On Fri, Nov 16, 2007 at 03:48:32PM -0800, rq17zt wrote:
> [Slight rant]: I _still_ wish the timestamps were recorded in a manner
>  that was crystal clear. I'd like to have the automatically parsed
>  logs begin each line with, "2007. <data>...",
>  etc.

You know if we had a simple record count from the MCU we could use that
as a time stamp, given the fact that the MCU data sampling and reporting
didn't suffer from the latencies that the Linux system did.

I'm not convinced this would work yet as there are more data records
than just acceleration in the telemetry stream, but one thing to do is
to not use the time stamps of the Linux box and use an external clock
for the dynamic and navigational computations.

>  > Some one should run the x/y acceleration data through and FFT to see
>  > if there are any acoustical effects getting sampled.
> This has been done for LV1b data, but the results were just as noisy
> as the time domain. I'm not sure this means there's nothing
> there. Perhaps some smoothing is required. Possibly one of the more
> popular windowing filters?

I'll take another look at the frequency domain this weekend.

There may not be any x/y data to be had other than to really understand
what we have and why its the way it is, and work on making sure next
time we get data thats more useful.

This data should be looked at as if it holds valuable secrets.  Within
in it are system issues that need to be addressed that will influence
you avionics, payload, and software designs.  Its full of gold.


Attachment: signature.asc
Description: Digital signature

psas-software mailing list

Reply via email to