Yeah, it appears, in all probability, from your description that the voltage spikes are causing this abnormal behaviour. Else, how would the data be written to a 0 when the unused value is a 1 ?
So, as suggested by Zeeglen, get hold of a medical centre to help you in exposing the tags to a bunch of X-Rays & simulate the error to correlate with the actual findings.
Actually we haven't isolated the failure to any component yet -- the PV cell just seems most sensitive to radiation. Good point about the scanner, though I wonder how much RFI it would generate nearby.
Only the tags that pass initial testing are deployed; these failures are observed after they've been in use.
Excessive sunlight? The PV cells handle a 10 000 lux desk lamp with no problem ("bright sunlight" is supposedly around 6000). There's a series diode protecting the battery from discharge through the LDO.
Only thing I can think of adding to the controlled test is the cold and shock environment when attached to a downhill ski ;-)
From your description it sounds like you have determined that the PV cells are the failing components. Didn't think carry-on goes through the full body scanner along with the passenger.
Are the failures discovered before or after they have been in use? Are they tested by the user before attaching to the birds? If they fail in use could be excessive sunlight. Is there a focusing lens that might be overheating the cell? Is there any chance a fully charged battery could be reverse biasing back through the LDO in the dark?
For Xray test can you get a local medical clinic or dentist to zap known good tags then retest?
Or better yet try a more thorough controlled test. Have your boss approve a field trial at an environmentally hostile place such as Hawaii or Bahamas or Mediterranean or Rio. Send some of your tags UPS, FedEx, some checked baggage, some carry on. Test them all at arrival. Then bring the good ones to the beach and let them soak up the bright sun for a few days of more testing. (Don't forget your suntan lotion). After about a week you can return home, repeat the tests on the remaining good units from the beach, and write your test report. Would be interesting to see whether it is Xrays or sunlight causing the failures.
Rich, can you supply a little more information of the failure mode? What element of the logging tags is failing? Do they fail the same way each time?
If the tags are tested good, then shipped and arrive with failures it could be simple electrostatic discharge, especially if the receiving site has naturally low humidity.
Part of my work involves designing small solar-powered logging and telemetry tags for studying bird migration, etc. We've encountered some mysterious failures in logging tags and are wondering if some could be due to X-ray irradiation during air shipment (FedEx, checked and carry-on baggage inspection.) We feed the output of an 8-pin SOIC PV cell through an LDO regulator to sense light levels and charge a lithium coin cell which powers a PIC microcontroller. I wonder if the PV cell might be particularly sensitive to X-rays (or the RF from full-body scanners?) -- the manufacturer hasn't done any X-ray testing, citing prohibitive cost. Can anyone point me to some info and guidelines on the subject?
Rich Gabrielson, Research Engineer
Bioacoustics Research Program
Cornell Laboratory of Ornithology
95 Brown Road Mail Slot 1012
Ithaca NY 14850
Blog Doing Math in FPGAs Tom Burke 18 comments For a recent project, I explored doing "real" (that is, non-integer) math on a Spartan 3 FPGA. FPGAs, by their nature, do integer math. That is, there's no floating-point ...