Friday, May 15, 2015

Prusa i3 Rework- MOSFET Repair, Part 2

After I got the headers soldered onto the RAMPS 1.4 board, it was time to make the breakout board for the MOSFETs. This would allow them to be able to be heatsinked properly, which helped to make them last longer.


I ended up buying entirely new transistors- they are TI's CSD18532KCS, which come in a TO-220 package, and are equivalent for our purposes as replacements for the STMicroelectronics STP55NF06L, which is given in the bill of materials for RAMPS 1.4. You can compare the CSD18532KCS datasheet and the STP55NF06L datasheet and see that they are quite similar. The TI MOSFETs are a bit pricey, at around $2 a pop, but they seem to be solid replacements for whatever was on the board previously (mine were not the recommended STP55NF06Ls).


Some of the items needed for the breakout board build include TO-220 heatsinks, 3-pin 0.1 inch female headers, and the replacement transistors. What isn't pictured is the perfboard for the breakout, and the hardware for securing the transistors to the heatsinks. The male headers in this picture are for the main board modification.


I cut the female headers from a long strip with wire cutters, then filed the edges.


Here are the three tiny screws (M3 x 6mm) and the M3 nylon washers that go under the screws, used to secure the transistors to the heatsinks.


Here is one of the MOSFETs, a TO-220 heatsink, and the mounting M3 screw and washer.


And here are the components being attached together. At this point, the screw should be tightened loosely with a hex wrench so that the MOSFET can still slide around.


Here is one of the three MOSFETs, now completely assembled.


Here are all three of the MOSFETs, ready to be mounted and tightened.


Here is the project board that I used. It's a small board made by Radioshack, but anything will work, as long as it has 0.1 inch hole spacing. I marked out three of the patterns on the board, which consist of 3 pins, then a gap, the another three pins, then a single pin. I marked the holes with a permanent marker.


The single holes were drilled out with a 0.015" drill bit, in order to accommodate the mounting pins for the heatsinks.


Another shot of drilling out the holes.


At this point, I am putting in the MOSFET assemblies, but not soldering them in yet.


Another shot of how the pins line up.


Here are all three of the MOSFET assemblies resting on the board. Again, I waited to solder them until later on in the process.


The reason that they were on the board was so that I could tighten the mounting screws. Now that the mounting pin for the heatsinks was lined up, I could tighten the mounting screw and have the assembly oriented correctly.


Another shot of tightening the screws.


At this point, it was time to mount the female headers onto the board, in the innermost 3-pin slots.


A trick that works well for soldering lots of components of the same height is to mount them all at once, and flip the board over on a flat surface. They will all be supported. I helped to do that method here by placing an extra length of female headers on the fourth side of the board. This way, the board would not tip over.


I put an index card over the top, and flipped the board, so that the headers would stay in.


Another shot of the board, flipped over.


A side view of the board. This was right before I began soldering.


At this point, I had done quick solder joints on each header, to keep them from moving around. Soon they were all complete.



Now, I put on the MOSFET assemblies.


Another shot of the assemblies on the board.


I flipped the board over, with the same index card trick as described earlier.


Then I soldered in the heatsink mounting pins.


Soon I connected the header pins and the transistor pins with bridges of solder.


Another shot of the soldering, completed. A bit messy, but there was no accidental bridging.


I trimmed off the tall MOSFET leads with wire cutters.


I then used flux remover to rub off the solder flux.


And here is the completed board!

The last step in the process was to get the cable harnesses set up that connect the RAMPS board and the breakout board. More info on that soon.