In the past, when I’ve done surface mount soldering, I’ve either used a hot air gun or my soldering iron on parts that I had access to, and avoided designing circuits that required QFN, DFN, LGA, BGA and etc.
A few years back I invested in a 2000W cooking hot plate to allow me to use solder paste for those parts with no access underneath. I have yet to get my solder paste out of its’ tube and stumbled on a YouTube video showing how you can tin your pads with normal solder and re-flow in the same manner. This requires a very thin layer of solder , just enough to make a solid connection but not too much lift the parts away from the PCB. After this add a little no-clean flux and away you go.
Here is the Test1 PCB design from the WearWare: G-Force Data Logger project. One of the benefits of having three PCBs from OSH Park is that you can experiment a little and not worry about it too much.
I recently purchased a vacuum vice to help with various soldering tasks, and for £7.99 it has been well worth it. I am only tinning the pads that are going to be soldered to during the reflow process, the rest will be soldered to when / if the through-hole components are soldered to the design in the usual manner.
The pads need to shine when you look at them, and not be dull as if you have wicked them away with some braid. I found placing a very small blob on my soldering iron was enough to cover all the pads as seen above, to give you an idea of how much (Little?) you need.
Because we’re not using solder paste, we lose the tactile stickiness that solder paste provides, meaning that when the board is transported from where it is populated to the hot plate it is very likely that things will move out of place. To compact this I did my populating on the hot plate itself whilst it was cold and unplugged. After this comes the fun part. Because the thermostat on the hot plate is simply Off->1-5-ON there is no real gauge of how much heat is being produced. What I did was slowly turn the temperature up and let the hot plate indicate it has reached temperature, check to see if everything has re-flowed and if not increase the heat a small amount and try again until it does. Unfortunately for me the control on the heat plate oscillates a lot meaning it likes to over shoot. This caused the boards underside solder mask to start smoking.
During the first test the flux being used had taking in water whilst in storage leading to popping and bubbling occurring leading to a rather odd display of components flying everywhere. Not what I was hoping for. And on top of this C1 had managed to solder itself to one of its pads and the neighboring pad of R3 causing a bridge.
After cleaning all the solder and remaining flux of the PCB the pads had to be tinned once more. This time, without any excitement generating flux, the board re-flowed successfully. One of the resistors had tomb stoned but otherwise everything is where it is supposed to be and without any shorts / disconnects. One of the benefits of this method is that it is hard to create solder bridges because if you apply too much solder things don’t usually re-flow nicely without flux.
Here is a another shot of the board after my attempt of fixing that resistor for comparison. You’ll notice how out of line it is and the amount of solder needed to make a connection.
One of the challenges I found with this was populating the PCB with components with tweezers. It is by all means possible, and much easier with solder paste because of the tactile nature of it. Moving forwards I will be making a manual pick and place of which I have used before as they make the process much easier.
Thanks for reading, see you next time.