[Blabber] Fire and 3D printers

John O'Brien jweob at cantab.net
Sat May 19 16:50:40 UTC 2018


Very late response to this.

My meltdown happened when the board was in a completely unresponsive state,
I think is  because the ground connection for the Melzi's power supply came
loose. I therefore don't want to rely on any firmware running on the main
board in future.  In response to my meltdown I put an independent arduino
on my printer which under certain conditions will cut off all power to the
printer using a relay. The secondary arduino has a secondary thermistor,
but James as you say this doesn't help if the heater cartridge falls out of
the block. I tried to protect against the case where the hotend falls out
of the block and so is no longer thermally coupled to the thermistor by
using the secondary arduino to monitor the voltage across the heater
cartridge. If the heater cartridge is on continuously for more than a
minute the safety system will trigger and cut off power to the printer (one
minute is slightly more than is needed to bring it to 200C from a cold
start).

There's a description and video of the safety system here:
http://jweoblog.com/?p=835


On Sat, Mar 17, 2018 at 3:24 PM, James Carpino via Blabber <
blabber at list.hackmanhattan.com> wrote:

> missed some words, edit:
> "The cartridge can get hot enough to MELT ALUMINUM as John O'Brien has
> shown us in a previous post."
> http://jweoblog.com/?p=831
>
> see previous.
>
> ------------------------------
> *From:* James Carpino <jamescarpino at yahoo.com>
> *To:* Hack Manhattan! <blabber at list.hackmanhattan.com>
> *Sent:* Saturday, March 17, 2018 3:22 PM
> *Subject:* Fire and 3D printers
>
> Just to make sure everyone understands that 3d printers can catch fire,
> here is another good account;
>
> https://www.thissmarthouse.net/dont-burn-your-house-down-
> 3d-printing-a-cautionary-tale/
>
> This came up in one of the 3DP facebook groups.
>
> I've given considerable thought to the scenario where the heater cartridge
> graually slips out of the aluminum block.  I'll abstain from a full
> treatment here, but basically:
> - Regular use leads to gradual loosening of the setscrew until it is ripe
> for failure, in this case a gradual escape
> - Coupling with the block is changing slowly
> - PID loop is adjusting, and increasing the heater cartridge temperature
> to compensate for the weak connection.
> - The thermistor is in the aluminum block, not in the cartrdige.
> - The cartridge can get hot enough to as John O'Brien has shown us in a
> previous post.
> http://jweoblog.com/?p=831
> - Thermal runaway protection in the firmware WILL NOT kick in, as there is
> no apparent thermal runaway.
> - Extreme temperatures cause fire of whatever the cartridge lands on
>
> It's time for someone to demonstrate this under controlled conditions but
> I haven't found the time myself, and I'm considering it as a back-burner
> project.  But I definitely see that this is a distinct possibility and this
> is the 2nd probable case of cartridge overheating/fire that I've seen
> posted on 3DP groups.  It's different from what John O'Brien experienced,
> and John I'm contemplating whether your device would catch this form of
> failure.  This is not exactly thermal runaway if thermal means "measured
> temperature," from the perspective of the controller I mean, it's rather
> "on-time"- or "heat-input"-runaway.  It would seem to me that some
> awareness of this could be written into Marlin.  Major changes in the duty
> cycle (on-time) needed for a given temperature, would be a red flag.
>
> The printer could theoretically stop before the hotend falls out; leave
> all fans running and turn off all heat sources.  I'm not sure whether to
> leave fans running but shutting down the whole printer shuts them off so
> that has to be considered carefully.  The trade-off is damage to the
> printer's remaining parts vs. damage to the environment.
>
> James
>
>
>
>
> _______________________________________________
> Blabber mailing list Blabber at list.hackmanhattan.com
> https://list.hackmanhattan.com/listinfo/blabber
>
> Unsubscribe: https://list.hackmanhattan.com/options/blabber or send a
> blank email to blabber-unsubscribe at list.hackmanhattan.com
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://list.hackmanhattan.com/pipermail/blabber/attachments/20180519/07a59194/attachment.html>


More information about the Blabber mailing list