I've been installing fuel lines on my 900 van which is designed with a fuel return. As I was equipped with the materials and tools for this job I decided to make yet another modification to my 500.

The carburettor return spigot and the decision over whether to blank it off or route it back to the supply pipe has been discussed and debated a million times here. I have decided to make a full fuel return pipe back to the petrol tank, as fitted on the Fiat 126 and as very well described here:

https://www.flickr.com/photos/ricambifiat500/albums/72157622038829612

Feel free to suggest the alternatives again or to describe this modification as pointless or OTT, but some people may want to try this so my images might give some pointers.

There has only been one occasion in three years when the benefit of the cooling and de-pressurising that a return line gives might have helped avoid a problem. That was in very hot weather, driving hard and then going downhill for quite a while when in gear, with the fuel almost static at the carb and in the pipe. It developed a bad vapour-lock which took ages to sort itself out. But in the interests of completeness and to stop myself from background worries about the possibility of the blanked off return pipe "letting go" I've bitten the bullet.

At the tank I will be modifying the fuel sender mechanism and soldering-in a piece of copper fuel pipe as the return. I may be able to adapt a spare one I have for the van; it is the same size and fitting.

I've drilled through the bulkhead just next to the hole for the supply pipe and then routed the flexible pipe down the inner wheelarch and footwell alongside the supply pipe. I didn't want to put more pipework inside the already congested centre tunnel so the flexi is tucked against it, under the floor-covering on the passenger side. It exits under the car through another, new, grommeted hole.

From there I've used copper pipe to follow the route of the existing solid fuel supply pipe. You can see that I also need to renew that pipe!

The trickiest part of the route was finding a way up to the carb. next time the engine is out I will add a couple more clips. I have put two clips close together where the short flexi from the carb spigot is connected in order to resist the effects of engine vibration.

The pipes have all been secured using rubber insulated steel clips and self-tappers and the ends of the copper-pipe have been flared to give a friction grip on the flexible pipe.

It's not in operation yet and ordinarily it's not going to make any discernible difference. But given the extremes of use my car is given, this will give me even more confidence and peace of mind.

Probably a good idea and looks well thought out. You could go the easy route a buy a new 126 sensor with the twin pipes but that would probably make it too easy for you and you like a challenge plus it means more ££££’s to spend!

I find that if I leave my 500 with the 28 IMB and not run it for a while, because I have a ‘T’ in return pipe, it seems to have a problem getting keep the natural flow of fuel down to the pump. I have to attach a little hand pump onto the fuel line to the pump to prime it. It’s fine once it’s running. All I can think is the the T piece is interfering. I could fit a one way valve in the return line but it isn’t really a big enough problem to get too stressed about.

I put my return via a ‘T’ piece back into the main fuel line in the engine bay as it goes from the original metal pipe to the fuel hose to the pump..

When I bought my Fiat, it already had a 650 '126' engine installed, so it had the 28IMB carb fitted. The fuel return on the carb had been blanked off, but somebody at sometime had tried to fit a return because it has a 126 tank unit and there was a length of non-connected fuel pipe running through the car.I decided to simplify the system so took the return fuel line down to a metal 'T' piece on the side of the engine bay, ahead of the fuel pump. The 'return' pipe is led behind the cooling-air trunking, through a length of small-bore water hose to protect it. All is held in place with rubber faced 'P' clips. To date, I have had no hot starting problems at all. I have retained the same fuelreturn system for the FZD carb now on the engine.

Vitesse said:

You could go the easy route a buy a new 126 sensor with the twin pipes but that would probably make it too easy for you and you like a challenge plus it means more ££££’s to spend!

Click to expand...

I thought about that but they're going to be matched to the random shape of the 126 tank. I also have real confidence in my original setup telling me when the fuel is down to five litres, which is important when you're fuel gaugeless and regularly travel hundreds of miles into the wilderness.

I'll maybe find away to fit my existing sender rheostat and float/arm onto the double-elbow fitting. If not, I can see that soldering does the job OK.

fiat500 said:

I thought about that but they're going to be matched to the random shape of the 126 tank. I also have real confidence in my original setup telling me when the fuel is down to five litres, which is important when you're fuel gaugeless and regularly travel hundreds of miles into the wilderness.

I'll maybe find away to fit my existing sender rheostat and float/arm onto the double-elbow fitting. If not, I can see that soldering does the job OK.

Click to expand...

Just drill a hole in the existing sender as you said and weld a bit of pipe to it. I’m sure you could probably do it in situ with it attached to a full fuel tank, it would stop it moving around whilst welding. I’m sure it will be fine! ?

Vitesse said:

Just drill a hole in the existing sender as you said and weld a bit of pipe to it. I’m sure you could probably do it in situ with it attached to a full fuel tank, it would stop it moving around whilst welding. I’m sure it will be fine! [emoji91]

Click to expand...

Good idea... that'll save me a bit of time!. But to be on the safe side I'll wait until. I'm low on fuel and use the welder on low power. [emoji3]

Vitesse said:

Just drill a hole in the existing sender as you said and weld a bit of pipe to it. I’m sure you could probably do it in situ with it attached to a full fuel tank, it would stop it moving around whilst welding. I’m sure it will be fine! ?

Click to expand...

That is just a ridiculous thing to suggest!!!!!!!...………….Surely metal swarf will fall into the tank and have the potential to be sucked through to the pump...…..

The Whitakers said:

That is just a ridiculous thing to suggest!!!!!!!...………….Surely metal swarf will fall into the tank and have the potential to be sucked through to the pump...…..

Click to expand...

No, you idiot, obviously I'll turn the engine off whilst I drill the hole.

fiat500 said:

No, you idiot, obviously I'll turn the engine off whilst I drill the hole.

Click to expand...

Good call...….should be 'reet' then...…….

It turned out that my alternative sender unit was all kinds of wrong so adapting the existing was best.

It's 50 years old and apart from a little plastic inspection cover that has been lost at some point, it looks like new. Note the very fine filter on the intake tube and the complete lack of any debris over it.

I did remove it from the car, (just in case anyone took previous comments literally!), and drilled a hole out to 5mm. Then I used the 6mm section of a stepped drill to punch out the final 1mm for the copper pipe. This made a good, tight fit and created a slight lip which provided stability and something for the solder to get a grip on.

I used plumber's solder and flux after cleaning back to bare metal; the whole effort took about half an hour.

Obviously I had to test and demonstrate the amount of fuel that is returned to the tank.....and it's a lot. I got about half a litre in a couple of minutes. The video (sorry...out of focus and there really was a helicopter overhead at the time!) is on a fast tickover with the car not warmed up. When the choke was fully on and the engine revving high the flow was obviously greater.

It seems that whilst possibly unnecessary, this modification does promise to give good results in extreme conditions.

[ame]https://youtu.be/ZgNEF7nFQ2g[/ame]

Vitesse said:

All I can think is the the T piece is interfering. I could fit a one way valve in the return line but it isn’t really a big enough problem to get too stressed about.

Click to expand...

So are you saying that no amount of cranking will pull fuel through once the pipework is empty?

It sounds familiar from the first rebuilt engine I tried to start; it look ages to get fuel through until I realised my bench testing perhaps needed the petrol can placed a bit higher. The 500 setup, with the fuel tank above the level of the pump, means that once the delivery pipe is full, fuel gets to the inlet of the pump by gravity. Obviously once pumping there will also be a vacuum created in the pipe to assist this. You normally don't lose that head of siphoned fuel when the engine is turned off, although I think that fuel does evaporate from the carburettor bowl over a period of days.

It will be the non-return valves in the pump which maintain the vacuum which stops fuel draining back to the tank. Is it possible that the three-way connector is allowing a bypass of those valves so that it loses its vacuum to the carburettor inlet valve? This will gradually open to atmosphere once the level in the bowl drops, thus allowing fuel to reverse siphon back to the tank.

I've read somewhere that the way the 500 and 650 pumps work is different, although I have no evidence or experience to prove it. But the 126 petrol tank is below the pump so it seems logical that it might be designed for more positive suction. Which pump do you have?

fiat500 said:

The video (sorry...out of focus and there really was a helicopter overhead at the time!)
https://youtu.be/ZgNEF7nFQ2g

Click to expand...

It was probably the fire brigade!!!!!!!!

Looks very professional. I am not being sarcastic for once.

fiat500 said:

So are you saying that no amount of cranking will pull fuel through once the pipework is empty?

It sounds familiar from the first rebuilt engine I tried to start; it look ages to get fuel through until I realised my bench testing perhaps needed the petrol can placed a bit higher. The 500 setup, with the fuel tank above the level of the pump, means that once the delivery pipe is full, fuel gets to the inlet of the pump by gravity. Obviously once pumping there will also be a vacuum created in the pipe to assist this. You normally don't lose that head of siphoned fuel when the engine is turned off, although I think that fuel does evaporate from the carburettor bowl over a period of days.

It will be the non-return valves in the pump which maintain the vacuum which stops fuel draining back to the tank. Is it possible that the three-way connector is allowing a bypass of those valves so that it loses its vacuum to the carburettor inlet valve? This will gradually open to atmosphere once the level in the bowl drops, thus allowing fuel to reverse siphon back to the tank.

I've read somewhere that the way the 500 and 650 pumps work is different, although I have no evidence or experience to prove it. But the 126 petrol tank is below the pump so it seems logical that it might be designed for more positive suction. Which pump do you have?

Click to expand...

It starts fine on using what is in the carb bowl but will conk out after a few minutes due to lack of fuel getting to the pump. If I pull the fuel line off the pump there will be no fuel coming out of it, even when it’s running. What I have to do is get my little hand pump, that basically is egg shaped rubber ball and has fuel pipe on each end of it. As you say stick one end in a can of petrol up high, the other end to the fuel pump. A couple of pumps so that it primes the carb again, start the engine and after a few seconds of running, the fuel from the can will then get returned from the carb via the return line to the T piece and come out of the pipe that is meant to be attached to the fuel pump from the tank, then you just quickly swap over whilst the engine is still running, obviously you don’t have fuel squirting everywhere as you have you thumb over the end of it until the last second.

I think your explanation about the vacuum and the T piece makes sense, effectively I’m recreating the vacuum by filling the empty piece of pipe back to the T piece. It has the pump that you can take apart, with the bolt on the top and the screws around the middle of it.

I too have found that after a period of not pulling fuel from the tank (engine out etc) it can be very difficult to pull fuel through. In my case, I gently pressurise the tank with a small tyre inflation pump, simply sealing around the filler neck with rag. This has always been all that is needed to push the fuel through to the pump. I have a 500 with a 126 engine (and fuel pump) installed.

Vitesse said:

It has the pump that you can take apart, with the bolt on the top and the screws around the middle of it.

Click to expand...

Mine's the same. I think the key identifier is that physically it has a sloped base to clear the alternator as Tom has noted previously. But whether there is some difference in the internals? I don't know.

I guess your problem (which is not a problem ) happens after a week or so of neglect disuse? On the rare occasions when my car goes, say three days not used, it does need about three short crankings to get primed, but that's all. But up to now I've obviously just had the blanked-off return. We'll see what happens now.

I've had one short run in it with the re-jigged fuel system and the differences so far are; idle speed has increased with no adjustment and the very slight accumulation of misty, oily residue that comes from somewhere? and gathers on the drip-tray seems to have stopped.

Vitesse said:

It was probably the fire brigade!!!!!!!!

Click to expand...

It was a bit worrying because just as I started the engine I did smell burning; I think the helicopter was looking for a woodland fire!

Vitesse said:

Looks very professional. I am not being sarcastic for once.

Click to expand...

.....I'm still waiting for the punchline....

You have possibly introduced a little more air into the system due to the additional joints, which is maybe burning off the fuel better hence the lack of oily residue? I am not sure whether that would increase revs? I suppose it’s possible that maybe the mixture was a little bit out and you’ve unintentionally improve it hence the increase of rev’s?

Thinking back to when I installed the 126 'power house' and utilised the 'T' piece I had problems priming my fuel lines. My drive is slightly down hill which with the small amount of fuel in it at the time (I don't like repairing cars with large tanks of fuel) I probably lost the self prime scenario and from memory I took off the return pipe at the carb and blocked it of with my thumb whilst the car was cranked over to get it to prime to which it did no problems. (I had a fuel filter hanging dangerously hanging over the alternator so I could view the fuel scenario)……….it starts on the button pretty much now but I've never let it run dry so cant comment on if that scenario..

I hope you are NOT going the weld ?solder any thing in situ on the fuel tank or you will need a new garage when it blows up in your face .