January 30, 2007
Not everything goes according to plan. After unpacking most of the hardware into the field lab the previous day the first order of business was to top off the charge on the twin litihium-ion power stacks used to run DEPTHX, since all subsequent electronics check-outs are based on being on system power.
Unbeknownst to us (possibly from vibration on the trip down) a loose wire on the charging cable crossed circuits and sent full stack voltage (48 volts) down a line normally used for digital communications to control a number of electronics boards on the battery. Before we caught the problem it had fried all three (two primary and one backup) of the E-Stop (emergency stop) power disconnect boards on site here at Zacaton. For human safety reasons DEPTHX has the equivalent of a “big red STOP” button on the top of the vehicle. But because the full power draw from the vehicle can be large (scores of amps) that E-stop switch is not like your normal house light switch -- it’s a digital switch that in turn “talks” to a very high power relay board in each battery which in turn enables high amp flow from the batteries. And it was the communications line (and all the microchips) to that “Disconnect” board that got fried. It took all of today to sort out the source of the problem that caused the loss of the boards and to permanently preclude the incident from re-occurring. But that still left us with a dead bot until the E-stop could be renabled. No solution had been found as of midnight and the team retired with the unpleasant knowledge that we had just 4 days in the jungle to sort this out before the arrival of the main research team.
Bill Stone
Stone Aerospace
Above: source of “El Nacimiento” (the spring) lies just 100 m west of the main field lab at Zacaton. The horizontal black slot at center is the entrance to the Pasaje de la Tortuga Muerta (dead turtle passage) leading upstream and underground to the main Zacaton cenote. It was first explored in 1989 by Jim Bowden and Gary Walton and leads 220 m underground to where it re-emerges on the southwest corner of Zacaton.
January 31, 2007
We awoke early to the knowledge that the bot was dead until we resolved the disconnect board issue. John Kerr began a continuous stream of emails with Nigel Jones back in the states while Vickie Siegel, myself, and two visitors coming north from the 2007 Tabasco caving expedition made our way into Aldama (the nearest town) in search of a Mexican cell phone to see if we could get voice communications out from the ranch. We considered Skype but it doesn’t work well when run off satellite relay (we were using a customized version of a HughesNet link).
Above: Nacho, the main ranch hand at Rancho la Azufrosa, brings over one of many loads of sand used to level out the “bot garage” so that sheets of plywood could be laid down to serve as a surface for maneuvering the bot (it’s lower orange frame rides on four heavy shop casters, but these need a flat surface to work on.
So off we went while John continued his asynchronous stream of emails with Nigel (our lead embedded systems electrical engineer). A plan was developed to produce more boards. Component parts were FedExed to Jones and he spent most of the day assembling the prototype boards - a tedious process that involved placing 2x3 mm surface mount chips on the board and using a large magnifying lens to help with guiding a fine point soldering iron to get all the chip pins properly soldered to their respective pads on the board. Nigel also programmed the onboard logic chips and then re-FedExed them to Harlingen, Texas where, if all went according to plan, they would be picked up by Marcus Gary who was due to begin driving back on February 2.
Bill Stone
Stone Aerospace
Above: John Kerr debugs the disconnect board following email discussions with Nigel Jones.
February 1, 2007
A good day today. We received word early via email from Marcus Gary’s Blackberry wireless that he had made it to Harlingen and that the replacement disconnect boards had arrived intact at FedEx. It would still be another six hours before he would arrive at Rancho la Azufrosa. Meanwhile it was a full day for those of us at the ranch. Vickie Siegel and I began an assembly line to clean all of the electrical connectors for DEPTHX and replace all of the oring seals associated with them. Given that this would be the final system assembly before DEPTHX could see service at up to 1,000 m water depth this was a necessary preventive maintenance task.
 |
|
 |
| Above: Vickie Siegel takes on the tedious task of replacing all the oring seals on the electrical cables for DEPTHX. Some of the eleven electronics housings have more than a dozen such connections each, so this was an all day task. |
Once that was done each cable end was individually sealed to prevent ingress of dust... of which there was a lot flying around outside as stormy weather blew in from the north.
 While waiting for the disconnect boards to arrive John Kerr made good use of time by integrating the main cPCI computer housing, the science payload computer, and two of the sonar array digital signal processors into the vehicle.
Marcus Gary finally arrived at 7pm with the replacement boards. When installed, however, they did not deliver the necessary output voltages. Further emails ensued but by 11 pm the boards were still non-functional and the team went to bed. Nonetheless a lot of progress had been made today.
Bill Stone
Stone Aerospace
|
