Fuel Pump Return line

[john carroll]

Dumping the bypassed fuel back into the inlet side of the pump is a bad
deal.  The sudden depresurization and turbulance encourages the lighter
fractions to bubble out of solution.  After a few passes through the
regulator and engine compartment the fuel warms up, compounding the problem.
A friend did this despite being warnined.  He is alive, his airplane is
bent.  Put it back into the tank.

john carroll
jac at wave.sheridan.wy.us




At 03:13 PM 8/26/96 -0500, you wrote:
>>Matthew Lamari <mlamari at origin.ea.com> wrote:
>>
>>I was just thinking of your idea (quite good) for the return line to a pump
>>going to the back of the pump (thus saving further butchery of the fuel tank
>>area.)
>>
>>I was wondering, though, that, in the flow circuit, it is possible for the
>>pump to end up pulling fuel through the regulator.  I guess this would
>>depend on the flow resistance between the fork and the tank, and the pump
>>and the return fork. 
>>
>>Since it was your idea (don't want to take credit) I was wondering if you
>>could mention it to the group and see if anyone has tried it or shoots it
>>down for some reason.
>>
>>It sounds like a cool way to add such a thing; but I was wondering if it had
>>any hazards I should know about.
>>
>>thanks,
>>Matthew.
>>
>
>Okay, here goes:
>
>On the Holley EFI system (Matt, you could call Holley tech. support
>at 502-781-9741 about this), a return petrol line is required from
>the pressure reg which is located at the TBI.  The fuel pump runs
>continuously with the excess fed back to the tank.  I understand
>some other systems do this as well.
>
>Matt wanted to avoid installing a return line (I put mine into the
>filler neck).  I suggested perhaps putting a tee into the pump intake
>and feeding back to there.  Then, if no fuel being used by car,
>the fuel would circulate in loop from pump to TBI back to pump.  Only
>fuel drawn from tank would be actual amount used.  The way the return line
>system works, the fuel is constantly circulated through the tank,
>its filter sock, and whatever other filters are in line.
>
>Any comments?  Any reason why this wouldn't work?
>
>tom cloud
>
>
john carroll
jac at wave.sheridan.wy.us