To print this text file, a 10-point/12-pitch fixed font (e.g. Courier) is best ============================================================================== BULLETIN FSB-102 - WORKHORSE AND BROADBAND SPEED OF SOUND ERROR DATE/REV: 15NOV96 / 0 ============================================================================== ABSTRACT: Under certain conditions, there is an error in the speed of sound calculation. INSTRUMENT/PROGRAM AFFECTED: BBADCP CPU firmware v5.50; Workhorse Monitor or Sentinel CPU firmware v8.12 and v8.13; Navigator CPU firmware v9.04, v9.05, and v9.06; NOTE: These units must also meet the conditions described below. DESCRIPTION: In June of 1996, a customer pointed out that the model we normally use for speed of sound calculation is only good to about 1000 meters. We therefore converted to a different formula for speed of sound that is good to 8000 meters depth. Unfortunately, we introduced an error into the coding of this new equation. The following indicates the conditions under which the new equation is used. (Salinity >= 30 ppt) AND (Temperature <= 30 degrees C) In other words, if your deployed system meets ALL the following conditions, the calculated speed of sound will be incorrect: 1. You have one of the firmware versions listed under INSTRUMENT/PROGRAM AFFECTED. 2. The ED-command is set to a value greater than 0. 3. Salinity is equal to or greater than 30 (but not 35). 4. Water temperature is equal to or less than 30 degrees C. For the conditions under which this new equation produces an error, the error grows linearly with depth for a given salinity value. A value of 35 ppt for salinity produces no error, since the affected term is zero in this case. However, the error grows almost linearly as the salinity moves away from 35 ppt in either direction. The following table gives you an idea of how large the speed of sound error is (expressed as a percentage). The table shows the speed of sound error as a function of salinity and depth. For positive error, the calculated speed of sound is too high; for negative error, the calculated speed of sound is too low. Speed Of Sound Error (in %): -------------------------------------------------------- SALINITY | DEPTH (ppt) | 20m 50m 100m 200m 500m 1000m 6000m -------- | ---- ---- ---- ----- ----- ----- ------ 30 | 1.7 4.2 8.4 16.7 41.7 82.9 472.0 31 | 1.3 3.3 6.7 13.4 33.3 66.3 377.3 32 | 1.0 2.5 5.0 10.0 24.9 50.0 283.0 33 | 0.7 1.7 3.3 6.7 16.6 33.0 187.9 34 | 0.3 0.8 1.6 3.3 8.3 16.5 94.0 35 | 0.0 0.0 0.0 0.0 0.0 0.0 0.0 36 | -0.3 -0.8 -1.7 -3.3 -8.3 -16.5 -94.0 37 | -0.7 -1.7 -3.3 -6.7 -16.6 -33.0 -187.9 38 | -1.0 -2.5 -5.0 -10.0 -24.9 -49.5 -281.7 39 | -1.3 -3.3 -6.7 -13.3 -33.1 -65.9 -375.3 40 | -1.7 -4.2 -8.3 -16.6 -41.4 -82.3 -468.8 -------------------------------------------------------- SOLUTION/ACTION: Short Term Work-arounds - Doing any of the following steps will prevent the instrument from computing an invalid speed of sound. However, in each of these work-arounds, you will have to scale the data appropriately during post- processing. 1. Enter a fixed value for speed of sound through the EZ and EC-commands. No error occurs when a fixed value for speed of sound is used. --OR-- 2. Enter a fixed value of 35 ppt for salinity (ES=35). No error is produced if the salinity is 35 ppt. --OR-- 3. Enter a fixed value of 0 decimeters for the depth (ED=000). No error is produced if the depth is set to 0 decimeters. Long Term Solution - New versions of the affected CPU firmware code are being developed. However, we do not plan to release these new versions until early 1997. When the releases are ready, we will announce them via the RDI Electronic Newsletter. To automatically receive this newsletter, send an e-mail message to "rdifs@rdinstruments.com" with the word SUBSCRIBE as the Subject. In the body of the message, include your name, organization, address, telephone, and system type. ==============================================================================