3D Printer Hardware Failure

Yes. Yes. I guess that this half of the year is going to be hard on me for hardware.

I was 3d printing a plant pot … weird I know … but I had an idea for growing some moss in my office … weird I know :slight_smile: … and I needed a break from printing parts to mount my CH Products MFP to my Virpil desk mount and from printing a model of the Falcon Heavy.

So I was in the middle of the print (and it is hard to hear the printer in the other room after I upgraded to a new motherboard (for the printer), which has new drivers for the stepper motors, making it very quiet) and I heard a small bang. My cats flew out of the printer room like hell was on their heels and I went to investigate. Now I should probably leave that bang out of the story because, if you know cats, they tend to like knocking things over and it is not unusual for them to do just that and freak themselves out but it was what got me to check on the print.

And it was stopped. The printer head was stationary, in contact with the surface of the print like the power had gone out. Odd. I was immediately concerned because a 215° C printer head has a tendency to melt all the plastic around it, which it is intended to do as long as it keeps moving. It if sits somewhere for long (like a belt freezes or something) the extruder can keep sending plastic through it and with no where to go, it can quickly surround the head and case fire problems.

After getting the computer out of it’s dead-screen state, I was ‘relieved’ to see that OctoPrint (running on a Raspberry Pi3 connected to the 3D Printer) had received an error from the printer…

:fire:“Ender3 Pro Error: Thermal Runaway, system stopped! Heater_ID: bed”:fire:

Yikes! (Fire emojiis added for dramatic effect)

So one of the additional benefits of the new motherboard in the 3D Printer is that the up-to-date Marlin firmware has a couple of safety mechanisms enabled. A few of these mechanisms relate to detecting when either the bed heating or extruder heating has run amok and is trying to burn your house down. This can happen if either heater enters an uncommanded (listen to me sounding like mission control), uncontrolled heating cycle or of the temperature sensors attached to those heated elements starts giving false data.

The firmware had detected a problem and shut the system down. Everything was at room temperature. Whew.

So lets take a look at what might be happening. I rebooted both the Printer and the OctoPrint RaspberryPi and asked it nicely to set the bed temperature to 60. And nothing happened. I mean the graph displayed the intent to get the bed to 60 but the temps from the temperature sensor continued to read room temperature.

Well that likely means that either the sensor has become detached from the bed or the bed ain’t giving me the degrees that it should be. Quick look at the underparts of the Printer and the sensor looked like it was in place.

Looking at the back plate, however and I notice this:

Mr. Ground Wire is no longer connected to the plate. Well there’s your problem! A little disassembly and it is clearer.

I could not get the temperature sensor wires to get out of the shot but it looks like the ground wire to the heating plate was not soldered correctly and just the tip of the copper was connected to the solder. The red wire looks better but if there is as little copper available under the solder as the ground one, it is not great.

So this should be a relatively quick fix … but then I notice this:

The power supply wire, running from the supply to the motherboard has obviously been having a hard time of it. The connector is fused together and, to make it worse, melted. Oi! I am not happy to see this.

Incidentally, does anyone with electrical knowledge know why the positive wire on one side (bottom) and the negative wire on the other side (top) would be the ones melting? The other two wires ‘look’ fine while the ones showing the obvious problems have melting on the heatshrink wrap and discolouration on the actual connector.

Looking at the table that it was sitting on (note the corners of the white squares - calibration squares as I was dialing in the bed leveling sensors):

It is hard to see but there are two faint discolourations in the top of the table. Where the wires above were situated. The top position was where the printer was last week (or the week before) and the bottom one is where it was moved to recently - toward the back of the table to allow better feeding of the filament into the printer. Those long prints (plant pot and Falcon Heavy) of 8-13 hours were causing some ‘burning’ of the top of the table.

I am extremely glad that I caught this. I do not think that I was in a strong potential fire situation but it could have gotten worse pretty quickly. I need to figure out a way to deal with this - likely with a fire-resistant mat under the printer.

It should be noted that one of the problems with some of these ‘cheap’ printers is the quality of the parts that they are built with. Some of the XT60 connectors used are counterfeit parts - not true XT60 connectors which should be able to handle the current/voltage. This is something that has been discussed on various forums related to 3D printers, robotics, etc. I am experiencing it first hand here, I think.

It could also be that the bad ground wire connection was drawing more power than it should have due to the increased resistance of the connection. I think (@troll (or anyone with electronics experience (@TheAlmightySnark) can correct me if I am wrong) that bad connections like that can show up as increased resistance to the electron flow and that can lead to heating at the site of the bad connection and increased power draw as the system tries to compensate and give me the power necessary to get the heated bed to the requested temperature.

I also think that the bang I heard was, indeed, my cats as I am pretty sure nothing that I am seeing here looks like a rapid failure event. That ground wire has not scorching or post-arching evidence at the disconnect site.

On the inside the power supply looks unaffected. So there is that.

So, for anyone out there using ‘hobby’ grade equipment with some power consumption … check your connectors on a regular basis. Catch this before it becomes a fire hazard! We take it for granted but a quick check can save you time, money and having to find a new house!

Stay safe out there.

7 Likes

While that is essentially true, something like that rarely happens on its own in that, there’s either more resistance than there should be (through a faulty solder spot, e.g.) or there isn’t. The circuit resistance degrading by itself is unlikely to happen at the power levels you are facing (unless the wires are sheared mechanically).

This looks more like a short that wasn’t caught by some kind of overcurrent protection. The reason i say that is that if you’d have had an increase in resistance and the voltage regulator had tried to compensate, your wires wouldn’t be fried. All that additional power would be dissipated at the point of failure (where the resistance increased). Instead if you have an overcurrent situation, suddenly the entire circuit has to dissipate way more heat than is usual.

So my thoughts are, either the solder joint on the pad failed mechanically and caused a short or there’s a short somewhere else in your circuit that caused an overcurrent, that heated the bad solder spot and caused it to fail.

Before you try to repair this, check the resistance of the heating circuit. If that is unusually low, that would explain the failure. Minor note: Those solder spots on the pad look kind of cold. You should maybe give them a do-over with a bit less solder.

It depends on resistance and heat dissipation ability. It’s easy enough in theory but if you want the exact numbers of why that length of wire was the one to heat up the most, it gets complicated. Could be that the soldering joint to the plug was slightly worse on those wires compared to the other two. Nothing to worry about really, as during the failure those seem to have been exposed to a serious overcurrent. Plugs are usually the first thing to heat up due to higher resistance than the wires.

1 Like

Yeah, my thoughts too.

@Fridge, When you resolder that negative wire, are there really no wire strands left inside the solder blob…? Did they really solder such a small part of wire, to the plate?

3 Likes

I will do this. The both look odd to me.

And I will check this as well. I am curious as well :slight_smile:

1 Like
© 2019 Mudspike.com | Articles Website | Forums Rules & FAQ