As I mentioned in some earlier random posts, CFL failure is few and far between, but they seem to temporally cluster sometimes. Maybe that’s a coincidence, but lets take a look at the innards of a few globes. I’ve looked at the IKEA 11w ES globe previously, here’s the ballast of the BC version. The BC version looks almost identical externally, except for the base – but interestingly the ballast inside is actually quite different.
The other CFL that’s failed recently is the Philips Genie 11w ES. It was a pretty popular globe. The tube itself is a newer “slimline” sort with thinner tubes, and thus it’s composed of three U-shaped sections. As it had been running for a long time, the adhesive that secures the tube had failed.
The circuit board also shows a similar desire to cram as much into a small space as possible. Hefty inductors, transistors and capacitors litter the board as expected, but interestingly, the tube is connected to the board by wirewrap!
The underside of the PCB suggests that the board had developed some high amount of heat before it failed. This is backed up by the yellowing of the CFL plastic shell. Also evident is the main filtering capacitor’s soldering joints which had cracked! From the colour of the components, it looks like a polyester cap may have failed by arcing through, as the tube doesn’t look blackened and the filaments are still show continuity.