaper.

---

-=( blog )=- -=( thingiverse )=- -=( 3Dindustries )=- -=( Aluhotend - mostly metal hotend)=--=( Facebook )=-

Quote
thejollygrimreaper

Quote
tmorris9
I have been thinking about this since this thread started.

What about putting a second thermister in the heater block (just drill another hole) run it to an Arduino that is powered seperatly (wall wart) and then set it up to turn on a relay for the mains power to the machine as long as the temperature is below a set point (say 300c). This would assure that if something happens to the heat or to the Arduino that the power would be cut to the printer.

I checked with several manufacturers of thermal fuses but none have high enough temperatures for our use.

So for the cost of an Arduino, thermister, relay and wall wart (say $50 or less) you could have a way to power off the machine in case of thermal runaway.

I am not a programmer but I would think the program for the Arduino would be dead simple.

the program would be very easy and quick to do, i would do it on a small pic chip or a small avr chip as opposed to a full blown arduino

i've been looking at putting very specifically chosen fuses in line with the hotend so that the large current draw blows the fuse when the mosfet fails but not during normal operation, what is interesting though is that most of the automotive fuses are only rated for a service life of 100 hours at 100% of their rating but for 45c a piece it's a cheap disposable option

Interesting idea. I would use the ATC type fuse versus the older glass ones. The ATC fuse link and tabs are cut from one bit of metal where the old glass fuses had a fuse link that was soldered to the 2 end caps. More than once I have seen them get hot and unsolder the ends before the link failed.

So when heating up from cold is the voltage still pulsed? If not, not sure a proper fuse would hold.

I have been reading about this topic and doing my own thinking about using a proper thermal fuse on the hot end. As of now I have a j-head hot end with peek. I will be designing a larger all metal hot end soon enough. One of the things that interests me is the thermal fuse solution. I have seen this applied many times before. I think some testing is necessary to assure the right measures. The idea is to use a fairly poor conductor as a sandwich for the thermal fuse. Stainless steel sheet seems right. Then wrap the fuse in the stainless on one end and have it stick out like a blade on the other. The stainless should be insulated to avoid environmental effects. This can control the heat reaching to the fuse. We can experiment with a spare cartridges and some thermal fuses of different ratings until a sweet spot is reached.

1 - Thermal fuse option (100-150c) seems appropriate.
2 - Stainless steel of set dimensions
3 - flame proof insulator
4 - thermal fuse should be connected directly to the hot end. No signal wires going out through a moving head. This way if a wire breaks it just kills the hot end power.

Anyone care to comment on this view?

---

My Personal Blog. Build blog.
[engineerd3d.ddns.net]

Modicum V1 sold on e-bay user jaguarking11

Have a simple idea that I'd love to try out, but would like to get some comments first.

If a falling out or faulty thermistor is the top reason in causing fire, can I just add a second thermistor to the heating block of the hotend and monitor the temperature? If the temperature exceeds a preset (indicating a thermal runaway), heating and printing would stop. Would like to some of your comments on the feasibility and effort to do so.

Quote
chngyian
Have a simple idea that I'd love to try out, but would like to get some comments first.

If a falling out or faulty thermistor is the top reason in causing fire, can I just add a second thermistor to the heating block of the hotend and monitor the temperature? If the temperature exceeds a preset (indicating a thermal runaway), heating and printing would stop. Would like to some of your comments on the feasibility and effort to do so.

I think the bigger failure rate is a shorted mosfet so the hotend gets full uncontrolled voltage.

Quote
jaguarking11
I have been reading about this topic and doing my own thinking about using a proper thermal fuse on the hot end. As of now I have a j-head hot end with peek. I will be designing a larger all metal hot end soon enough. One of the things that interests me is the thermal fuse solution. I have seen this applied many times before. I think some testing is necessary to assure the right measures. The idea is to use a fairly poor conductor as a sandwich for the thermal fuse. Stainless steel sheet seems right. Then wrap the fuse in the stainless on one end and have it stick out like a blade on the other. The stainless should be insulated to avoid environmental effects. This can control the heat reaching to the fuse. We can experiment with a spare cartridges and some thermal fuses of different ratings until a sweet spot is reached.

1 - Thermal fuse option (100-150c) seems appropriate.
2 - Stainless steel of set dimensions
3 - flame proof insulator
4 - thermal fuse should be connected directly to the hot end. No signal wires going out through a moving head. This way if a wire breaks it just kills the hot end power.

Anyone care to comment on this view?

I print with ABS so I would need a thermal fuse around 300c and they don't exist. I would think if you print with PLA you would need 250c or so (also does not exist).

We already use a thermal break between block and filament feed. Why not use another thermal block between fuse and block is the point. 300c thermal fuses exist. It's just that some cost more than the hot end. Look at thermal fuses meant for gas fired furnaces. They are made of a ceramic something or other. However even those are using a thermal spacer these days. Why use a 150 dollar part when with 10cent worth of steel and a 1 dollar fuse do the same job.

Edited 2 time(s). Last edit at 08/09/2014 02:24AM by jaguarking11.

---

My Personal Blog. Build blog.
[engineerd3d.ddns.net]

Modicum V1 sold on e-bay user jaguarking11