search

OceanGlidersCommunity / Salinity_SOP

Salinity Standard Operating Procedure (SOP)
https://oceangliderscommunity.github.io/Salinity_SOP/README.html
Other
6 stars 7 forks source link

platform integrated sampling rates #71

Open isgiddy opened 2 years ago

isgiddy commented 2 years ago

Section 3 Data sheets associated with sensors provide power consumption estimates based on sampling rates. However these are not always the sampling rates that are typically used/possible during deployment.

CT sensors are limited by the processing system of the platform they are integrated with

This issue is associated with PR #55 and issues #49 and #48

isgiddy commented 2 years ago

Comment from @bastienqueste

Sampling frequency is actually more often a function of platform capabilities than sensor capabilities.

Seaglider limited to 0.2Hz unless inclusion of scicon board (which is no longer manufactured), because of old TT8 processor and single-thread processing. Uncertain for Slocum - I can't remember I haven't used them enough. SeaExplorer have a dedicated ARM processor for handling sensors and can currently handle 16Hz for RBR Legato for example.

Analog sensors can essentially be sampled as fast as the platform can handle. Sensors that do onboard analog to digital conversion are limited by their A2D converters (generally 10s to 100s of Hz so not actually the bottleneck). Most often, the issue is if they do any onboard calculations and logging, which slows it to several Hz or 10s of Hz.

The only real limitation is the platform.

Typical settings for salinity are 0.2Hz for Seaglider, ~0.5 to 1Hz for Slocum if I recall correctly, and ~2Hz for SeaExplorer for the average user, and up to 16Hz for power users (or anybody with a fast Legato set up to log onbaord the sensor).

isgiddy commented 2 years ago

Comment from @gkrahmann

As far as I remember for old Slocums with a persistor the default sampling rate is 1 every 4 seconds, i.e. 0.25 Hz. This can be changed to 1 every 2 seconds or 0.5 Hz when the science persistor has 'free' time, meaning when you do not have too many other sensors. I am not sure, but I think the new processors in the G3s can go faster. While we have a G3, it is a hybrid with an old processor. So the text in the main md is wrong for Slocum unpumped (as Pierre wrote). There 4 Hz is mentioned when it should 0.25 Hz... As Bastien wrote, the platform has until recently been the limiting factor, with newer gliders getting their recording speeds into the ballparks of speeds of the CTDs (which I am not sure of but should be 0.5 Hz or faster).