! 448888 !>
For comments, suggestions, and questions; please contact frontal[dot]harvest[dot]25[at]gmail[dot]com.
2015-03-29. By Patrick.
For a while now, we have wanted to get some of our PCBs manufactured rather than home built. Toward the end of last year, we purchased some ATmega328 TQFPs from digikey (they were out at SparkFun). A small digression here: It was fascinating to see the number of out-of-stock items on the SparkFun site increase as Christmas approached. Any, we also got some great little TQFP breakout boards from Adafruit. I tried soldering one of the ATmegas on there with paste. It was my first time trying this technique and it worked like a charm. And wow! It really was easy.
But really, there is no point in getting a TQFP and then soldering it onto a breakout board that is larger than the DIP version of the chip. Or rather, the only point is for developing a circuit on a breadboard when there are TQFPs available but no DIPs. So we needed to design the TQFP directly into each circuit or into a module to be soldered into a circuit. And here is the problem. From my point of view, the problem is not that TQFPs are hard to solder. It's that it is damn near impossible for me to home build a PCD that will accept a TQFP.
That said, I did try it. Direct etching or hand milling the copper to the TQFP pitch was a total failure. I did manage to create a mini dev board by bending the even pins of the TQFP up and over the top of the chip. I then etched the copper to handle a pin pitch twice the normal TQFP pitch. As the traces move away from the TQFP, I had them splay out so that I could fit new traces between them. Let's call these the 'even pin traces'. I then soldered the TQFP odd pins directly to the board. The idea was to add short wires from the even pins to the even pin traces. But that was difficult. These bunt up even pins sit above the package. During soldering, wires placed on top of them tend to fall to one side or the other where they often make contact with the adjacent pin. And they don't just fall of when being soldered to the pin. When being soldered to the PCB, the heat travels up the wire to the pin, melts the solder there, and boom the wire falls off again. The same type of thing happens trying to solder to the board first. Finally I got a bunch of them soldered, not all of them but enough to test out the board. It tested OK, blinking some LEDs. But I had to do so much repeat soldering of wires that kept falling off that I wouldn't be surprised to find that some part of the chip has suffered heat damage.
The solution is definitely to have the PCB manufactured for us. And this we did using the OSHPark service. But that's another blog entry.
The text we wrote, pictures we took, and system we used to layout our web pages are licensed under the Creative Commons Attribution 4.0 International License. Link to License. This means that you can copy anything you want, use it and alter it as you want, as long as you give credit to us for the original work, provide a link to the license, and explain any changes you made. We try to make sure all other content we display is also available to use. But check the sources we list before using them.