Hi folks,
I used a TeleMetrum as the primary controller in my successful NAR L3
certification flight a couple of weeks ago. (Woohoo!) Folks at TCC's
October Skies really got a kick out of the voice synthesis flight report
over the PA system. In fact, I'm pretty sure I sold a few of them. :)
The flight was nearly perfect, but when I gathered all the data, I found
that the TeleMetrum had actually reset in mid-air, just as it blew the
charge for the main parachute. I suppose its job was done, but I'd really
like to root-cause the issue before my next flight.
Data points:
- This was the first flight of this TeleMetrum unit (#340) and of this
rocket, but I've flown a different unit (#301) several times in a smaller
rocket, so I'm pretty familiar with the TeleMetrum.
- The TeleMetrum, TeleDongle, and OSX AltosUI were upgraded to the latest
release (1.02).
- The TeleMetrum was connected to J-Tek e-matches, which set off Rouse Tech
CD3 CO2 canisters for both drogue and main.
- I had used the TeleMetrum's remote firing feature earlier in the day to
ground-test the drogue and main with no apparent resets.
- I used a Raven connected to BP pots for backup, so it was completely
independent. The Raven recorded a normal flight (though a misconfiguration
caused a slightly early drogue deployment).
- The TeleMetrum has the stock LiPo battery, which was fully charged
shortly before flight.
- The TeleMetrum is bolted to a G10 sled with nylon spacers at all 4
corners. The battery is encased in a cut-out fiberglass "sandwich." There
was no indication that anything moved around in flight.
- The power switch is a rotary voltage selector, soldered to a wiring
harness made of 20-gauge stranded copper and a crimp-on 3M connector.
- The TeleMetrum connects to terminal blocks outside the av-bay bulkheads,
which are sealed to prevent blast leaks.
- BP and CO2 residue suggests that the Telemetrum did pop the main CD3
before the Raven set off the main BP, just as intended.
- The TeleMetrum only contained one flight recording, #001, and the last
line in the resulting CSV file is the transition to "main" state. There
were no dips in recorded battery voltage, even during the drogue firing.
- The ground station actually recorded two valid .telem files, flight #001
and #002. The second file appears to be from the descent on the main, but
lacks GPS coords.
- The ground station also recorded four other .telem files, but they look
bogus. The file names indicate flights 21954, 22054, 22154, and 22254, and
they each only have one or two data points.
- My on-board video camera picked up the beeps from the TeleMetrum
resetting into "Pad" mode as the main deployed. It then beeped the "no
continuity" warble the rest of the way down.
- Data from the Raven suggests that the deployment of the main parachute
wasn't abnormally violent -- about 20 Gs. (The CD3 gives a relatively soft
"poof.") The drogue deployment was actually worse (50 Gs) because the
Raven blew the BP early.
Collectively, the data suggests the reset occurred right at the deployment
event, seemingly before the main parachute even inflated. The only causes
I can think of are:
1) Momentary loss of power from a physical break/short in the battery or
switch wiring
2) Voltage drop caused by the current draw at ignition
3) Firmware crash/reset
None of these seem very likely. 1) I'm fairly confident in the wiring
harness and connections. 2) The Telemetrum had tested these same ematches
without issue, and the battery voltage didn't dip at all for the drogue.
3) I've never noticed a firmware crash, but if it did, would it record it
in a log somewhere?
I attached the .eeprom file, since it's pretty small. I'm happy to post
the other files, if useful. Lacking anything better to go on, all I can
think to do is maybe hang a big capacitor across the power terminal. I'm
hoping you guys will have better ideas. :)
Thanks,
Casey Barker
