Deployment 1 - PAR ball bent upon recovery of the glider

By Lori Garzio, on 9/1/16

CE01ISSM CAMDS stopped working

The CAMDS experienced some kind of failure wherein the sampling schedule suddenly changed from every 4 hours to approximately every 5 minutes. The camera stopped taking pictures 10/14 after either blowing through its batteries or disk space. This problem has not been seen before.

By Mike Smith, on 12/6/17
Redmine Issue #12957

The high voltage lines providing power to the MFN via the CVT for CE09OSSM-00004 shorted out yesterday 2017-01-22 at 08:31 UTC:

<snip from 20170122.pwrsys.log>
2017/01/22 08:30:06.391 PwrSys psc: 24.87 2933.33 96.5 0100 00000000 00000000 ... cvt 1 376.31 142.50 1 13.65 0000000c No_FC_Data 526d
2017/01/22 08:31:06.395 PwrSys psc: 24.55 4253.33 96.5 0100 00000000 00000000 ... cvt 1 56.45 1902.50 1 13.84 0000000c No_FC_Data 51dd
</snip>

Sea state was high, but not unusual for this time of year (see attached plot). Have disabled future power scheduling to the MFN and applied an override to disable the CVT. No further data will be logged from the instruments on the MFN. This affects:

On DCL35:
* ADCPS (had already failed upon deployment)
* PCO2W -- should continue to autosample
* PHSEN -- should continue to autosample
* PRESF -- should continue to autosample
* VEL3D -- should continue to autosample
On DCL37
* CTDBP (plus DOSTA) -- should continue to autosample
* OPTAA
* ZPLSC -- had stopped reporting data early on in deployment, may still be autosampling

By Mike Smith, on 3/23/17
Redmine Issue #11837

The PAR (PARAD) and ECO triplet (FLORT) was later turned off on 2017-03-02 to further prolong battery life to recovery date.

By Mike Smith, on 3/15/17

The ZPLSC internal clock on CE01ISSM is slowly drifting out of alignment with GPS time. On two occasions now, I have had to shift the sampling window on the DCL to catch the reported condensed profile. Most recent data shows and estimated offset of approximately 1 minute 6 seconds.

CI needs to develop methods for correcting internal instrument clock data when working with recovered instrument data for cases where the recovered instrument data provides more information (e.g. full acoustic profiles rather than condensed). If left uncorrected, the recovered instrument data could be out of alignment with co-located sampling.

By Mike Smith, on 1/30/17
Redmine Issue #11809

The ZPLSC clock has drifted approximately 45 seconds, putting it outside of the original sample window. Sent a new mission file yesterday to change the time window. Went a little too far and have backed it back today. DCL37 will now turn on the ZPLSC port at 35 minutes after the hour and
leave it on for 2 minutes. ZPLSC is reporting it's first condensed profile at ~36 minutes and 13 seconds after the hour. Will need to monitor over the course of the deployment to make sure any further clock drift is accounted for. But, should be good for now.

There will be a data gap of about 7 days in the telemetered data. The recovered data will have the full record.

By Mike Smith, on 11/16/16

The ZPLSC scheduled sample time has shifted out alignment with the DCL sampling window. Last sample was on 2016/09/26 17:36. Sent new mission file changing the sample window on 2016-10-13, but that did not correct the issue. Will send new mission file with broader sample window.

Have attempted to expand sample window. Will see if that corrects issue

By Mike Smith, on 10/20/16
Redmine Issue #11404

Instrument not deployed during Fall 2016 cruise because its refurbishment was not complete before the cruise.

By Mike Smith, on 10/25/16
Redmine Issue #11427

Instrument not deployed during Fall 2016 cruise because its refurbishment was not complete before the cruise.

By Mike Smith, on 10/25/16
Redmine Issue #11427

Instrument not deployed during Fall 2016 cruise because its refurbishment was not complete before the cruise.

By Mike Smith, on 10/25/16
Redmine Issue #11427

For no apparent reason, the CAMDS stopped reporting data on Friday October 21 at 12:30 UTC. Do not know if the system is continuing autonomously.

By Mike Smith, on 10/25/16
Redmine Issue #11434

The Power System Controller on CE07SHSM indicates the power connection to the MFN short-circuited on 2016-08-12 at 23:30 UTC. Failure is indicated by decreased voltage (70 VDC instead of 380 VDC) and increased current draw (1900 mA instead of 120 mA) to the CVT. This symptom is indicative of a short-circuit in the power lines of either the stretch hose (most likely) or EM Chain. In the past, this failure has been associated with a failed stretch hose.

Have disabled all further power and communications with the MFN. This means no more telemetered data from:

ADCPT2
VEL3D
CTDBP2 + DOSTA
PRESF
PCO2W
PHSEN2
OPTAA2
ZPLSC

By Mike Smith, on 10/20/16
Redmine Issue #10988

The Power System Controller on CE07SHSM indicates the power connection to the MFN short-circuited on 2016-08-12 at 23:30 UTC. Failure is indicated by decreased voltage (70 VDC instead of 380 VDC) and increased current draw (1900 mA instead of 120 mA) to the CVT. This symptom is indicative of a short-circuit in the power lines of either the stretch hose (most likely) or EM Chain. In the past, this failure has been associated with a failed stretch hose.

Have disabled all further power and communications with the MFN. Therefore, there is no telemetered data for this reference designator.

By Mike Smith, on 10/20/16
Redmine Issue #10988

pH data was looking really good until 2016-10-14 10:30 UTC. Possible fouling, although also possible system was damaged during large storm event.

Will need to monitor to seeing if fouling clears.

By Mike Smith, on 10/19/16
Redmine Issue #11391

> Plots of the pH data for CE02SHSM indicate that unit may be experiencing biofouling, causing increased variability in the data signal.

Whatever was causing the spiky signal appears to have cleared and resolved itself. Signal looks good for now.

By Mike Smith, on 10/14/16