Discovery

Test 3 was precipitated by a serendipitous slowdown of the ship to change which engine was being used for the autopilot. Electrical interference was detected when the ship slowed down.

(click here for full size)

This figure has temperature of os75 in the top panel, and a deep signal return in the second panel. The next panel is beam 1 velocity from the os75 narrowband mode. The last panel is beam1 speed in a shallow (uncontaminated) reference layer. The horizontal axis is floating point day of year (UTC), in this case mostly March 18.

Since the electrical interference was apparently back, we immediately started testing. This time, we started with the usual long RDI cable and the deck unit in the rack. We worked out a plan with the engine room to test the two engines engines in autopilot and to try different power sources. The electrical noise was persistent.

(full size: "part08" discovery of interference)

(full size: "part08" discovery: beam velocities and processed data )

(A) Engine and Power Tests

We tested the use of autopilot on either port and starboard engine, and tested a variety of power sources (main lab UPS, rack-mounted UPS on clean power or dirty power) and the electrical noise was persistent. These tests uwed the long RDI cable, with the deck unit and the power all in the computer lab. The interference is identifiable because (1) it occurs below the range that the instrument should normally acquire good data, about 700m in this case, and (2) it is strongly biased. The bias is most obvious when the ship is sailing at 5-8kts.

(full size: "part10" test 3A - beam1 velocity)

This figure has temperature of os75 in the top panel, and a deep signal return in the second panel. The next panel is beam 1 velocity from the os75 narrowband mode. The last panel is beam1 speed in a shallow (uncontaminated) reference layer. The horizontal axis is floating point day of year (UTC), in this case mostly March 20.

(B) Laundry room, short RDI cable, UPS

With the short RDI cable and the Rack-mounted UPS, we collected data in the laundry room. This data collection showed no signs of the electrical interference.

(full size: "part11" test 3B - 4 beams, no interference)

This figure shows panel plots of beam velocity for the OS75 narrowband mode. The vertical axis is depth in meters, and the horizontal axis is floating point day of year (UTC), in this case mostly March 13.

The range is about 700m and there is no sign of any signal deeper than 700m, therefore no sign of electrical interference.

(full size: "part11" test 3B - beam 1, no interference)

This figure has temperature of os75 in the top panel, and a deep signal return in the second panel. The next panel is beam 1 velocity from the os75 narrowband mode. The last panel is beam1 speed in a shallow (uncontaminated) reference layer. The horizontal axis is floating point day of year (UTC), in this case March 20.

The range is about 700m and there is no sign of any signal deeper than 700m, therefore no sign of electrical interference.

(C) Computer lab, UPS, long RDI cable

Collecting data again with the long cable revealed the interference was still present.

(full size): "part12" test 3C - beam 1, and processed

On the left of this figure we have temperature of os75 in the top panel, and a deep signal return in the second panel. The next panel is beam 1 velocity from the os75 narrowband mode. The last panel is beam1 speed in a shallow (uncontaminated) reference layer. The horizontal axis is floating point day of year (UTC), in this case mostly March 20.

The processed velocities are on the right, with E/W ocean velocity in the top panel and N/S velocities in the second paanel. Backscatter is in the third panel, and Percent Good is in the last panel. The horizontal axis is floating point day of year (UTC), in this case mostly March 20.

Extra Bonus: Temperature Stability

Plotting the temperature for the whole cruise showed that when we did not see interference, the temperature deviations were very small. But when interference was present, the temperature deviations were larger. This was particularly striking in the last short-cable test. The temperature deviations before and after the short-cable test were high (when the long cable was used and the electrical interference was obvious). But during the short-cable test, when no interference was visible, the temperatures were stable.

(full size: TN400 temperature and interference)

This figures shows OS75 temperature in the top panel, with each separately-logged data segment (Part08, Part10, etc) in a different color. The second panel has detrended temperature (cyan) and a running (31-point) standard deviation in the lower panel. Of note: (1) stable temperatures at the beginning, when we conducted Test2 (which did not show interference) (2) sudden onset of noisy temperatures about decimal day 73.8 (unknown cause) (3) briefly stable temperatures during Test3(C), "part 11".