Electric fuel pump to prime carbs

[Dick Steinkamp]

It seems that most of us with 38/38 or 32/36 Webers experience dry carb float bowls if our car sits for a relatively short period of time (>4 days?). I've put up with this for a year or two, but it's annoying and tough on the starter and battery and time to address the problem.

I've owned many collector cars with carburetors. Some had the same problem, some didn't. I could never figure out a common thread for those that did. New, rebuilt, old...it didn't seem to matter. Some did, some didn't. I could never pin down where the fuel was going. It couldn't be backing out of the carb up through the needle and seat. Liquids don't flow uphill. It couldn't be leaking externally out of the float bowl or there would be a gas smell and stains on the carb and manifold. It couldn't be leaking internally or the engine would be flooded on the first start attempt and the oil would be getting thinner and higher on the dipstick. The only thing that makes any sense is evaporation, but that sure seems like a stretch. There is only a pinhole size vent in the float bowl. It doesn't seem it could let that much vapor out. There is no gas smell in the garage. If anyone has a proven explanation why our Weber float bowls empty in 4 days I'd sure like to hear it.

I put an anti backflow valve in the fuel hose at the outlet of the tank. That didn't make any difference. I didn't expect it to. If the fuel in the line was draining back to the tank the float bowls should (theoretically) still be full and provide plenty of fuel to start and run the car until the pump filled the line again.

So now I'm ready to install an electric pump for the sole purpose of filling the float bowls prior to turning the key to start the car. I want to keep the mechanical pump as the primary. I think I've read all the threads here about this process. I want to put the electric pump next to the gas tank in the trunk. It might perform better if it was lower then the tank but I don't want to cut the hard line under the car and have the pump in a location that is difficult to get to.

The Pierburg EF1 7.21440.51.2 seems to be the pump of choice here for my purpose. I found this diagram on the Pierburg site for what they call a "pre feeder pump".

03 is the pump, 02 is a coarse filter just to keep big junk out of the pump. They advise against a fine filter here because it could restrict the flow to the pump which would overheat the pump and burn it up. 05 is an anti backflow valve so that the output of the pump doesn't just go in a loop. 06 is an existing pump. In my case, the stock mechanical pump. The Pierburg is shown mounted above the fuel level which tells me it can "suck" fuel and would work in the application I'm considering.

How does this look to those that have installed an electric pump just to fill the float bowls prior to starting?

Does the fuel flow through the stock mechanical pump pre start OK?

What are some alternatives for mounting the switch/button to activate the pump?

 

[JFENG]

(1) go read BavBob’s recent experience adding this pump to his CS.
(2) the diagram looks right to me. You need the parallel path because the main pump can’t pull fuel thru the pierburg (vane type pump). The backflow preventer is also needed so the output of the pierburg goes toward the carbs and not just back to the tank.
(3) rather than Tee fittings I suggest you get u-bend type Y fittings to keep the hoses tidy.
Here's a link to a nice one: u-bend fitting

 

[Dan Wood]

Dick
@bavbob and I both have done this recently without the parallel path. My post is below and Rob's is later. We followed what @HB Chris had previously completely. Per Rob's suggeston, we both used the rear defog switch.

What did you do to your E9 today?

Fabricated a bracket for the in line fuel filter mounted to the firewall

e9coupe.com

 

[bavbob]

Here is how I set mine up in parallel. I ground the edges off the one way valve to lower its profile to fit and not cause the cover that goes over this to wobble. I did this slowly as to not overheat the thing and ruin the internal valve. Held down with a bit of Velcro. I used 8mm T's and I give a big thank you to halboyles for that little plastic up size piece that goes from 6mm output of the sender to 8mm. When I had the set up in series I got a vapor lock and the car died. Chris has not had that experience with his in-series setup.

I call this my Roman Aqueduct system.

 

[Dan Wood]

Same as Chris, I did not have any issues vapor lock issues with the pump in series.

 

[Dick Steinkamp]

Hmmm...

John says...

You need the parallel path because the main pump can’t pull fuel thru the pierburg

Pierburg shows a parallel path.

@bavbob tried the series setup but it worked only with the parallel setup

It looks like @Dan Wood and @HB Chris didn't use the parallel path and had success.

I guess it wouldn't hurt to try it in series and go to parallel if it doesn't work for me.



I love the idea of using the rear window defroster switch and circuit to energize the pump. the switch and wire to the rear of the car is already there and I can't imagine a time when I'd be driving the car when the weather is nasty enough to need the rear defrost.

 

[HB Chris]

Yes, no issues for me and no loop, this is in my NK, it pulls fine after my five second initial fill. (And the rear defrost probably doesn’t work anyway!)

 

[Dan Wood]

I added a 5 Amp fuse btwn the defog wire (behind the seat) and the pump wire. The pump is 2 Amp. On mine, fuse 8 is for the defog, AC, and left power window (25 Amp).

 

[Mike Goble]

What is wrong with a single inline electric pump? I have a VDO pump in my tank that feeds my carburetor directly.

 

[JFENG]

You need the parallel path because the main pump can’t pull fuel thru the pierburg

My hypothesis: if this is a sliding vane type pump, it can block flow thru depending on where it stops in its rotation. Depending on the pump design, this might only happen at a specific rotational orientation range. That means it won't happen every time, but it can happen sometimes. So, I hypothesize Rob's electric pump was turned off and stopped in a position that blocked thru-flow, preventing his main mechanical pump from drawing any gas from the tank. Had Rob 'bumped' the Pierburg once, it might have solved the problem.

I'm happy to be proven wrong on this because it'll save me $15 of stainless u-bend Y fittings.

 

[Dick Steinkamp]

"Had Rob 'bumped' the Pierburg once, it might have solved the problem"

Good tip.

 

[Mike Goble]

How about one of these?

 

[JFENG]

"How about one of these"

remember those old cars where you had to periodically operate a foot or hand pump to add air pressure to the fuel tank so it would deliver gas to the carb?

 

[JFENG]

What is wrong with a single inline electric pump? I have a VDO pump in my tank that feeds my carburetor directly.

Nothing, and that's what I had on all my E3's and carb'd 2002's. But .... BavBob's goal was to keep the original mechanical pump and add a priming pump.
FWIW, I have 38/38 Webers on my E9, and it does not take tons of cranking to start after being parked for a month. IDK why some have issues and some don't.

 

[HB Chris]

My Zeniths can sit for a month and it only takes 15 or seconds to get gas and start the coupe.

 

[lip277]

My Zeniths can sit for a month and it only takes 15 or seconds to get gas and start the coupe.

I was the same on my 2800CS. They worked great that way. The 32/36 DGAS Weber on my 2002 would take a while to prime up after sitting for a while though. Both cars were on their original style (mechanical) fuel pumps.

 

[bavbob]

BTW, recall the pump flow rate is 1.5 liters/min.

Now John can determine the x-sectional area of his 8mm tubing and multiply by the distance to the mech pump to determine the volume to prime, known as the VTP. Then know how many seconds to prime, install a timer in series with the defrost button and voila, the perfect set up..............or count to 5.

 

[Dan Wood]

My hypothesis: if this is a sliding vane type pump, it can block flow thru depending on where it stops in its rotation. Depending on the pump design, this might only happen at a specific rotational orientation range. That means it won't happen every time, but it can happen sometimes. So, I hypothesize Rob's electric pump was turned off and stopped in a position that blocked thru-flow, preventing his main mechanical pump from drawing any gas from the tank. Had Rob 'bumped' the Pierburg once, it might have solved the problem.

Is it possible that the pump is still rotating after being shut off as long as the mechanical pump is pulling fuel?

 

[jmackro]

05 is an anti backflow valve so that the output of the pump doesn't just go in a loop.

I'm not seeing how "the output of the pump could go in a loop", unless you installed the one-way valve backwards. But I do see how the one-way valve provides a path to the primary pump ("06" in the diagram) that doesn't impose the resistance of the auxiliary pump in the "off" state.

Question: What are sources for the one-way valve and those u-bend type Y fittings that keep the hoses tidy?

 

[Dick Steinkamp]

I'm not seeing how "the output of the pump could go in a loop",

It could be my wording is misleading.

Without the one-way valve, the fuel coming out of the pump might meet too much resistance to go toward the carbs and make a right turn and return to the inlet of the pump.