Out Banjo persistant leak

[sgnimj96]

       I may have to open the top of the tank and look inside.      When I took the outlet line off the tank fuel didn't run out very fast-  poked a stiff zip tip end in it,  made no difference.        That poor electric fuel pump had a lot of resistance to work against, pushing and pulling.  Its own filter seemed clogged when I checked on the road but you can't bypass it without a threaded adaptor.      It was the green Mr. Gasket Diesel pump.  Still comes on.    At least now I can drive my other car in the meantime.     I will be running biodiesel in it when it gets straightened out (they sell B99 here).       I got a lot to do before I can get back to the banjo leak.     Thanks for all the posts.   It was amazing to get back and see the good discussion that had started about the possible causes of return banjo leaks.   

[TylerDurden]

Return line pressure readings:

1985 Jetta, 1991 ECO engine w/ 1989 TD IP.

That's an air/gas gauge.

[libbydiesel]

Thanks for the test.  Your gauge is weird, tho.  It looks like 10 marks between 10 and 15 psi but only 8 marks between 0 and 10.  It looks like you're counting the marks as 1 psi each doesn't make sense in either case.  Regardless, apparently there is slightly more flow from the OUT bolt than will bleed back to the tank by gravity alone, unless you have a bit of restriction in your return line somewhere.  

[745 turbogreasel]

OP, you might also want to consider running a biocide fuel additive.

My boost gauge is like that too, but has honest numbers that start at 2.
might be to stop chattering on the peg at near 0, or to mask  poor calibration at 0.

[TylerDurden]

Not too surprising that 500ml/sec flow from the out bolt to the tank would generate some pressure, albeit quite small compared to typical fuel line pressures.  I would have expected slightly higher pressures at 3000rpm.


Had I known the pressures were in that small of a range, I'd have found a more appropriate gauge.

[sgnimj96]

       ...sounds like all new fuel lines is in order for this.      I do have some Killem biocide I can put in the fuel after.
         Nice pics with the gauge on to see actual normal pressures!    

[sgnimj96]

Off topic somewhat, but where does one find replacement copper washers, my import supplier could not find the ones I need for my oil return thanks.

I got em from Autohauz.    You might want to get some other stuff while your at it.

[TylerDurden]

DIY manometer, ATF for the fluid.




Tee into the return line, mostly level with the raintray, since about 4 feet of fuel will push into the tube.





Engine off: Negative pressure from the weight of the fuel trying to flow back to the tank.





1000 RPM:





2000 RPM:





3000 RPM:





Keep in mind that the reading is including the negative pressure the flow the line to the tank is creating; but that resulting sum (the reading) is the pressure the OUT banjo encounters.

So it would appear the Watts gauge is a bit inaccurate at low pressures. lol

The u-tube manometer is quite accurate if there are no leaks. It is a common "primary standard".

For those who like math, the specific gravity of ATF is ~0.87.

[92EcoDiesel Jetta]

Nice science project. The pressure you measured is after the banjo bolt orifice. Now measure it before the orifice from the outside of the copper washer with the banjo bolt loosened.

[TylerDurden]

Nice science project. The pressure you measured is after the banjo bolt orifice. Now measure it before the orifice from the outside of the copper washer with the banjo bolt loosened.

Lol... I'll let you handle that one.

[8v-of-fury]

Pressure .

From 12 with the engine off to over 40 at 3000rpm. It rose the liquid in the manometer by 28 what I am guessing are inches?

Care to go a bit further in to detail TD? Interesting.

[TylerDurden]

Ja, them is inches... inches of ATF.

40 inches of ATF should be about 46" of water ("WC). (40/.87)

46 / 27.7 = 1.66psi.   So, the banjo outlet is handling ~1.7psi.  


12/.87 = 13.8" of vacuum due to the tank means the IP out bolt orifice is actually outputting 59.8" WC of pressure, or 2.2psi... but the tank vacuum lowers the pressure in the fitting.

[8v-of-fury]

Right on, I learnt about it in the times between our reply but I came to some different pressures. Being that it was not my experiment I don't have accurate readings of what you had there.. so it probably threw it off.

Is 2.2 psi enough to cause a leak between those copper washers and the aluminium if not crushed? I think so.

[TylerDurden]

Is 2.2 psi enough to cause a leak between those copper washers and the aluminium if not crushed? I think so.

I rekon so too... Every time I leave that fitting loose, it leaks just fine.

[libbydiesel]

Now measure it before the orifice from the outside of the copper washer with the banjo bolt loosened.

...while doing a handstand, riding backwards, blindfolded, on a horse...