more likely the fuel level sender seal is leaking.....

 

From what Ive read so far if cars have an intank pump and owners are concerned wrt the condition isn't the solution to remove the pump, replace the small section of hose and re-install?

 

That's what I did with mine (while I was replacing the main pump & filter as part of PM). The hose on mine was split but no bits had fallen off into the flow.
1990 S4 with 130,000 km.

Myles

 

I'll have to go back and read the details of what happened to Jim. But, I will throw-in my experience.

My '91 has stranded me on the road exactly twice in the 18-ish years I've owned it. Both times the fuel pump seized due to ingestion of in-tank pump bits. Fool me once. Shame on you. Fool me twice? Shame on [the in-tank pump.] I have a new in-tank pump on the shelf, but installed a new main pump with the screen from the pre-'89s when the second in-tank pump self-destructed last year.

The second in-tank pump appears to have lasted exactly - almost to the day of install - 9 years.

I installed Doc Brown's hose on RedFlash's 90GT this winter. The in-tank pump had already had it's hose replaced with "some kind" of (unidentifiable to me) hose that had lasted about 10 years, but was starting to go.

Maybe if I plan to drive through Death Valley I'll re-install a good in-tank pump before I go.

 

So two reasons to look inside.

The fragrance can be any of several things, incldung the sender seal that Jim mentions, or a vent hose that's cracked or not secured, or a filler neck that isn't connected well, or...

You can check the sender seal through the access cover under the rear deck. Filler neck stuff can be done by the feel-and-sniff method with the right rear wheel removed.

Has the tank ever been out before that you are aware of?

 

All good stuff!

But after reading Jim C's adventure story, I'm not sure it takes a near-death valley experience to conclude that the in-tank pump serves a purpose somewhere shy of superheated fuel conditions. His was working and it still vapor-locked at the pumps. Things that make you go hmmmm...


I'll get a couple pieces of Greg's magic hose and the matching clamps added to my next Roger parts order. Then pump the tank over to the Honda and go into it all.

 

Well, I'll be the guinea pig. I spent last Saturday (at Tass' gathering) doing my best to get the twin-screwed '91 heat soaked. Lots of starting and stopping and hard-driving in between and in full sun-shine with temps in the mid-to-upper 90s. New main pump though.

I know fuel-pump state is - like a rodent - usually binary: they're either alive and healthy or dead. However, is it possible that a main pump could have a failure mode or degrade to an in between state? A seized pump will sink enough current to blow the fuse. Might a very-tired pump sink not-just-quite-enough current? Everything becomes heat. See where I'm going?

EDIT: I forgot. The trip back was 3 hours of stop and go traffic due to Cape Cod-destined cars being used to perform stupid tricks. That's got to be really good for heat soak.

And no A/C in the '91. So, no fuel cooling. (A/C will likely get fixed this weekend.)

 

Already checked the sender seal and hose. Tank has never been out.

 

So rereading this whole thread start to finish brings to mind the idea that we need a better way to mechanically support the in-tank pump, or perhaps just lengthen the hose a little. The little stub of hose is managing a relatively long moment arm in a short distance. The mechanical "reinforcing" in the hose is cross-wrapped fabric something in the rubber sheathing, sometimes with SS braid added for abrasion protection. Still, nothing to keep the hose from trying to make a too-small radius bend, which cracks the rubber as the rubber ages/deteriorates in the fuel, ultimately breaking the plastic liner inside the hose.


So questions to those who have gone before me:

-- Is there any reason why we couldn't lengthen the piece of hose between the pump and the bulkhead fitting? Space limitations, like interference with the well that the level sender sits in?

-- This one primarily to Greg B but others chime in please-- Is there a good way to add a stiffening sleeve of some sort around the hose to keep it from flexing so much? With the clamps as they are, there's not enough room to put something around the short hose and clamps. Maybe something around the outsde of the hose, then clamped with the hose inside. I was initially thinking about nylon tubing around the hose, with the ends feathered where the clamps sit to transfer clamping force to the hose.


I haven't had mine (or any others) apart to look inside, so I'm relying on pictures and the expertise of others for suggestions and comments.

 

All this clamoring about pumps and hoses. Not a word mentioned about the fluid the pump is meant to use.

Bad pump! More pumps! Different pump! New pump! Bad hose! New hose! Broken hose!

I guess, when all an engineer is given is one hammer, everything he see's is a nail.

 

Simply replaced my in tank pump with the strainer from Roger and installed a new external pump about 12 months ago. Hasn't missed a beat and have felt (seat of the pants - SOTT) no perceptable differences i performance, fuel usage, etc. Seems to work just fine with out it. As previously stated incorporated into the 928 design to mitigate vapour lock issues in high temp conditions/countries.

Cheers

Tony
928 S4 MY89

 

when greg built my engine he put the o44 pump with intank strainer and no intank pump.....he said I will have no issues,,,,didnt elaborate..though

 

As usual, there is actually more than one way to solve this kind of problem. Reflective top on the finned roof-mounted tank should do the trick, once we get the aero implications worked out. Gets it out of the way of the Audi 5000 behind you at the toll both. Should free up some room for bigger stuff in the boot too, although still not high enough to carry clubs standing up. Lots of benefits to be realized!

Or the primer pump in the existing tank bottom.



It's tough being an engineer with just a hammer.

 

Customer complained of his car stalling out after driving for 45 minutes or so, said it was going in to "limp home mode", it's a '91. After doing a harness and a few other things I went out for a 30 minute drive, ran fine but had a slight off feel to it. Drove it for 45 minutes this morning and it pooped out on me. Get it home as I knew exactly what I'd find.








Funny thing, as I was talking to the customer while I was on the side of the road he stated that his normal mechanic had replaced the pump twice in 4 years so I asked him if his mechanic replaced the internal pump.

Nope. The pump part was floating around inside the tank and had to be fished out.

 

So what are you going to do? Fix it, or replace it with a strainer?