[urq] Fuel Pump Redux

Ben Swann benswann at verizon.net
Mon Oct 22 18:14:34 PDT 2007


Here is the follow-up on fuel pump replacement outcome.  I went through a lot of trouble
to make sure that I was solving a problem I originally suspected was the fuel pump,  but
wanted to make sure it wasn’t something else, especially since many commented there were
other possibilities.  In the end, I had a bad fuel pump and of that I was sure, but
also, there was validity in comments about possible starvation to the pump.  I made sure
to eliminate those items prior to replacing the pump, rather than just installing a new
pump.  I could have achieved the same end result using a stock pump, but found a Walbro
pump I consider to meet and exceed the need of the aftermarket Megasquirt EFI as well as
the stock CIS system at about half the cost.  The main downside to this pump is fittings
and connections needed to be changed, but partly I had done that already and actually
ended up with fuel line sizing closer to stock and some of those lines needed replacing
anyway..

When I drained the tank, I noticed the flow started out good, but eventually went down
to a trickle until I blew clear with compressed air – then full flow again, for awhile.
I had suspected some flow issues, and some prodded me to check this, so went about
checking for a clogged screen.  After removing the tank and flushing and seeing the
screen was a sizeable screen “sock”,  I concluded there was still some varnish left from
long ago when I got the car – had been sitting for about 6 years and one of the initial
tasks in bringing this car back to life was purging the jellied gasoline from all the
fuel system components.  I did good at that time, but there apparently was still some
residual that was probably straining the flow.  After removing the tank, inspecting and
flushing and blowing the sock clear, I’m confident this won’t be a problem, and I left
the “sock” in place.  I found the tank can be removed for cleaning in about 4 hours out
and back in.   Other things that contribute to slow flow to the pump include the lines
from the tank and they will kink if given the slightest opportunity. If you do this,
make sure to route them so they won’t kink and too much line will kink as will too
little.

I had originally installed a pre-pump filter when I got the car – mine did not have one
when I got it, but with installation of larger NG pump I procured the largest OEM filter
with ½” diameter hose and that matched the converted hose to the pump being used from NG
equipped late CGT and model 80’s.  I found even with new lines and newly cleaned tank,
this pre-filter tended to “cavitate”  a little bit and that was visible, but I don’t
attribute this at all to causing the pump to fail.  In fact, the pump was making noises
from the getgo.  A pump that it replaced did not last 2 sec from the box, despite being
given clean fuel and good flow.  I later learned that there was a bad run on some Bosch
pumps and these were apparently among them.  I believe they were remanufactured in some
unscrupulous plants and sold as new pumps - evident by peeling paint on what should have
been new pumps.  Anyway, I thought I’d see if there were some other Bosch alternative
that would do the job – I no longer have a Bosch CIS system in this car, and experiment
I will, especially if it saves me money.  So I got the Wallbro at around $100.   

Prior to removing the old pump – it still worked, but made the same noises and dropouts
that it always had – I re-did the lines and installed a see-through filter with
replaceable elements.   I noticed that the OEM filter seemed to be cavitating by virtue
of it large diameter, and this was closer to being a tube, so worth a try.  It was
difficult to redo the lines and filter to keep the hoses from kinking, but made sure no
kinks and nice curves prior to the pump inlet.  The old pump ran a little better, but
still exhibited its old tricks, so out it went.

The Wallbro uses ring terminals, so I installed some larger than those supplied since
the wire guage for Bosch pump was larger – 12 guage I think.  I had some fun coming up
with the right combination of fittings mostly 3/8”  but retaining the ½” feed from the
tank.   If you do this, be very careful to install the hose wit nice large radius curves
so it doesn’t kink.  I finally got the hose routing right.  The new pump was a little
more noisy than a typical Bosch, but fairly quiet, and the sound was consistant
indicating no flow interruptions as before.  The engine now runs much smoother and
consistently than it did with the bad pump – Duh!   The new line setup probably helps
too, and I can see clearly there is no cavitation going on before the pump as well as
see if I’m picking up debris.  I know that I could have gotten the same results with
stock pump, so not trying to say otherwise, just the pump was indeed bad.  It had always
caused hiccups, but never nailed it down while sorting all the issues I had with the
original engine and CIS.  It took installation of a new engine and EFI to discover it
was a problem.

I still have a problem with the engine falling flat at 5000RPM, so the pump did not
resolve this.  I felt confident that it is not a fuel delivery issue and so can go on
narrowing down this problem.  Please don’t comment on the cutout problem or speculate as
a reply to this post.  I am going to address it in another post as well as enquire to
the Megasquirt collective.  I think it is ignition, but again, don’t want to cross
threads.

Thanks all for your suggestions, even if I did not necessarily want them at first, in
the end, it made me to work through additional possibilities and eliminate other
potential problems.

Ben
www.gtquattro.com



More information about the urq mailing list