Heat of the night

[Jeroen3]

However, 3D printing seems to require NASA level tinkering skills! I’m surprised that any noob has ever printed anything!

It does not have to require tinker skills. When I bought my mk3s I only had to assembly it following the lego-technic level assembly manual. Follow the firmware's setup instructions, and it printed properly first try.

Designing things that print properly is a different game.

Most printer fires I've read about are due to bad connections, not properly tightened screws, bad crimps, or just cheap connectors or worn wires. (the thing is moving)
Typical stuff even your blender can kitchen oven fire by, except your oven had tertiary protection by thermal fuses. But even ultimaker has some reports of molten pcb terminals on the web.
The entire debacle started with, iirc, ANET printers catching fire more than incidental. Then there were concerns, because the open source firmware most printers work on didn't have thermal protections enabled by default. And many cheap printers just had the default settings, because why put in effort.

If you want a hobby with a high level of tinkering, get into CNC machining.

[janoc]

I would like to able to print some useful stuff but I also like to “tinker” with things so that I understand them better.

However, 3D printing seems to require NASA level tinkering skills! I’m surprised that any noob has ever printed anything!

Nonsense. However, a 3D printer is a machine tool. The same as a lathe or a CNC mill. So yes, you do need to know what you are doing, how to do it safely and how to adjust and maintain your machine. The cheaper the machine is the more "tinkering" and maintenance is usually going to be required.

If you want the "push-button-and-done" experience, then you need to order your parts 3D printed online.

BTW has anyone ever had or even heard of actual fires started by them?

Plenty of reports around. Did you try to Google actually?

https://www.reddit.com/r/3Dprinting/comments/3lv5ld/heres_a_reminder_to_not_leave_your_printer/

or the infamous Anet 8:

https://www.fabbaloo.com/2018/12/3d-printer-safety-another-anet-a8-burns

https://www.thissmarthouse.net/dont-burn-your-house-down-3d-printing-a-cautionary-tale/

Etc.

So yes, 3D printers do catch fire, especially the cheap crap where every corner has been cut - e.g. the Anet 8 had a very poor heated bed connection and lacked the thermal runaway protection in the firmware. Even today you can still buy machines that have the same flaws!

But even the expensive machines had some spectacular failures.

[thm_w]

For tinkering, a 24V heated bed might be safest, if you don't want to be worrying about mains power.
Ender 3 is 220W (9A), which seems reasonable.

[PerranOak]

Cheers all.

Yes, the tinkering comment was tongue-in-cheek. Anyway, I like to tinker, I'm just surprised that so many people seem to have them; they can't all be tinkerers, surely?

My purpose in asking about fires here is that I trust this community and wanted this community's opinion. Googling gives every kind of response to every kind of question: my aunt (newish silver-surfer) has Google-diagnosed herself with every kind of cancer under the sun and is still fit as a fiddle!

[Whales]

My first and only 3d printer is cheap kit:  Geeetech i3 for about $200 AUD.

The printer took two whole days of work to assemble.  I had to fill in the gaps of the instruction manual and had to work out how to tweak many, many things myself to get it to a good working state.  Some parts were missing (lubricants, decent washers, etc), others were incorrect (spade connectors for connecting to limit switches).

It now prints beautifully and happily at 1/2 the total print time of my boss' multi-thousand-dollar 3d printer and has been doing so almost uninterrupted for a year, barring me replacing the hotend recently because the PTFE in the old one had swollen.

I print with a max speed of 100mm/s in PLA, 0.4mm nozzle and 0.25mm layer height.  For some PETG models I was easily able to up it to 150mm/s max linear motion.  Those vorons shown previously can do 1,000mm/s alas that doesn't mean 10x the speed of mine as corners are still much slower, but it's still much more impressive.

My print bed is masking tape (on glass).  It works really well (too well for PETG sometimes) and it's soft so it increases your tolerance for non-level beds.  I used to have to level my bed all of the time, until I discovered (1) how to quickly & reliably straighten the Z axis on its two steppers and (2) mounting it on my shelves/wall instead of the bench keeps it from changing shape.  Now I re-level the bed once every couple of months.

I do NOT recommend getting a cheap printer like this if you are not mechanically minded and do not have lots of spare time.  My printer is very much a hobby to use, not an off-the-shelf tool like a dropsaw or drill press.  If you want a tool that's already in working condition, and is more likely to stay working without random fiddling, then you're probably looking at spending around $1000 on a 3D printer.  I have not used one, but I've heard that the Prusa i3 series is reasonably reliable and predictable. 

I have not had it catch fire yet.  At the end of the day the hotend in any printer is basically a soldering iron, so you can't really leave it unattended. Marlin's hotend runaway detection seems really good (good thing too, I tried running it once with the heating cartridge uninstalled).  I have however had plastic char on the outside surface of the hotend and I don't trust the power supply with my life.

[Whales]

If you're not up to spending the money on an expensive one and you're not up for the tinkering on a cheaper one: see if you have any local Universities that run free or open hackerspaces.  Often they let you use their (pre-calibrated) printers.

As other people have mentioned: there are services for ordering 3d printed parts.  I had OK luck last time I tried them, but I can never make a part perfect on the first try, so having a machine within arm's reach is super useful.

[PerranOak]

Sorry, I’ve given the wrong impression. I do like tinkering it’s just that it sometimes seems like you have to be a genius just to get one of those little boats printed; I’m probably exaggerating.

I did see a YouTube vid where the guy had a mains relay device (110V unfortunately) in place but I can’t seem to find one - a fully complete one rather than the actual component relay.

[Mechatrommer]

Dont put or use combustible/flammable material near or in your 3d printer. You can make external monitoring circuit to cutoff mains supply in case strange thermal happens.. etc etc... if one thinks its nasa level tinkering is because he is not interested in mechanical or electrical field..nobody can help him with that.. car burned up, are you afraid to buy car?.. matches can burn house, are you afraid to buy matches? Car can hit you while you walk outside, are you going just to stay in house? Its a risk management issue if you want to do anything. And to  manage it, you need knowledge. Dont want to take risk or manage it? Pay premium or die in your sleep, no offense my friend

[PerranOak]

None taken. 

[janoc]

I have not had it catch fire yet.  At the end of the day the hotend in any printer is basically a soldering iron, so you can't really leave it unattended. Marlin's hotend runaway detection seems really good (good thing too, I tried running it once with the heating cartridge uninstalled).  I have however had plastic char on the outside surface of the hotend and I don't trust the power supply with my life.

I am just in the process of fixing that sort of issue on my Mendel90 too - the fan shroud for cooling the hotend and the part (Mendel has only one fan for both things) has cracked and the back was touching the hotend. So it is nicely melted and charred now and I have found that only by accident while doing some unrelated periodic maintenance (cleaning, oiling, etc.)  Bullet dodged ...

Really, don't underestimate these risks.

Sorry, I’ve given the wrong impression. I do like tinkering it’s just that it sometimes seems like you have to be a genius just to get one of those little boats printed; I’m probably exaggerating.

Well, one certainly doesn't need to be a genius. However, if you expect a 3D printer to be like a regular paper printer or a coffee machine that you just turn on, push a button, wait a bit and a good part comes out,  you are in for a very steep learning curve and a rough ride.

Even the 1000€ Prusa machines (which are otherwise excellent and have a first class support) or 3000-5000€ Zortraxes require maintenance ("tinkering"), setup and an operator who has a clue and isn't just winging it, thinking it will be all fine. With the cheap machines like the various Creality CR10s and Enders it is even worse because corners were cut and things wear out or break faster (or weren't even working properly to begin with). So one has to invest time and money into "tinkering" to get (and keep) them working.

That's the same thing as with cars - if you don't maintain yours properly and drive it like an idiot to boot, it will not last long and it will fail you at the worst possible moment (or you kill yourself in it). It used to be a  completely normal thing that a car driver carried a box of tools and spare parts in the boot in order to fix the inevitable breakages on the road. These days most people don't even know how to replace a blown lightbulb or change oil - but that doesn't mean it doesn't need to be done, only that they have to pay a mechanic to do it for them.

If you are comfortable with mechanical and electronic work and don't mind having to spend time working on your machine, then the cheap printers like the Enders could be good investment (and a lot of fun too). You could even build your own printer, whether from scratch or a kit (like those Vorons or various RepRap machines).

However, if your goal is to use it as a tool that has to work whenever you need it, reliably producing good parts and you don't want to spend time on the machine itself (which is totally fine!), you better add a zero (or two) to your budget - or use a printing service, letting someone else deal with the problem for you.

You need to decide where between these two extremes your comfort zone and needs are and shop accordingly.

[Mechatrommer]

I had few cases where people managed to screw up their normal 2d paper printer.. among the worst are ear ring or cloth pin/clip etc got stucked inside paper feed and mouse renovated the ribbon cable to give way for the sweet home inside.. the most common is they dont know how to do nozzle cleaning.. mostly families and friends that want it to get done for free.. i cant imagine when 3d printer become common in household. Sometime i feel like its better for me to look like a fool who doesnt know a rat ass about printer..

[thm_w]

I did see a YouTube vid where the guy had a mains relay device (110V unfortunately) in place but I can’t seem to find one - a fully complete one rather than the actual component relay.

There are relay modules that will shut off the printer after the print is completed: https://www.biqu.equipment/products/bigtreetech-reply-v1-2-automatic-shutdown-module-after-printing

Another option, for the same $10, is a WIFI mains switch. But you'd have to either have it turn off the power at a set time daily, or enter in a timer when you think the print will be done.

[PerranOak]

This thing looked good in the vid (https://youtu.be/tTJfASOHojo) but is only 110V despite it being a UK vid.

[Ian.M]

This thing looked good in the vid (https://youtu.be/tTJfASOHojo) but is only 110V despite it being a UK vid.

Why would you use such a hack?  240V AC Zero volt release (NVR) switches are readily available.  Ideally get one with an externally accessible coil terminal, e.g. https://uk.rs-online.com/web/p/push-button-switches/3985279

Then its simply a matter of your supervisory circuit interrupting the coil power of the NVR switch by cutting low voltage power to a NC relay. 

However I think the video's reliance on thermal fuses as sensors is also flawed.  Certainly they may be beneficial as a fail-safe for the PSU and SSRs or MOSFET modules, but for simplicity/reliability they should be wired  to cut power to/through what they are monitoring directly, provided they are rated appropriately.

What is needed is redundant monitoring of bed and hot end temperatures, and heater power by an independent monitor circuit (although it may tap into the controller's bed and hot end thermistors to reduce the number of extra thermistors and connections required).  Setup the printer controller firmware to limit heater power to 99% (PWM) and if it ever goes to 100% after the MOSFET, the MOSFET has failed.    If the hot-end heater power is high but temperatures aren't responding appropriately its likely the heater cartridge has become loose or has even fallen out. Carefully positioned flame and smoke sensors could also be beneficial, though if either are detected you probably are in need of an automatic fire suppression system, not just a power shutoff.   Current sensors for all heaters may also be beneficial as they could be used to detect unexpected increases in heater circuit resistance which would indicate a failing connection or wire, hopefully before the bad connection burns up.

IMHO AC powered heaters are inherently more dangerous - if you are going down that route for a moving bed heater, a 120V heater powered from a 110V grounded center tap site transformer would be the way to go, with redundant SSRs, one in each lead controlling it.

Its also dubious to put 3D printed fan ducts in close proximity to the hot block unless fire retardant filament is used to print them.

[janoc]

IMHO AC powered heaters are inherently more dangerous - if you are going down that route for a moving bed heater, a 120V heater powered from a 110V grounded center tap site transformer would be the way to go, with redundant SSRs, one in each lead controlling it.

That's more a feel-good setup than anything really practical or actually useful IMO. Even that 120V powered heater can kill you or start a fire. Even a DC powered 12 or 24V heater is good enough for starting a fire, like on those Anet 8 printers. The doubled SSRs will only protect you from one failing short but not from e.g. the firmware crashing and leaving the SSR conducting full time. Failed SSR is not the real danger there, those things are normally pretty reliable - plus a well set up printer would notice due to the thermal runaway sensing and abort/alarm.

The main dangers there are twofold:

- Many printers have crap wiring for the bed (insufficient strain relief, poor quality connectors, etc.). So as the thing moves, the wires either wear through or the connectors become high resistance - and you get a fire (that was the Anet 8 case) or electric shock from the mains heater wiring. Transformer could potentially prevent that but won't stop two adjacent wires wearing through together and shorting.

- Mains powered bed heaters are problematic because many of these machines are not earthed at all (despite metal construction) and having good earthing on a moving bed is difficult in any case.

Instead of messing with an isolation transformer one shouldn't use a mains powered heater at all - and actually make sure the printer is properly earthed. The bed is the least of one problems when the AliExpress special power supply blows and sends mains to the frame of the machine!

Its also dubious to put 3D printed fan ducts in close proximity to the hot block unless fire retardant filament is used to print them.

Pretty much all printers do that - and not only fan ducts. If the printer is constructed correctly and has the thermal runaway sensing, there is no way that plastic would get anywhere hot enough to catch on fire even after prolonged period. Melted, deformed sure (like mine). Fire, no. The hot end doesn't reach high enough temperatures for that - unless you have a thermal runaway for some reason - and the machine is unable to sense it (e.g. because idiot vendor didn't enable the option in Marlin).

Heck, you have a plastic filament going directly through the hot end, often sitting there for long minutes without movement while the machine is waiting for something - e.g. getting the bed hot or doing some preparatory work.  Good luck with a flame-retardant filament there ...

A smoke detector and not leaving the machine unsupervised is a much more practical solution. Technology can never account for all possible unexpected issues.

[Ian.M]

Yes, I agree there is little justification for AC mains bed heaters.  I believe the argument for them is that the lower current allows thinner, more flexible wires to be used, and causes less problems with connectors, but opposing that are the safety issues you have noted and also that the higher voltage almost guarantees that an internally broken wire will continue to arc, and significantly increases the risk of overheating terminals or PCBs developing conductive carbon tracking across the AC supply, both significantly increasing the fire risk over even a 12V bed of similar max. power, provided said 12V bed is wired with appropriate ampacity fine-stranded extra-flexible wire with connecters rated for the current with plenty of safety margin, and appropriate strain relief is used.

The flame retardant plastics issue is more nuanced.  Obviously, if the hot end heater cartridge comes loose, the firmware *SHOULD* detect that from the thermistor reading, abort and shut off the heaters, but assuming the failure is a shorted heater control MOSFET, on a printer without a software controlled PSU, there is little the firmware can do to alleviate the situation.  Its therefore worth minimizing the fuel available to the potential fire.  Hopefully the heat break cooling fan will be able to keep the filament temperature at the top of the heat break below its flash point, and the firmware certainly shouldn't continue to feed filament into an over-temperature hot end.  Personally I use a thin aluminum part cooling fan duct bent up out of heavy gage stiff foil and held together by interlocking tabs and aluminum foil tape, rather than a 3D printed one.

 

[janoc]

Yes, I agree there is little justification for AC mains bed heaters.  I believe the argument for them is that the lower current allows thinner, more flexible wires to be used, and causes less problems with connectors, but opposing that are the safety issues you have noted and also that the higher voltage almost guarantees that an internally broken wire will continue to arc, and significantly increases the risk of overheating terminals or PCBs developing conductive carbon tracking across the AC supply, both significantly increasing the fire risk over even a 12V bed of similar max. power, provided said 12V bed is wired with appropriate ampacity fine-stranded extra-flexible wire with connecters rated for the current with plenty of safety margin, and appropriate strain relief is used.

The main reason for the mains powered heaters is not really the wiring. Even for 12V the wires aren't that thick - we are talking cca 10A only for a normal sized bed. If the printer is one of the more modern ones, they tend to use 24V and the current is about half. That doesn't need any sort of thick or unwieldy cable - e.g. my 12V Mendel uses a regular 20ish conductor ribbon cable, split in half and the conductors soldered in parallel. The cable never gets significantly warm (apart from the radiated heat from the bed). Moreover, if you use mains, what you save on copper has to be "paid back" in the insulation thickness because of the higher voltage and risks if the wire wears through.

Mains powered beds are used mainly because of much faster heat up times because more current is available without requiring beefy (and thus more expensive) power supplies. Common printers use about 200W supplies, with about half of the capacity going to the bed and the rest powers the electronics, motors and the hot end. The common PCB heater 12V bed on a printer without enclosure takes about 5-10minutes (depending on the type of the bed) to get to 90 degrees. Mains powered one only about a minute or two.

A bit faster bed preheat when starting the print would be certainly welcome but I have never felt it was limiting me in what I do. I may upgrade the machine to 24V at some point, though.

The flame retardant plastics issue is more nuanced.  Obviously, if the hot end heater cartridge comes loose, the firmware *SHOULD* detect that from the thermistor reading, abort and shut off the heaters, but assuming the failure is a shorted heater control MOSFET, on a printer without a software controlled PSU, there is little the firmware can do to alleviate the situation.  Its therefore worth minimizing the fuel available to the potential fire.  Hopefully the heat break cooling fan will be able to keep the filament temperature at the top of the heat break below its flash point, and the firmware certainly shouldn't continue to feed filament into an over-temperature hot end.  Personally I use a thin aluminum part cooling fan duct bent up out of heavy gage stiff foil and held together by interlocking tabs and aluminum foil tape, rather than a 3D printed one.

The firmware could at least send out an alarm (assuming you have that set up).

You also likely have a lot of other things that are potentially flamable there. E.g. the extruder parts (unless you are using bowden, of course), enclosure (usually made out of polycarbonate/plexi), touch probe like the BL Touch (which is completely plastic and right next to the hotend), the filament itself which is in direct contact (by definition) and which will conduct the flame upwards where there is likely much more plastic to be had, the wiring (PVC insulation) ...

Also, if the cartridge comes loose (as opposed to the thermistor), then it could land on the plastic on the bed.

If the MOSFET shorts, it is likely that something on the electronics/driver board will catch fire before the hotend does. So this really doesn't worry me that much - or rather it is a risk that one needs to mitigate against in a different fashion along with the other issues. Such as by having that smoke detector there, a camera pointed at the machine, not leaving it completely unattented, etc.

A blown MOSFET or SSR aren't going to cause much damage if the operator checks on the machine every once in a while (I usually do every 30 minutes or so at least) and monitors the print progress (and any alarms). Even if there is a failure it does take a while for the hotend or bed to get sufficiently hot that they could start a fire. And even then it does take some time before it could do significant damage.

[thm_w]

Yeah my heated bed is 500W AC. I could use a 24V DC power supply but the one I had was noisy, heavy, etc. 110V AC though, fair bit safer than 220V.
I'm not too worried about anything falling down on the bed and burning as its a glass bed.
Earthed the metal portion of the bed, and the frame.

Ideally everything would be in a large metal enclosure but I don't have the tools or materials.

[boyddotee]

I would like to able to print some useful stuff but I also like to “tinker” with things so that I understand them better.

However, 3D printing seems to require NASA level tinkering skills! I’m surprised that any noob has ever printed anything!

BTW has anyone ever had or even heard of actual fires started by them?

No fires,  I keep mine in a closet. I'm no pro I've had a ctc maker bot clone (with USB key debacle, I was involved in it going out the window 🤣), a copymaster (no longer available but a good direct drive printer). And I have 2 creality ender 3, one is now a mix of the copymaster for larger format.

Have to say it's the bed leveling that's caused most of the issues you can have the perfect surface of x material (I've tried a lot) but go with a auto leveling system. I use creality's own but the bl touch is also an option and glue stick. No problems since on a glass bed.

[Lindley]

 I have 2 creality ender 3,

Have to say it's the bed leveling that's caused most of the issues you can have the perfect surface of x material (I've tried a lot) but go with a auto leveling system. I use creality's own but the bl touch is also an option and glue stick. No problems since on a glass bed.

Which Glass bed have you fitted , Crealitys own like on our V2 or something else ?

What was your problem, filament not adhering to the bed or not being able to remove it when finished ?

Do you use a particular brand of glue stick, Eg Pritt

[janoc]

 I have 2 creality ender 3,

Have to say it's the bed leveling that's caused most of the issues you can have the perfect surface of x material (I've tried a lot) but go with a auto leveling system. I use creality's own but the bl touch is also an option and glue stick. No problems since on a glass bed.

Which Glass bed have you fitted , Crealitys own like on our V2 or something else ?

What was your problem, filament not adhering to the bed or not being able to remove it when finished ?

Do you use a particular brand of glue stick, Eg Pritt

(Attachment Link)

Throw away those glue sticks and use something like Dimafix:

https://dimafix.com/

Yes, it does cost more but works wonders. Once the bed is hot the print sticks like, well, glued on. And once it cools down, it just pops off, unlike with glue sticks where I had the print either not wanting to stick properly, coming off mid-print or stick so well that it took out the top surface of the glass too. I am regularly printing PETG and ABS with this on glass - which wouldn't stick at all otherwise. PLA doesn't need any glue and can be printed directly on glass as long as it is heated.

I am using normal borosilicate glass. Nothing special. Even normal window glass or cheap mirror will work but I wouldn't trust it to withstand heating and cooling for a long time without cracking.

[Lindley]

Throw away those glue sticks and use something like Dimafix:

https://dimafix.com/

Yes, it does cost more but works wonders. Once the bed is hot the print sticks like, well, glued on. And once it cools down, it just pops off, unlike with glue sticks where I had the print either not wanting to stick properly, coming off mid-print or stick so well that it took out the top surface of the glass too. I am regularly printing PETG and ABS with this on glass - which wouldn't stick at all otherwise. PLA doesn't need any glue and can be printed directly on glass as long as it is heated.

I am using normal borosilicate glass. Nothing special. Even normal window glass or cheap mirror will work but I wouldn't trust it to withstand heating and cooling for a long time without cracking.

Thanks,    Do you prefer the Stick or the Spray Can

Will be looking for a new Glass as the V2 coating has already come off with a print during a very difficult release with PLA.

[janoc]

Thanks,    Do you prefer the Stick or the Spray Can

I have actually bought both. The spray is easier to apply on a larger surface (do it away from the printer - you don't want to gum the machine up with the glue!) but then after the print is removed, you will want to re-apply the glue at the used-up areas. If you don't, in those places the new print could stick so well it will rip off the surface of the glass. Go figure - one would assume it won't stick there at all since the glue was removed by the previous print but I have damaged the glass plate multiple times like this already.

Re-applying is easier with the stick because you only touch-up the used areas. Otherwise you will need to wash the glue off (water removes it easily), dry the glass and re-spray the entire surface. With the spray can it is difficult to apply it to only a smaller area without having a very uneven surface coating.

[MarkF]

I use hair spray on glass with a heated bed.
Just print, let it cool down, pop off and start another print.  No cleaning in between prints.
One thin coat of hair spray last me 10 to 20 prints before needing replacement.  I usually clean with 99% IPA or just water until the glass is clear without streaks.  Sometimes I'll spray a small touch-up on some bad places.

Leveling the bed for that first layer was a long long learning process for me.  That made all the difference in getting my prints to stick.  I print PLA and PETG with just an increase in temperature for bed and hotend and a lower fan speed are basically the only changes I do for PETG.

[gmb42]

I use the standard coated borosilicate bed on a CR6-SE, just clean it every few prints with soap and water, dry off with a microfibre cloth, job done.

Great adhesion with PLA, PLA+ and PETG, prints just slide off when the bed has cooled, no need for any messy glue or extras.