Does anyone know what conditions cause an OBS-3 to return NaN instead of a number for turbidity? Does this reflect extremely high turbidity? OBS is connected to a CR1000, programmed with ShortCut.
Which output option did you purchase and post the program section you are using to measure the sensor.
I do not know which output option was purchased, as the instrument is rather old and predates my arrival at the lab. All I know is that it is an OBS-3, not a 3+.
Here's the applicable portion of the logger program:
Scan(1,Min,1,0)
'OBS3 measurements
PortSet(9,1) '9=switched 12 volt
Delay(0,3,Sec)
VoltDiff(FTU,1,mV5000,2,True,0,_60Hz,0.4,0)
'Multiplier = (2000FTU)/(5000mV)=0.4
'measures differential channel 2
Next Scan
This program fragment was made in Short Cut.
* Last updated by: SMM on 10/2/2010 @ 8:41 AM *
The OBS-3 manual says the standard output was 0-5 Vdc. Have you checked your wiring against the OBS-3 manual?
A suggestion to see if the sensor is working would be to:
1.Remove the red wire from the SW+12.
2.Remove the White and Green wires from the logger and attach them to a volt meter set to measure DC volts.
3.Connect the red power wire to a +12V terminal on the logger.
Wave you finger in front of the sensor optics and see if you get any indication on the meter.
I *do* get data most of the time, but sometimes, when I would expect turbidity to be very high given the weather and tide stage, the sensor outputs NaN.
I see.
Section 19.1.4 of the CR1000 explains various reasons and possible solutions to NaN such as over-range and open sensor input. If they are very infrequent you can tell your output instructions to ignore them in the averages.
The OBS3 could be outputting >5V at high turbidity levels thereby over-ranging the datalogger. You could trap the NaNs with an if statement and if encountered replace with the maximum NTU the probe is designed to measure.