It's back in my tank, i.e. the plastic box full of water :) and it's now ready to go. I leave it in there normally for a week or so, loading up the prop every now and then just to make sure everything's working and checking for water coming out of the tube that the wires go down into. Up until now this has happened quite regularly, fingers crossed it won't be doing from now on.
Here are a few pictures of the thruster as it stands at the moment.
This one is fully assembled, with the two bars down the side fully tightened up, to compress the o-rings at either end.
This next one shows the brushless DC motor controller PCB (designed to be pressure tolerant, more on that later) fitted into the end of the thruster at the rear end of the motor and the prop removed. The control to the thruster is a speed signal (analog 0V to 5V) and a direction signal, either 0V or 5V (digital). The only other wires going down the tube to the PCB are the power supply (+24V and 0V (return)) and the returns for the two control signals. I have put a ramp in the software so that the minimum time it takes to go from full thrust in one direction to full thrust in the other direction is about a second. This puts less stress on the mechanical components of the thruster design, and eventually where the thuster mounts to the ROV.
The one below shows the plastic tubing and barb/fitting where the power and control wires enter the rear end cap of the thruster.
The whole thing is sealed with 5 o-rings in total and one shaft seal! Lots of seals in one place and I'm sure the number could be minimised, but I think it would make it harder to machine the components then. BTW the tube is from the plumbers merchants, the rest of the items are machined on my small CNC machine.
I'm intending doing some thrust tests next, to start with without a shroud but eventually with a shroud to see if it makes much of a difference, which I'm hoping/suspect it will.