MK2 Heated Bed stopped working....

I think my heated bed has short circuited. It does not heat up anymore.The sensor appears to be fine. What should I look for to replace? I am hoping its just the LED's and resistor not the bed itself. Is there anyway of diagnosing this with a multimeter without taking apart everything? The wiring seems fine.

edit: The 1k resistor still reads as 1k. So I presume its not that. I tested a spare LED with my mulitmeter and it lit up so I decided to try the ones on the board and they did not light up. Would they if they are soldered to the bed? I also tested the connection to the bed and the multimetre beeps so I presume thats fine also.....

Edited 2 time(s). Last edit at 02/18/2013 05:35PM by jininjin.

What electronics are you using?
If you have a ramps board, disconnect the 2 leads on D8 (Ramps) going to the heated bed and check the resistance- should be around 1 ohm or so. Did you check for an output voltage on D8 ? (with the bed switched on)

Sanguinolulu 1.3a.
I will check ohms and output voltage, thanks.

Ok, I have seemed to have found the issue. The molex plug to bed connection output of the Sanguinolulu melted. Only the negative side. See pic.

I redid the wires and plug it in. Now the bed heats up normally but I decided to put my finger on the bed output of the Sanguinolulu and it gets pretty hot. Why is this getting hot? I noticed the LED#1 of the bed does not turn on....When does this light turn on?


Also to answer the previous question which may not apply anymore:
I measured 1.5 ohms and it also beeps when I checked the resistance of the leads from the bed. The output from the sanguinololu heated bed output is 12.3 volts.

Edited 1 time(s). Last edit at 02/23/2013 09:13PM by jininjin.

It gets hot and burns out because the connector is not rated for 10A. I connected one lead from the bed direct to 12V on the PSU and connected the other to the tab of the MOSFET with a ring terminal. That avoids the connector completely.

---

[www.hydraraptor.blogspot.com]

Thanks Nophead. I may just re-solder some molex connectors rated 10A or higher then. I presume that should work too.

Edit: Molex does not seem to have anything that fits the PCB.... So I will go for the screw terminals. It looks like people are using only 2 of the 4 through holes on the PCB. Does it matter if you only use 2 holes to power the heatbed?

Like this:
[4.bp.blogspot.com]
[www.resco-research.com]

Edited 1 time(s). Last edit at 02/24/2013 04:34PM by jininjin.

To answer my own question. No it does not work. I am bit inexperienced so I thought each through hole had 12v but obviously its 6volts each. While doing this I somehow fried something too.....so now I am worse condition then before, but I am used to something failing every time I print something so I will get over this....

Hi, I'm tripped up here searching for my company references (www.resco-research.com).
Here my experience: we use a 5.08mm two poles connector rated for 15A without problems. A bit of care must be taken to assures that the connector is soldered on both sides of the pcb. In this way the current flow in the traces on both sides. I made some hundreds of print with this configuration, using a 0.8Ohm bed with a 12V supply (about 15A current) without an issue.

Thanks Dario,
Good to know. I think I know what happened then. The through hole eyelets loosened so I had very little connection which caused it to short. I have decided to go with a Arduino/Ramps configuration (since this is my second Sanguinolulu failure) but I am hoping to come back to it and try to repair the eyelets but I figure its toast now.

I wonder why my Sanguinololu was not sold with the correct connector seems kind of bound to happen sooner or later.

If you toasted only the connector you can replace it soldering a new one. If a pcb trace is damaged, you can solder a wire "over" the short (using two solder join connected through the interrupted trace).

There are many Sanguinololu assembled with incorrect connectors, I never understood why.

---

Dario Corsetti
R.eS.Co. research
www.resco-research.com