Hot start problem resolved due to failed check valve in the DMCH fuel pump

[David T]

It only takes about a one or two hour soak in gasohol if you have some calipers to do the test I did.

Measure each before you soak, Tag which is which, soak and remeasure. I would like another verify of what you find.

The problem is we really do not know exactly what the originals were made of. What they sell today and call Buna N is probably not what they used to make. Whatever the originals were made of, it seemed to be OK. The problems seem to start once you change them. There are plenty of mixture units out there with the original "O" rings in them and they function just fine. If you should replace them I would suggest you try to use Viton of the correct size since what they pass off as Buna N is not going to work. I believe one of the problems is that the Buna N swells and since this is a dynamic seal (the "O" rings are on parts that move) when it swells it gets chopped as it moves over the slots. If it was just used to seal a part that doesn't move (like a drain plug) where it is captured and cannot swell, Buna N seems to be fine.

[Ron]

There are plenty of reasons to go with viton, but I don't think I know of any where it comes to dealing with cars.

...Automatic transmission o-rings and seals.

...I used a Viton o-ring from rebuilt kit leftovers and have had no problems for years.
Back when Dave first posted measurement/results, I had very similar results.

[jawn101]

1) PLEASE - install an external inline check valve. Trust me, you'll be glad that you did. It's $18 and all of 20 minutes of time. It's definitely worth it IMHO!

2) When you installed the pump, are you sure that you had the bottom of the pump as close to touching the bottom of the tank as possible? The unit, as you probably know, is spring loaded and isn't very smooth in it's action when decompressing after being compressed (it's easy to get it stuck in a compressed state). It also doesn't fit the tank when fully decompressed, thus some level of compression is necessary. I test fit mine about 4 times before I was able to get it where the bottom of the pump was just barely off the bottom of the tank. I've only been running mine for the last 3 months and about 4,000 miles, but have always run it down low enough to trip the low fuel light. No issues such as those you've described thus far.

That's about the only thing that I can think of.

You may also want to check the RPM relay - if that's opening/closing while you're driving, it could potentially cause what you're describing as it's stopping and re-engaging the pump.

Best of luck!

Hey - this is a little late for a follow up, but that's how long it took me to run the fuel tank down low enough to inspect. I just pulled the blanking plug off the level sender hole and inspected the pump. It is fully extended and resting on the bottom of the tank. There is about 1/4" of fuel surrounding the baffle area (that is to say that the baffle is about 1/4" submerged) and the gauge is reading a tick below 1/4 full. But I still have the "collapsed pickup hose" symptoms under hard acceleration when the tank is this low.

Anyone have any other ideas what could be causing this? Dave S, are you still out there reading this thread?

[NightFlyer]

Hey - this is a little late for a follow up, but that's how long it took me to run the fuel tank down low enough to inspect. I just pulled the blanking plug off the level sender hole and inspected the pump. It is fully extended and resting on the bottom of the tank. There is about 1/4" of fuel surrounding the baffle area (that is to say that the baffle is about 1/4" submerged) and the gauge is reading a tick below 1/4 full. But I still have the "collapsed pickup hose" symptoms under hard acceleration when the tank is this low.

Anyone have any other ideas what could be causing this? Dave S, are you still out there reading this thread?

Sounds like a crack somewhere within the assembly is introducing air into the system when the fuel level gets low enough. Is the pump still under warranty?

[jawn101]

Sounds like a crack somewhere within the assembly is introducing air into the system when the fuel level gets low enough. Is the pump still under warranty?

Probably not. I'd have to pull the whole assembly to check for one.

FWIW, I did measure the fuel levels in the tank and it seems like there is more around 1" of fuel down there. Perspective made it tough to estimate properly, and I had a hard time getting a real measurement as the fuel mostly evaporated off my rulers before I could get them into the light to read.

[NightFlyer]

Probably not. I'd have to pull the whole assembly to check for one.

FWIW, I did measure the fuel levels in the tank and it seems like there is more around 1" of fuel down there. Perspective made it tough to estimate properly, and I had a hard time getting a real measurement as the fuel mostly evaporated off my rulers before I could get them into the light to read.

I would think that an inch should be enough fuel to still operate the pump normally, but then again, I really don't have any basis of comparison / experience. It's possible that the sender float is incorrectly adjusted and giving an incorrect reading on the dash gauge.

One way to check would be to siphon the remaining fuel in the tank into something that you can use to measure the amount of fuel actually present and compare it to what the gauge is reporting. You can siphon via the sender hole. Inclining the car (by parking on a steep incline or jacking it up) will concentrate the remaining fuel into either the front or back of the tank where you can effectively siphon a vast majority of the remaining fuel out (only a nominal/insignificant amount will remain). Remember that the tank's capacity is 13.2 gallons, thus if your gauge is reading 1/4 tank, you should have roughly 3.3 gallons of fuel remaining within the tank. If the gauge is close to what your measured results yield, then you'll know for certain that the problem lies in the pump housing, unfortunately...

Again, best of luck!

[David T]

I would think that an inch should be enough fuel to still operate the pump normally, but then again, I really don't have any basis of comparison / experience. It's possible that the sender float is incorrectly adjusted and giving an incorrect reading on the dash gauge.

One way to check would be to siphon the remaining fuel in the tank into something that you can use to measure the amount of fuel actually present and compare it to what the gauge is reporting. You can siphon via the sender hole. Inclining the car (by parking on a steep incline or jacking it up) will concentrate the remaining fuel into either the front or back of the tank where you can effectively siphon a vast majority of the remaining fuel out (only a nominal/insignificant amount will remain). Remember that the tank's capacity is 13.2 gallons, thus if your gauge is reading 1/4 tank, you should have roughly 3.3 gallons of fuel remaining within the tank. If the gauge is close to what your measured results yield, then you'll know for certain that the problem lies in the pump housing, unfortunately...

Again, best of luck!

You really cannot properly inspect the baffle assembly from the sender unit hole. Pull the pump out and inspect the pick-up hose. If is is swollen or soft either put a spring in it or replace it.

[DMC5180]

You really cannot properly inspect the baffle assembly from the sender unit hole. Pull the pump out and inspect the pick-up hose. If is is swollen or soft either put a spring in it or replace it.

David,

Jon has the NEW DMCH Pump combo unit. No Pickup hose on this one.

[jawn101]

OK, after trading some emails, photos, videos and some call time with Dave S (have I mentioned lately how great he is?) I've got a plan of action.

First step was to test the level sender. I held the float down gently with this threaded rod:



Then verified that the sender and low fuel light were working properly



OK, that's all fine. On to the next step - running the engine with the sender plug removed to observe the pump operation.



Um. Well, I'm pretty sure that isn't supposed to happen. Next step was to remove and inspect the pump assembly for leaks/cracks:









Clean as a whistle and no obvious cracks/splits or anything concerning. So, I tried reinstalling the assembly. It's a DeLorean after all, sometimes a simple R&R is all it takes to fix a problem, right?

Wrong. Same symptoms after reinstalling. So I pulled the pump up out of the tank just enough to observe it in operation from directly above instead of from the side. Here's what I saw:

https://www.youtube.com/watch?v=3bKBobEnZos (sorry, can only embed 1 video per post for some reason?)

Uh oh. The leak is right at the top of the pump itself.

So right now the working theory is that the pump was still providing adequate pressure to the engine (no issues at all with running or driveability unless the gas is very low) but due to this leak, the fuel in the baffle was being depleted very quickly - causing the poor running conditions with low fuel in the tank.

Long and short of it is that 2100 is off the road while the pump assembly heads back to DMCMW for a thorough check and repair by Dave. A bummer but at least it's a resolution!

Will update again when I have more info on the situation.

BTW: Dave said he was going to return the unit with an inline check valve supplied.

[DavidProehl]

Great job documenting this. That is surprising to see gas leaking out the top of the unit! No wonder it was so hard to track down.