Hi all,
I finished putting together the Makerbot Kit and had the Plastruder connected, calibrated, and working to the point that it could extrude a small amount of plastic fed into it. However, after about 2 days of success with the heater, the heater stopped operating.
I measured the resistance of the nichrome wire to be 6.8 ohms, which is the same as when I measured it when the heater was working. I also also determined that there is no problem with the nichrome wire insulation by measuring infinite resistance between the nozzle and each end of the nichrome wire, respectively.
After commanding the heater to heat to 220C in ReplicatorG control panel (and confirming that there was connection by successfully moving the motor via the control panel) I measured the voltage across the heater terminals with the heater wires attached (measuring 0V with LED not lit at all) and with the heater wires detached (measuring 3V with a dim LED). I also used an oscilliscope and observed that (with the heater wires detached) there was a PWM signal being sent when I commanded a relatively small temperature increase.
I know that it is not a problem with the entire board because the Thermistor and the motor for the Plastruder are both working properly. I also didn't see anything visibly wrong with the board (nothing seemed to be smoking or burning). I think it might be a problem with a mosfet near the heater pins because I am able measure 12 V on the board up until the mosfet. Has anyone else had a similar problem and can provide a solution?
Some users reported that their mosfets were not soldered in very well. Apparently they were able to press on the mosfet until it made good contact. Try adding a bit of solder to the pins if you can. Be careful not to leave a solder bridge between the pins.
Hmm… what really would be interesting is how your multimeter was set when you measured the 3V. If there was a PWM-Signal and your multimeter did not measure the peak value but some kind of RMS value, its pretty clear that you never get a 12V reading out of your multimeter. The PWM signal could at the same time dim the LED brightness too if the pulse is short enough.
So, lets try to narrow it down a bit further:
When you measured the PWM signal with detached heater wires, how big was the amplitude of the signal? Was it still only 3V peak-peak, or did it go the full 12V range? Did you measure at the drain-source of the FET, or only at the gate, or both? Can you measure a different output signal from the FET (with your oscilloscope) when there is a heater attached or no load present, when you enabled the heater?
Your ATX-Powersully is totally okay, right? you did not see the complete 12V supply on the extruder controller drop to a lower voltage when you tryed to enable the heater?
Did you measure the amplitude of the signal that drives the FETs gate (best to use oscilloscope here too)? Just to make shure its not the microcontroller that could have a defective output pin. If your signal looks good there, i am almost 100% shure it must be a FET failure.
For my taste it's quite a big design flaw that they forgot to add a gate resistor between the FET and the microcontroller… Knowing that the gate basically acts like a capacitor there can be some nasty currents trying to get sucked directly out of the controller, and we all know microcontrollers just dont have the pins to provide a lot of juice and they usually dont have any way of limiting the current, so the weak port-pin drivers in the controller could actually get blown to bits if we are unlucky.
(On a sidenote, these spikes can at the same time make your 5V logic supply very noisy and unstable. It just adds up to more and more flaws that could have been eliminated by adding some cheap and clever capacitors and resistors at the right spots… Ah, well, guess i have to live with it for now, or make it better myself. For now i wont have time to work on the extruder controller as i am planning to get the mainboard replaced with a cortex M3 based developper board from Luminary and porting that piece of code over will probably give me enough brainfood for the near future…)
the NIF5003 isn't a raw MOSFET, it has a built in gate resistor so you can drive it direct from a logic pin. It is also protected against over voltage, over current and over temperature, so I am really surprised to see things like "don't make the heater less than 6R or you will fry the MOSFET." They should be bomb proof, but they will cut out if they get too hot.
Ooops.
My bad. I knew it was a protected FET but i did not have the data sored completely in the back of my head and missed the fact that it has a built in gate resistor too. Sorry about that.
Anyway, i think there are still some valid points in my posting above to figure out wether its the FET or controller that has a problem. And did anyone actually see some information about how big the included gate resistor is in the NIF5003? The datasheet does not offer a lot of information about how the internal logic should be suppled, and i cant see any data about gate capacitance and other interesting values…
Anyway, i have just seen that the NIF5003 seems to be at the end of his product cycle as it is not recommended for new designs according to OnSemi. So for the future, you (we) should start to look for an alternative FET before they get rare.
