Is surface tension enough?

Oven test part 2.
The question I have set about asking is, Is surface tension enough to keep components on the underside of a board whilst it is reheated to solder the top layer.

If your in a hurry: Yes it is.
If you have a few seconds longer, you may like to observe the evidence:

I baked some components onto a bit of gash stripboard (more about this later). It was a good job did this test as i discovered that the solder was all fully melted when the thermometer read about 160degC – really? i was expecting this to be at about the 200degC point. Ok is the thermometer lying? – it was cheap ebay fodder. Cue up the pan of boiling water. This came out at 99devidenceegC so within the tolerances I need. For now I will assume that the discrepancy at high temperature is due to the probe having a greater thermal mass and so lagging in temperature rise.

Step two was to mount the stripboard inverted so the components can fall off under gravity and repeat the heating step. With some solder paste on the top side to verify that a sufficient temperature has been reached, the parts behaved admirably. None have fallen. I pushed the heating step for longer than needed to make sure, this lead to my second discovery of the evening. When over heated, strip-board omits lots of smoke and fumes. Quite disarmingly, burnt stripboard smells of Naan bread! What does this mean, will i be able to eat naan in an Indian again?

Oven Calibration!

In past years, the student robotics hardware has been fabbed by Pcb/Assembly houses. Partly encouraged by the quality of last years results, I decided to undercut the quotes that the companies offered and build the boards myself.
As I will be fitting lots of surface mount components I decided its high time I investigate Reflow ovens, well ovens for reflow. There are a fair few people expressing opinion on what sorts of ovens are acceptable. So where do you buy cheap reflow ovens….. well argos of course! I dropped in on the way back from a job and picked up the cheapest heating enclosure they had – A cookworks desktop oven (423/0285) £20. On first examination, its exactly as expected – Made of badly fitting pressed toffee. In fact its so flimsy one could be excused from mistaking the oven housing for its packaging!

Anyway, to the more important issue, what is its natural temperature profile? Cue up the recently purchased digital oven thermometer, a spreadsheet and a 5 second beeper (watch -n5 xkbbell). The oven was powered on at full power until 230 degrees C. At this point the door was opened and power removed. The results are shown below:

Also on the chart are some recommended warm-up rates and an approximation to the profile suggested by a solder paste datasheet. It seems that the oven is close to the upper limit of commercial reflow ovens(1.5degC/s) and well under the rate the Spark fun guys suggest as a limit (<2.5degC/s). So what is the rate of temperature change. Taking just the linear section of the graph, a line was fitted:

Well it works out at 1.75 degC/s not bad but possibly a little fast. It looks promising for a pretty good profile if the heating is stalled briefly at about 180degC. This delay in the prescribed curve is apparently to let any gasses in the components to escape before the part explodes. So testing in anger with solder tomorrow. Also that question you’ve always wanted to know. Is surface tension enough to stick components to the underside of a board when the topside is being reflowed? More results soon.