Clear skies and another day with lots of sun. Noticeable increase in smoke/haze, although the impact on visibility/incoming solar was pretty minor.
Onsite: Matt and Mack
Today:
- Data check identified no obvious issues.
- It looks like Rsw.in.t2 is being cut off at 1028 W/^2, but I wasn’t able to identify any reason why. Also, a morning shadow is also evident as a dip in SW around 7am on days with clear skies.
- Jacquie arrived this evening in time for dinner.
3 Comments
Jacquelyn Witte
This is what Matt is referring to:
Rsw.in.t2 (purple) flatlines at its peak. This started back in 30 July.
Steve Oncley
Yep. It is flatlining. The message is:
NR126 1027.960449 8388607 242.400100 1834419 -160.914718 -851464 141.041198 872305 36.022613 5013716 0.246946 828614 35.06\r\n
Where the A/D count of 8388607 = 0x7FFFFF or the maximum count of a 23 bit A/D converter. Apparently, the response of the Rsw.in signal is higher that the input range of the A/D. This isn't the case for t49, which is producing an A/D count of only 7119231 = 0x6CA17F. I'm not sure why the voltage gain for one NR01 is so much larger. Chris may be able to change the input range for this signal.
Steve Oncley
I appear to have fixed it by logging into t2, rs 7, then "@NR:ADG 2 3 3 3", which changed the gain setting on the NR01's PIC A/D for the Rsw.in channel from "3" to "2" – apparently a factor of 4 smaller. It seems that the PIC code on the NR01 correctly compensated for this gain change in computing voltages, so the resultant output looks okay (now 1037 W/m^2, whereas t49's is 1038).
We will want to remove the clipped values prior to now during QC.