compass measurements in HR mode

Hi,

I noticed something odd when processing our HR data recently.  We were sampling a single beam at 8Hz with a compass update rate of 1s.  The data strings for heading, pitch and roll were all the same length as those for velocity i.e. 8 points a second, so I assumed I'd just see one value repeated 8 times for each second.  However, if you look at the attached plot of a few seconds of data you'll see this isn't the case.  The pitch and roll data is interspersed with zeros (we know that around -25 is correct for the pitch measurement as that's what we set it to) and the heading data is interspersed with 90s.  There are typically 2 or 3 'correct' or measured values a second and these occur at the same periodicity (so I could just take these values and replace the zeros but I'd rather not).

This happens for both of our HR instruments, but only when they are in downward looking mode.  Perhaps there's a software bug somewhere?  Couldn't be my coding surely?

Any ideas of what's going on?

Cheers,

Julia

Hi Julia,

The compass update rate is a bit misleading when you do burst measurements. During burst measurements the compass (including pitch and roll) is in fact measured once every sample. The difference in behaviour depending on the instrument orientation makes me wonder if there is a problem with the connection between the compass and the main board. Have you checked the error code to see if it indicates compass error?
Can you upload a .hdr-file so we can take a look at the firmware version, serial number etc?
Or even better a small data file with this problem?

Cheers,

Sven

Hi Sven,

So the error code does indeed give us a value of 1 for those points when the compass and tilt measurements are zero/90.  I'll upload the hdr file (can I email you a matlab data file with data from 2 bursts -- it's 1.4mb?).    Isn't it unlikely to be a connection problem if it happens with both instruments?  Also they don't seem to have the same problem when deployed in 3 beam burst mode.

Cheers,

Julia

Hi Julia,

You are absolutely right, I missed the fact that this happens on both instruments. Not to mention that we very rarely see connection problems in the first place. I might need a couple of days to track this bug down, I will keep you posted. Regarding data files we always want the raw binary files when looking at data so we can get the whole picture. I will send you instructions so you can upload the raw data file to our FTP server if you want me to take a look at it. If only bit 0 of the error code was set your velocity data are most likely fine since the data was taken in beam coordinates. If you plan to use the heading, pitch and roll values you must correct the wrong values the way you suggested.

Cheers,

Sven