As many of you already know, OpenROV is going to be heading to Lake Tahoe in early June for an exploration of the wreck of the steamer SS Tahoe, which lies in 120-150m of water:
The principal vehicles we’ll be using for the expedition will be two heavily-modified OpenROV 2.8 units. They’ll be ruggedized to handle depths down to 150m, will have the high-definition 1080p camera from the upcoming Trident, new more-powerful batteries, neutral buoyancy tether, two pairs of external light cubes for illuminating the wreck, and a TriTech MicronNav USBL transponder to allow us to track the exact position of the ROV.
We’ll have further details on the ROVs once we have all the details perfected and ready-to-go, but for now I thought I’d post details of the modifications we had to do to the controller board to support two pairs of external lights and mounting of the USBL transponder.
A stock OpenROV can easily drive a pair of external light cubes, and revision 2.8 is prewired and pre-programmed to do so through the PWM3 channel on the controller board and external wires TP20 and TP21. It is easy to wire up the PWM4 channel to drive another pair of light cubes. But if you do so, and run both sets of light cubes simultaneously, then sense resistor R12 on the controller board ( see sheet 3 of the schematic here) starts to overheat. The current OROV 2.8 design can’t really handle more than about 2A of current being drawn on external loads without something overheating. Given that people are now starting to strap on more and more external loads, we’ll fix this issue in OROV 2.9, but for now you need to modify the controller board a bit to safely do this.
OpenROV 2.8 comes pre-wired to put (+) battery power (through R32 and R73) onto external wire TP20, while connecting wire TP21 to the low-side switch PWM3 (Q6). To attach two pairs of light cubes to the vehicle, we’re going to bypass the existing wiring to TP20, and hook TP20 directly to switched battery power, through a 4 amp polyfuse to provide some protection from external short-circuits. (+) battery power for both sets of light cubes will come from the TP20 wire. One pair of light cubes will use the existing wiring on TP21 to attach the (-) lead, and the second pair will have its (-) lead connected to the TP22 wire.
The first step to the mod is to desolder R73, which would be the normal way that battery power would get fed to TP20. We’re going to replace that with a higher-current wire.
We pick up battery power for the light cubes at the drain of MOSFET Q1, the main power switch. Here’s the polyfuse attached to Q1:
The red wire from the polyfuse goes down through one of the holes for the ESC2 leads, down to the bottom of the board:
On the bottom it connects to the pin of the DB-25 connector used for TP20:
For the (-) leads, the 2.8 controller board already has an attachment between TP21. We need to add another wire between TP22 and PWM4- to control the second pair of light cubes. Here’s the connection at PWM4-:
And here’s the other side of that wire at TP22:
The transponder for the TriTech MicronNav USBL system requires 12V power. We could attempt to run it off of battery voltage, but the battery bus voltage isn’t quite 12V, and dips even lower when the motors are running. A safer bet is to generate a clean 12V power supply from the 5V supply that is already available on the controller board. To get power to the transponder, we will use the power and ground connections that are pre-wired into OROV 2.8 for an external servo (TP24 and TP25).
External wire TP24 normally feeds +5V to an external servo, so we need to change this to +12V. So first we need to cut off the 5V connection by removing R71:
We’ll generate 12V power from the 5V bus by using a Pololu boost regulator board. The input and ground pins of this board will connect easily to the prototyping header pads on the controller board:
The output of the boost converter is wired to TP24. Here’s a shot of the completed wiring:
We’ll have more details on the other ROV modifications once we’re done with testing them.