3 ms·

It is a ruggedly designed... I get flashbacks to the college intern I interviewed that excitedly told me about the RPi telemetry design he created for a rocket

It is a ruggedly designed... I get flashbacks to the college intern I interviewed that excitedly told me about the RPi telemetry design he created for a rocket. Got it all coded up and working great. After some prodding he confessed that it failed on launch. The micro SD connector couldn’t take the vibration.

dTal 6y ago
That's odd - I've flown RPis on drones with far more vibration than the relatively smooth linear acceleration of a model rocket. In such setups it's recommended to directly solder power and USB connections, but I've never heard anything about the microSD connector. Did he say how he diagnosed the failure?
- DJBunnies 6y ago
  Maybe it was mounted poorly and the g force ejected it?
  - elcritch 6y ago
    A little lock tight would probably solve the problem too.
  - dTal 6y ago
    They only weigh half a gram; it would take hundreds of Gs.
- joezydeco 6y ago
  The suspected failure mode was vibration of the connector’s metal fingers away from the SD card’s traces, breaking off power and/or data.
  - skykooler 6y ago
    Good to know for high-vibration environments. I may try epoxying the SD card into place for my telemetry experiments.
    - joezydeco 6y ago
      Maybe think about upgrading to an SBC that uses eMMC. Costs more, but less trouble.
glouwbug 6y ago
Sure, but this thing isn't going to space. It probably as durable as a Thinkpad with an 7200RPM HDD.
polishdude20 6y ago
Yep, the connectors on those things can have intermittent contact. If you're uploading data to the SD card live while in flight, those intermittent connections can screw up your data. The workaround is to use a flash chip to write live data to it and then once the rocket lands, dump all of it to the SD card.
asdasfasdfasdf 6y ago
could probably hot glue the SD card in place, I used to use a lot of hot glue for extremely rudimentary waterproofing

dTal 6y ago

That's odd - I've flown RPis on drones with far more vibration than the relatively smooth linear acceleration of a model rocket. In such setups it's recommended to directly solder power and USB connections, but I've never heard anything about the microSD connector. Did he say how he diagnosed the failure?

DJBunnies 6y ago
Maybe it was mounted poorly and the g force ejected it?
- elcritch 6y ago
  A little lock tight would probably solve the problem too.
- dTal 6y ago
  They only weigh half a gram; it would take hundreds of Gs.
joezydeco 6y ago
The suspected failure mode was vibration of the connector’s metal fingers away from the SD card’s traces, breaking off power and/or data.
- skykooler 6y ago
  Good to know for high-vibration environments. I may try epoxying the SD card into place for my telemetry experiments.
  - joezydeco 6y ago
    Maybe think about upgrading to an SBC that uses eMMC. Costs more, but less trouble.

glouwbug 6y ago

Sure, but this thing isn't going to space. It probably as durable as a Thinkpad with an 7200RPM HDD.

polishdude20 6y ago

Yep, the connectors on those things can have intermittent contact. If you're uploading data to the SD card live while in flight, those intermittent connections can screw up your data. The workaround is to use a flash chip to write live data to it and then once the rocket lands, dump all of it to the SD card.

asdasfasdfasdf 6y ago

could probably hot glue the SD card in place, I used to use a lot of hot glue for extremely rudimentary waterproofing