Posts on the Alpisto engineering blog filed under #spl07.
Everyone who builds these devices knows you must not solder near the barometer, and that the third rework kills it. Nobody says by how much. This one came off the bench at 470 Pa/K against a rated 0.5, inventing 14.5 metres of climb while sitting still. Thirty lines of firmware removed 95 % of that and every bench test agreed — until somebody picked the device up and the vario went the wrong way. The number behind the folklore, and why the fix helps an altimeter and provably cannot help a variometer.