Around dawn, Steve reassembles the motor. But it still doesn't roll smoothly: the other bearing is stuffed, too. The first replacement bearing came out of our spare motor (the motor from the original Sungroper), but the spare has only one bearing this size, so we order new bearings from Coober Pedy, 150km south. The parents of one of the team members are in Coober Pedy, and they run the bearings up to us. An hour or two later, the parts arrive, and Steve fits them; but the motor is still not smooth: the drive shaft is bent.
How long has it been like this? We don't know. Possibly when the chain came off and locked the back wheel, the motor was put under undue strain, and the bearings toasted and the shaft bent then. Or possibly the motor was like this when it was delivered.
We push the car from where we're working on it back to the control stop, with only half the rear suspension connected because of the work we're doing. We serve our half hour, and push back. When we get there, and pop the array off, we discover that the half of the rear suspension we were using has not coped with the extra load, and is damaged.
In an attempt to get a working motor, we put the bearings back in the original Sungroper motor, solder on an additional plug so that we can connect it to the Tritium Gold Controller of Extreme Shinyness, fit a temperature sensor, and bolt it into the car.
When we test-run it, a nasty mechanical noise comes out of it, and we don't know why.
We pack, and trailer to Coober Pedy.
We roll out in the carpark of an underground backpacker's. Literally underground, of course, as this is the Coober Pedy style: a lot of housing and accommodation is simply tunnelled into the ground. This gives a nearly constant 27 degrees C all year round.
But not so in the carpark, blasted by wind which deposits a patina of red dust on everything, including the array. We clean the array with the last of our rain water. After we finish, it's still dirty, but it doesn't matter: if I push the whole power of the array into the batteries, two thirds of the battery pack heats up, one third dangerously so. So I turn on just one of the four array strings, to feed the batteries quarter power.
As we trailered to Coober Pedy, we had the back wheel hooked up to the remaining half of its suspension. When we roll out, we discover that this too has failed under the doubled load.
We acquire a crowd, several of whom are quite helpful.
John notices that we don't have any dash displays, despite 12V being turned on. We've not getting any telemetry either. I poke around with the multimeter, and discover that the 5V regulator, which is supposed to take 12V and cut it down to size for our 5V electronics, is not working. I replace it with one salvaged from another piece of gear. Now 5V works, but still almost none of the electronics that depends upon it is working. Andrew's +4 Voltage Sense Board of Rapid Prototyping is working, but nothing else. I swap in the master board chip from the original Sungroper, and now I get telemetry, but all the numbers are zero: every microcontroller chip in the car that takes that 5V and feeds data back to the master board has been zapped, presumably by the same thing that destroyed the 5V regulator.
(Note for geeks: I'm using a 7805 for 5V, instead of the TEM-1211 DC-to-DC converter used in Sungroper 1. This is because the TEM-1211 is now very rare and hard to buy, and no pin-for-pin replacement is available. A search on Google within Australia for TEM-1211 turns up exactly one hit: the Sungroper website. This means we do not have the isolation that the TEM-1211 would have provided. But I still have no idea how the boards got zapped.)
Steve fixes the rear suspension by removing it, and bolting the arms that used to go to it directly to the frame of the car. To provide some smoothness of ride, he decreases the pressure in the rear tyre.
John, Steve, and the audience look at the problem with the original Sungroper motor, and decide that it is a problem with the keyway. This is a slot in the shaft: when the sprocket slides on, it is this keyway that forces the sprocket to turn in time with the shaft. But it is loose, which will cause rapidly destructive wear. Options are discussed, and a solution involving glue is decided upon.
John takes the car for a short drive in the carpark to test the new non-suspension. Three metres in, the chain tensioner breaks again, and the chain falls off.
So one motor has a bent shaft, the other has a dodgy keyway, the tensioner is broken, the rear suspension is non-existent, the batteries won't charge, the array is red with dust, the electronics are all toast.
But the Tritium controller still works.
We go for pizza; John lays down the conservative strategy he intends to use: trailer to a few k short of the finish, set down, and solar across the line.
-- Doug Burbidge
