CDP Problems and Solutions

Similar to 2016, In 2017 the University of North Dakota (UND) Cloud Droplet Probe (CDP) failed to accurately size droplets after a few hours of flight time likely due to an alignment issue. The inaccurate sizing of droplets was evident by shape and concentration of the droplet spectrums.

The DMT spinning pin hole was used to test the UND CDP alignment. No counts were ever found using the spinning pin hole on the UND CDP. However, counts were obtain using the spinning pin hole on the Weather Modification International's (WMI) CDP. Counts were found to be between 29-30 in channel 6, which seems to indicate a possible alignment issue with the WMI CDP since the pin hole is between 18 and 22 um which should peak in channel 15 or 16. Additional, testing of the WMI CDP with beads will be conducted. Our understanding is that getting 29 counts in channel 6 is not the result of not having the spinning disk position correctly. While some changes in position were required to get the counts to move from channel 4 to channel 5, and then to channel 6; a correctly working/aligned instrument would not get 29 counts down in the 4 to 6 channel range. Hence, if bead tests confirm an issue with the WMI CDP, the instrument will likely have to be sent to DMT for repair/adjustment.

Note that when obtaining the counts on the WMI CDP, the different adjustment direction could not be locked into place. When locking an axis in place cause the counts to go to zero. Hence, I believe that locking a movement direction into place results in not being able to get counts. All three direction should be unlocked. The configuration of the different parts of the spinning disk does not even allow all three direction to be locked. The spinning disk configuration was changed; however, does not seem possible to enable all three axises to be locked and allow the disk to be within the beam.

December 22 Update: Today we tighten up the locks on the spinning disk so everything is a lot tighter. Once the locks were tight, we were able to get the WMI CDP to peak in channel 6 and be able to move the spinning disk around a lot and still get the peak. Also, was able to get counts with the Depth of Field correction applied. The UND CDP did not show any peaks, in any channel.

The UND CDP was pack-up to be sent to DMT for repair/alignment. The WMI CDP will be tested using beads.

During the POLCAST3 flights (Summer 2010), the Cloud Droplet Probe had sharp, uncharacteristic drops in the size spectrum. During the July 20, 2010 flight, a 40 second period shows that the 11-12 um diameter channel is unusually low (See plot below.)

This 40 second period, show relatively high (for this flight; however, low when compared to previous FSSP measurements in similar conditions) total droplet concentration; however, the integrated liquid water content is low compared to the King Hot Wire probe (See plots below).

The King Liquid Water Content is given by solid black line and the Cloud Droplet Probe Liquid Water Content is give by the solid blue line.

The December 16, 2009 calibration by DMT using 30 um beads show a good (no large dips) spectrum (See Plot Below.)

However, the 30 um performance check after the completion of the POLCAST3 project shows a low count in channel 18, next to the peak in channel 19 (See Plot Below). Channel 19 corresponds to size 24-26 um using the Pulse Height values sent to the probe and standard conversion.

Likewise, the December 16, 2009 calibration by DMT using 15 um beads show a good (no large dips) spectrum (See Plot Below.)

Here the 15 um performance check after the completion of the POLCAST3 project shows a low count in channel 10, next to the peak in channel 9 (See Plot Below). Channel 9 corresponds to size 10-11 um using the Pulse Height values sent to the probe and standard conversion.

The problem with the spectrum shape was related to using the wrong threshold values for the CDP. The A/D thresholds values (64,89,115,147,168,188,220,262,308,356,407,461,583,707,829,983,1148,1324,1512,1697,1909,2131,2365,2610,2864,3097,3337,3583,3879,4095) were updated on October 18, 2010.

Using these correct thresholds, the following perform check data were obtained for 15 um beads.

While for 30 um beads we get the following.

The 30 um beads peak in the 24-26 um channel (18); however, the 15 um beads peak in the 10-11 um channel instead of the 13-14 um channel as the DMT test have.

The 15 um beads (Exp Nov 2010) don't give a valid spectrum. On October 20, 2010. three different 15 um beads gave peaks in 13-14 um channel. Below is a plot showing the comparison between the poor 15 um spectrum (red line) and a good 15 um spectrum (black line).

During the September 25, 2010 flight, the CDP did not record data correctly. Data was only recorded in channel 10.

The M300 table was setup to acquire 196 bytes of data which is for 40 channels. At 10 Hz. This is probably to many channels to fast. Don't know why this was set like this. Need to change to 30 channels.

The M300 tables were configured correctly for 30 channels.

During the March 26, 2010 flight, the CDP did not record data correctly. During this flight, the data system was configured to record the particle by particle information.

The cause is probably related to the M300 data system table configuraiton.

On April 21, the tables were board table was changed to match the test example table. This cause the byte size error to not occur; however, the table stopped display any information include the probe voltage.

Update June 25: The configuration differences for the CDP for particle-by-particle and non-pbp are:

The cloud droplet probe has been move to the new serial acquisition card. However, while the CDP digital output can be set to 460,000 bsp, the fasted we could get the card to works is at 230,000 bsp.

During January 14, 2022 IMPACTS Flight, before take off, the CDP dump spot monitor was reading 3.17V. Right after take off, the CDP dump spot monitor was reading 0.0V and continued to read 0.0V for the remainder of flight. On January 15, 2022, the CDP was taken out of the hot wire boom, and sent to DMT for testing.

DMT sent an long-arm version of the DMT CDP in temporary place of the UND CDP. On January 24, 2022, aircraft testing was done on the long-arm CDP. The long-arm CDP control box was installed in the hot wire boom, and tested; no issues. The long-arm CDP probe head and control box were installed in the hot wire boom, and tested. After power was turned on for the long-arm CDP, the laser turned on and the dump spot monitor was reading 2.76V. The long-arm CDP probe head was taken out of hot wire boom and is waiting on NASA approval to reinstall (Jennifer Moore - 1/24/2022).