Sunday, June 16, 2013

Reflow Oven, Stencil, Guelph Triathlon


I’ve been testing my reflow oven setup and I’m having mixed feelings on using this specific toaster oven. Essentially it maxes out at 1 deg / sec but it can go a little faster if you “pre-heat” the oven. What this comes down to is that a cold oven has to heat up the metal and glass. If you pre-heat those things they will lose heat to the inside and the outside of the oven. This doesn’t help at high temps but at low temps it can allow it to match the curve better. I am concerned about the final heating to melting. I am afraid that it will take to long to heat up and therefore damage the FR-4 board. I’ve seen people do some real damage, scorching them for instance.


As you can see, doing this allows it to follow the intended curve more closely. This is at the limits of the generic lead reflow curve though.

While I ordered the circuit boards (6) from OSHpark, I needed a reflow stencil. This unexpectedly arrived very early compared to when I thought it would arrive. I chose Pololu’s slowest service, 5 day turn around. Shipping to Canada was about as expensive as the stencil, but was next day I think and I never chose the highest shipping available.


Below you can see the stencil. The QFN and peculiar Balun look like they are on the verge of not being usable, though I have no idea or experience. This has more to do with the limitations of laser cut out mylar I think. There was no bridging on the QFN cutouts. I don’t suspect I’ll have a problem with their work. I tried the Balun out to compare as it’s a non-standard layout. It is so small. I am having concerns about it and the 0402 components. If this backfires I’ll redo it with 0805 and a component based Balun. This means the board will get a little bigger but means that only the QFN would have to be reflowed.


During the weekend I attended the Guelph One Triathlon (as a spectator, due to my terrible swimming – if I don’t practice consistently I can’t make it around without drowning a couple of times). I did an unscientific survey of powermeters. Total count: Four. Three were Power2Max and one was Quarq. I was blown away.


I didn’t think many people would be running Power2Max in North America, let alone for it to be more common than Quarq or SRM. This was only the sprint distance, I didn’t attend the Olympic distance and suspect I would have seen more powermeters. When I’m closer to the Beta test group units I may start hanging out at Tri’s with a Booth to let people demo my device. However, this all leads me to feel more confident that there is market room for a $700 dollar unit that includes a crank – and not just a single arm that fakes the other side.


  1. I'd bet $5 that all three of those got them through some sort of team deal - hence the randomly high concentration. I'd agree, that sampling seems odd. Usually PT's are far above.

    1. There likely was some PT's but they are quite a bit harder to spot as people where running in and out with their bikes. I guessed it was due to a LBS sourcing them, but team deal seems like a good theory.