One of the most critical aspects of a properly functioning ROV is keeping a dry electronics housing. We've been using the double o-ring strategy with laser cut acrylic end caps and an extruded acrylic cylinder. The tolerances of the extruded tubes and any slight variations in a laser cutter can put the electronics housing at a higher risk of failure.
I've added some ways to test and troubleshoot potential problems with the electronics housing on the Wiki. Let me know what you think!
Also, I think it's a good idea to have tests for every section of the build so that potential problems can be caught early, before they're permanent. More on this soon...
looking at the tolerances at mcmaster-carr, the wallthickness is +-15%, have you done any measurements or observations on the tolerances. are the tubes oval, uneven in thickness?
do you know the tolerances from Evonik?(the supplier in the EU BOM)
I really like what David is suggesting about testing before all of the expensive electronics are in the tube. But the only way to test if the end caps are working is to have the pass through holes epoxied, and obviously one of the caps needs to have the wires going through the epoxy. If possible I am going to be just having wires in the tube without all of the electronics and doing some bucket tests and deeper before putting in electronics. And I think I have read somewhere about having a little kit on hand (DI water, alcohol, etc?) in case the tube floods a little... This stuff is worth thinking about - I think quite a few of us are going to need to do the ptfe "teflon" tape method.
By the way I found some 1/4" wide teflon tape. If I double that over it fits inside the o-ring groove. A few wraps around, put the o-ring back on, and the tap doesnt even show. (I worry about if its exposed it might eventually get caught on something during tube cap removal etc).
It must just be wallthickness. We haven't had any problems with oval tubes. The thickness seems to be consistent around the diameter but can differ over the length of the tube. For instance, it's possible that an endcap fits on one end of the tube and not quite on the other.
As Frank says, it's important that we monitor this closely...
I sent an email to evonik and got the answer today, they provide a slightly better tolerance(than mcmaster) on the wallthickness(around 11%) i've attached the specifications from them.
Thanks for that Thomas. We also found that there can be slight variations on the laser cut parts from machine to machine (laser cutters) or over time (like cutting something a week later.
Bottom line: test thoroughly.
Friends, I have been able to do a test on a tube with the end cap o-rings adjusted into proper place using the "teflon tape" fix referred to on that wiki page - It was a test of only a few minutes but the tube was sent down to 40ft of seawater and it held fine!! Additionally the tube was also put in a shallow seawater holding tank for approx 15 hours and it held fine there as well. so thats encouraging then :-) BTW without this fix the tube did have a slow leak so beware...
Why was the double o-ring design abandoned?