So, we got the bubbles problem out of the way, and still have relevel issues.

I was doubting the sealing of the Maxton check valve, and so took that all apart, cleaned it up, and replaced the valve stem seats, but it still releveled. Like always, it would initially level into any floor just fine, cycle the doors and park. Then after a couple of minutes, you would see the LU relay come in, and she'd relevel up. The packing was oil tight, no oil in the pit return sump, no oil going over the top of the jack. Everything was bone dry. Where was the volume of oil going?

I tried a pressure test, chained the push plank down to the pit channels, and put 400 psi on it. Watched the gauge, no bleedoff.

One day I put the car at the top floor, closed the gate valve in the machine room, yanked the power, then disconnected the feedline from the valve. Not a drop of oil came through that Apollo ball valve, all night long. Dems some great sealing ball valves, rite dere. Even with the ball valve closed, the elevator dropped 4" !!!

Where is the damn oil going, that is allowing the car to drift downward? It's not going back into the tank, it's not going back into the pit can, the jack's not leaking! Arrgh!

I consulted an engineer. Yeah, I'll admit this one had me baffled, and no, I didn't solve it without help. Wish I did. But the engineer measured the temperature of oil in the tank, and the temperature of oil at the jack, and we found that it had a difference of over 40 degrees! Sometimes much more, depending if the boiler room was hot where the tank was located, and if the elevator was being used a lot, etc. But generally, this point is that the hot oil would go through a long feedline, dissapating heat like a radiator, and then into a pit that was cold as a witches' tit. Must have been a high water table, bcause that pit was unusually cold.

Hydro oil, aka "32" oil, or ASO 150, has a coefficient of volume expansion of 0.027% per degree of Farenheit, (iirc, was over 10 years ago) and the contraction is the same when cooling. So for the given volume of jack oil, when the elevator is at the top floor, plus the feedline, equals "X", times 0.027%, times the degrees of temperature shift, equals "Y" volume lost, which came out to about 4" of piston length, when cooling over night, in worst case conditions.

It was just the conditions the job was installed under, and nothing at all was wrong with the equipment in any way. Chalk it up to the Fed's divine right to put the tank unit in a boiler room.

So I took the "FLO" fan/light output timer, which changes states after demand ceases according to the set time value, and drove a relay with it, which I used to inhibit the LU signal from getting to terminal 26, or up level input. When the car was running, FLO is dormant, car levels normally. After sitting for one minute, FLO changes state, the LU signal is inhibited, car drifts below floor level. It might drift even a few inches off floor level, but that's ok. There is no way to get into the car without having it relevel itself back flush with the floor, because any form of demand, whether Phase 1, hall call, car call, MLT, etc, would cause FLO to change states again, then it would level up flush. Even if the car had drifted out of the door zone, it wouldnt matter, because it won't open its' doors until it's in the door zone, after leveling in.

I don't reccomend modifying controls without getting the OEM to sign off on the change, because you could incur liability on yourself, even if some accident has nothing at all to do with the change you made. Liars, oops, I mean LAWYERS will distort anything if they think they can win a judgement. Draw it up, then send it to the OEM for an offical blessing.