As I was casually blasting video game psychos with my son, Jack, I started to smell something strange. The Cartesian 3D printer had been plugging away, layer by layer, and because I was lost in the game world of Pandora, I didn’t notice that the printer’s bed temperature gauge was steadily approaching 100 degrees Celsius. I paused the game, shut down the printer, and began to assess the damage.
I have heard tales of the fire hazards of unsupervised RepRap activities; of power supplies failing and sparking or extruder thermistors falling out of the place where they should be (resulting in an out of control, face melting nozzle), but I had not heard of this happening to a heating bed.
The first clue that the issue was running deeper than a quick switching of a wire or part came as I flipped the printer’s power on, and noticed that the bed temperature was steadily increasing, and that I could read that it was doing so on the LCD screen. If the thermistor was broken, the readout on the screen would not correlate with the physical temperature of the bed. I switched out the thermistor anyway, and surprise surprise: no change in behavior. The smell, a cross between burnt fish bone and a sterile dentist office, also became more pungent when the power was on.
Flipping the power on (again), I moved my attention to the TriGorilla mainboard (This is a proprietary design by Anycubic, based on the open source RAMPS 1.4 configuration). The smell was definitely coming from somewhere on the board. Switching the power off, and satisfied that I narrowed the search to a more manageable area, I took my problem to Thingiverse ( A forum for 3d files and 3d printing enthusiasts). I was directed to run a diode test to confirm if something called a MOSFET was shorted. In short (pun intended) the MOSFET was the culprit – but before I continue the story, here is a a little on MOSFETs:
A metal-oxide semiconductor field effect transistor, or MOSFET, is a device that makes use of the field effect as a gate to to regulate conductivity. In the case of the MOSFET on the 3D printer board, a small current is applied to the MOSFET gate when the controller determines that the heat bed needs to be on and that the bed is below the desired temperature. The small current closes the circuit between a power input (source) and a power output (drain) and the result is a hot bed to keep that precious filament adhering to the surface. I had gone through a few lessons on transistors, but the idea of the field effect was somewhat new (and exciting).
The trouble with tinkering with electronics as an amateur, is that there are too many places to dig. MOSFET? It is a transistor. Transistor? A type of silicon relay. Relay? And it goes on and on. Thankfully, the maker and reprap communities are there to keep nudging me in the direction of knowledge. For this particular problem, I took my questions to Thingverse.com, supplying the necessary pictures and symptoms, and it was quickly determined that my problem was indeed the MOSFET and that I had two options. I could scrap the board and buy a replacement (I had been wanting to ditch the TriGorilla board, as It required me to use an outdated Arduino IDE to update the firmware) or, I could de-solder the MOSFET and replace it with a new one. I opted for the latter.
Another place to dig. This time, knowledge AND skill was required. I turned my attention to youtube to figure out how to de-solder a MOSFET and quickly found this video:
I hate how these guys make it look so easy, just like the woodworking videos all over again. I should know better, but instead, pried all of the cables out of the board (careful to mark where they went) and attempted to emulate the skill displayed in the youtube video. I just had to heat up the gate enough to melt the solder that was adhering the tin plate and PCB board to the MOSFET, tap the source and drain, and voila. One de-soldered MOSFET. That’s what I saw in the youtube video, and that’s what I told myself was going to happen. Like all plans, this one didn’t survive contact with the enemy – err – the soldering iron.
Here is how it actually went: I don’t have a soldering station, so instead I used a cheap “helping hand” that – in retrospect – should be renamed to have the word “shaky” or “wobbly” included in the title. I set the iron to 450 degrees and placed it firmly on the MOSFET gate. What happened next, to which I will neither confirm nor deny that actual evidence exists, involved some melted plastic, swearing, and the eventuality of ripping the MOSFET off the board along with the PCB tin plates. OP success.
The board was trashed. Repairable, sure, but since I was dealing with larger than normal amounts of current passing though an area where I might have done a poor job of soldering, I instead decided to go with the other option and order an original RAMPS 1.4 board (RAMPS is an acronym for Reprap Arduino Mega Pololu Shield). The community didn’t exactly tell me not to attempt to replace the MOSFET, I just inferred that since the device failed due to a heat problem (most likely) I had better go toward the safer route. I would save the A4988 stepper motor drivers, but I could use the opportunity to de-solder every piece on the board. Maybe even re-solder everything as well.
So hear I am. Cartesian printer in pieces, chips, MOSFETs, capacitors and resistors successfully desoldered and scattered everywhere, and a new 3D printer brain waiting to be reassembled. The entire point of being part of this movement is not only a love of tinkering, but a direct path to understanding how this world works, one burnt MOSFET and one funny smell at a time. Thank you Thingiverse for all of the help with this.