I've been reading about using toaster ovens to reflow PCBs for years now, so I thought it was about time I gave it a go. Consider this a bit of a feasibility study. I've never done this before and wanted to get an idea of how hard it is, and what issues I need to account for.
The board I'm trying to assemble doesn't have many parts and all but one device could be soldered by hand. With a few part changes I could avoid the reflow process all together, but parts keep getting smaller and eventually it's a skill that'll be critical to know. Besides, when done properly, the resulting joints are also cleaner, and stronger.
Sometimes the most important thing you can do is just start. So with that in mind, I started testing using my toaster oven and an old breakout board I had. The oven has 4 ceramic elements, 2 at the top, 2 on the bottom, and I didn't want radiant heat from these to directly heat the board. I figured that would be a bit too harsh and uneven, I wanted the hot air to do the job. By putting a metal tray on the top and bottom shelves of the oven I could shield the board from the elements. Ideally I would have liked a fan forced oven to distribute the heat quicker and more evenly, but for now this will have to do.
The next step was to get an oven thermometer and hang it from the rack and verify the temperatures on the dial. I would set the oven to different temperatures, leave it to stabilise, and then read the thermometer. Surprisingly the temperatures on the dial matched my readings reasonably well. It's not the most accurate method, but it gave me a little more confidence in the equipment I'm using.
I've read quite a few articles on reflow profiles, how hot should things be, how fast should things be heated, and how long should you leave it at a temperature, and in the end decided to have a go at manually controlling the oven. My first test was to try and solder an 0603 resistor to a board by applying some solder paste, putting the part on the board, placing it on the tray, and turning the oven on full from a cold start. My plan was to observe the board, and as soon as I saw the solder liquefy, count 20 seconds turn off the heat and open the door gradually over the next minute and let the board cool naturally.
|Solder Reflow Test 1 - Unsupported board|
Things didn't go too well. I was slightly preoccupied and board was scorched. The first indication that something went wrong was when I started to smell burnt FR4. The board also made some cracking noises. I think it was starting to de-laminate.
The mounting of the resistor wasn't great either. The joint was irregular and a strange yellowish colour. It looked like lightly tempered steel. May have had something to do with the rate that the board was cooled, could have been flux, but I'm not sure. The resistor also rotated as I had a lot of solder on the pads and it floated to this position.
I think part of the problem was that the board was in direct contact with the bottom tray, which I suspect was hotter than the air as it is directly above the elements. To counteract this I placed the board on a fine wire mesh which I supported with some scraps of aluminium I hand on hand. This isolates the board from the heat of the bottom plate and allows air to circulate all around the board.
|Solder Reflow Test 2 - Supported board|
This time I basically did the same thing with regard to temperature, only I paid a little more attention and didn't turn the oven up so high. The results were pretty good.
|Supported Board - Reflow Results|
The joint wasn't spectacular and it still rotated, but the solder was the right colour and the board wasn't discoloured. There was also a nice fillet on the ends of the resistor.
|Reflow test 2|
|Reflow Test 2|
By my third attempt I had a better idea of how much solder I needed. I also had a go at recreating the JDEC reflow profile. I put the board into the oven at room temperature then turned the oven on and set the thermostat to 150 degrees Celsius. This is my preheat stage. Once the thermostat cut out I let it sit for one minute. This is the soak zone. After that I ramped the thermostat up to 240 degrees Celsius and watched the solder paste. As soon as it liquefied I timed 30 seconds and turned off the heat. After another 30 seconds I cracked the door open for thirty seconds and then all the way. The board was then allowed to cool naturally.
The results were pretty impressive. It looked like joints I'd seen on professional boards. Positioned well, nice wetting, nice fillet.
|Reflow Test 3|
|Reflow Test 3|
|Reflow Test 3|
It's a bit of a learning process. I've got some finer tips coming for my solder paste dispenser that should give me better control over the paste placement. I think I may still have a little bit too much. It really goes a long way. I'm probably also going to get a thermocouple to monitor the temperature of the board during the process. At first it will be for manual control, but at a later point I might get an automated controller for tighter control of the temperatures.
It's not the greatest set-up, but I'm not doing BGA chips here. They're relatively large components on boards for prototypes. If the need arises later on I may improve things, but for now I'm pretty damned happy with it.