so my external quick disconnect failed after two days on the playa. good thing the honda eu2000i runs for so long, it just meant i had to gas up once a day. still, i'm keen on getting the extended run time working.
i finally spent some time to fuss with this and figure out what went wrong. i was stumped at first, and started to just redesign the whole thing, at some point probably had multiple failure points too.
v2, cap
after paring things down, i realized i had a problem with a part i wasn't paying much attention to at all. the quick disconnect itself has an o-ring on the inside, which allows it to have a nice seal with the post. that o-ring was missing. won't work without that. any set-up can face this problem (no matter how novel my design was!).
along the way, i changed the cap design. twice actually. first just adding more o-rings to create a better seal on the cap. i'm guessing my original design was pretty good. i wasn't sure where the failure was, but i assumed the cap was the culprit.
original, barb-inside cap design
maybe for the better. this prompted me to change my approach and build something more along the lines what other folks were doing (stuffing a hose barb on the end of the gas cap). since i completed a project using a "tap & dye" to re-thread something now, i realized tapping out the cap would be easier than i thought (just hadn't really done that before). this is the optimal solution in my book for the cap design.
v2 finished!
so i just used the threaded end of the brass hose barb and started to jam it in the cap using a wrench for torque, and turning it into place. practically threaded itself with a nice tight fit. good.
but i tested it out and it still wasn't working? this is about the time i discovered my o-ring was missing -but to test it, i bypassed the disconnect all together. *still* didn't work. keeping a vacuum seal can drive you nuts!
so-- while it seemed like the hose barb was on tight, air still got past. on closer examination, i noticed that a *tiny* trench bored in the cap was enough to let air through (part of the honda cap design), not allowing a proper seal. so i put an o-ring at the top, and that did the trick. i believe the vacuum also 'pulls' the o-ring in. creating a better seal there.
it works!
like i said, my *original* design probably worked fine. but i'm pretty committed to the new one. with the o-ring and some sealant for good measure, should put that portion to rest.
for the remainder of the line i made everything else, including the fuel filter, all 3/8". no more fussing with hose-in-hose connections. while that worked brilliantly for the fuel filter (probably due to the hose being placed over it's nozzle for support) --more parts means more points of failure. more second guessing what could possibly *be wrong* when things *go wrong*. and while i liked the 1/4" tubing because it moves the gas faster (easier to see it working) and put let pressure on the fuel pump -everyone else seems just fine using 3/8" tubing and i'll stick with what works. again, the less parts, the better.
odds -n- ends :
threaded vinyl, from heat
curiously enough, with the prior version the vinyl hose actually threaded itself onto the plastic threaded hose barb (see photo). i just thought that was kind cool. not terribly useful in my circumstance now, but could come in handy for another project down the road.
johnson 3-prong tank connector
also, i finally saw where the *air* goes in the three-prong johnson connector. (really doesn't seem to effect my external tank operation at all). it's not *on* the third post as i had assumed (which is more of a stand-off). look for the hole below the pushpin for air, right in between the regular stand-off and the gas line. it's hiding at the base and angled down, the small divet maybe tough to see looking from 'above'.
also note on the three-prong: the pin on the right is definitely for the gas line. the pin on is probably for air. but i'm not sure how it works. since i have a good vacuum and the gas flows i don't think the air line does much of anything in my case. maybe some who knows just how the three-prong connector works could highlight what's going on. for the time being, it doesn't seem to matter. it works. i'm just curious.
"3-prong" assembly
backstory / follow-up :
generator backstory.
generator external fuel tank v2+.