Had a little mishap with my printer again. Apparently there was a bad connection and one of my plastic plugs melted. So to fix it I soldered new wires on the board, an inline rc battery pack quick disconnect and it seems to be doing ok.
The connector was overloaded. This happens constantly, because Chinese companies don’t use proper de-rating factors during connector selection. (A “12A” connector generally can’t handle 12 amps through two adjacent terminals at the same time.) They also don’t care, because it never happens early enough in the printer’s life for it to be an issue for them.
It’s amazing to me that this has only caused A FEW real fires that we know of. I’d guess there’s at least a hundred thousand ticking time-bomb printer connectors out there, slowly going into heat-oxidation resistance runaway failure, potentially able to burn people’s house down.
I’ve had this problem repeatedly on 12V machines with genuine RAMBOs and genuine traceable-source Molex SL 3953xx connectors, I think the biggest lesson is that we really should be avoiding 12V printers, the currents, especially for bed heaters, are just too high for reasonable sized connectors and wiring at that voltage.
I was running a machine with an enormous heat-sinked external FET and Mini-T connectors for the bed for a while before I gave up and converted the whole thing for 24V supply (and smoothieboard, and e3d head, and…).
Spelling pedantry: ferrules.
I’ve been doing that in my projects lately just for serviceability, an assortment of ferrules and a crimp tool can be had for a couple bucks from China, and it prevents fraying. I’m not convinced they really help with current-carrying capacity.
@Paul_Eberhart The phrase “reasonable sized connectors” should never, ever appear in a discussion of high-current electronics. You don’t size electrical components by eyeball or by gut feel. You use equations and manufacturer specs, and you do some engineering, and then the part will work reliably >999,999 times out of a million without failing. A properly-designed 12v circuit can handle 30A all day with cheap Chinese components.
Seriously, the problem here is bad engineering, and it’s a fire hazard, and people need to treat this as a deadly-serious design problem in huge numbers of popular cheap printers.
Not your fault, not trying to be mean, this is just a seriously effed-up thing the hobbyist maker community does constantly, and it’s a recipe for killing somebody sooner or later.
You have to buy a new board, and my advice is, use a Rele, between the heatedbed and the board, the rele can handle 40A without problem, and you eliminate any trooble with the board.
@Ryan_Carlyle By “reasonable sized” I meant physically reasonable compared to the rest of the electronics once you’ve derated appropriately. You shouldn’t be running a 12V bed heater of typical size through anything smaller than an XT30 or so once you’ve added up all the situational derating.
I was more making the point that there is no compelling reason to design 12V systems anymore, and there are a ton of disadvantages, some of them dangerous.
I agree that a lot of it is manufacturer BS, the already-bulky Molex 3953xx Eurostyle pluggable terminals UltiMachine is using claiming “15A” on are only rated for a nominal 18A, which is (demonstrably and theoretically) no where near enough overhead for something being physically wiggled by machine motion and pulsed with PWM.
@Paul_Eberhart I agree 12v is obsolete, but it’s not inherently bad to run a heatbed on 12v – it just needs to be fairly small and low power. Maybe in the neighborhood of 100w to use standard 15A parts safely.
Don’t get me started on heatbed flex wiring…
I’ve Learned to derate connectors at minimum to be 2x the maximum current. Since you’re going to have a hard time finding 2 pin 30amp connectors it typically ends up 4 pin, which helps because now you can use 4 thinner wires instead of 2 thick inflexible ones.
One of our sister hobbies (RC Aircraft) has, by necessity, made a pretty nice selection of high-current connectors readily available. The Amass XT30/60/90 connectors (those numbers are nominal current ratings in amps) seem desirable, and as noted above, I’ve used Deans connectors (nominal 30A) picked up as RC supplies on printers when the existing options failed.
it held long time…
as a member of geeetech g2s g+ group, we reported this more than year ago to them… it was fixed in later revision of the board…
it was one of the cheapest board at the time it came out, i counted 3 revisions, with latest being A+ that has all screw terminals and bit more pins broken out.
but since you have first rev., you have an option of powering bed on separate voltage… later boards had soldiered bridge on power pins.
if you are running on 12V (bed plug shows that you are), with >120W bed you are risking burning power/bed plugs and bed mosFET (without/weak active cooling).
i presume its geeetech gt2560 first rev.A board.
and if whole board is powered of 24V, its proned to buring 12V regular for fans (if overloaded)…
mine had power plug burned, but is still working.
the problem is most of the time not high current but sparking which generate the heat. This has to do with the PWM modulation (and originated by a bad connection).