DS21 ie fuel tank issue.

Buttercup

Buttercup

D Stracted
VIP Paid Subscriber
1000+ Posts
Fellow Frogger
Joined
Jan 20, 2013
Messages
3,907
Location
Lake Keepit NSW
I've probably missed previous discussion about this.......Sorry.

My DS21ie has recently been sucking air, when the tank is still about 1/3 full.
So I thought I'd investigate..... not knowing what the designers had installed in the way of baffles strainers and plumbing.
The symptoms indicated that (when less than half full) enough fuel would come through to maintain idle, but at higher power, after a few minutes, it sucked some air. After sitting roadside with the engine off for a while it recovered.... then did it again after a couple of minutes on the road.


Yes, I confirmed fuel pressure at the engine, with a gauge Tee-ed in, and watched it while throttling up.....
It never budged off the 2.0 bar setting, until it started to suck air.

1st step is drain the fuel.
While I could have done it via the drain plug at the left, I thought I would do it at the strainer plug at the right, and monitor the outflow..... as this was the chamber from which the pump sucks the fuel.
I removed the brass plug, and the plastic strainer.......
As expected, after an initial gush, it settled back to a very moderate dribble, for at least 1/2 an hour, delivering about 20 litres in that time.
Not a lot of crap came out, maybe a teaspoon of fine rust flakes.
I then took the gauge sender out, and peered in with a torch..... there's still a significant amount of fuel in there, but nothing coming out!
So I put a sucker hose in with an electric pump, and got a further 8 litres out.

So, I conclude.....
the pickup strainer is not limiting the flow.
The tank has plenty of fuel in it (20L should be plenty)
The fuel is not getting into the chamber around the pickup, at an adequate rate.
Peering in with a torch seems to indicate the chamber is a litre or two, and has a hole around the tube that encloses the pickup line, at about 1/2 tank level.
So when the tank is above 1/2 full, fuel freely goes in that hole, but at lower levels, inflow is severely restricted.

Has anyone experienced this before.
Has anyone discovered, invented or implemented a solution?

I have lots of fuel tanks in my collection, however only 1 other injection car.
I am prepared to do surgery on a tank to explore, and hopefully discover the designers intentions, and what has created this fault.
Apart from the return pipe, is the injection tank different to the carby tank?

All advice will be considered!

Thanks in advance
 
Injection cars were notorious even when new for an inability to maintain fuel supply on heavily cambered roads if less than a quarter full.
The transverse design and the fact that the fuel extraction hardware is on the right side of the car would have them sucking air instead of petrol. Your problem may be design fault or the result of vague and hopeless fuel gauge lies...or some age related pipe fracture.
 
fritzelhund said:
Injection cars were notorious even when new for an inability to maintain fuel supply on heavily cambered roads if less than a quarter full.
The transverse design and the fact that the fuel extraction hardware is on the right side of the car would have them sucking air instead of petrol. Your problem may be design fault or the result of vague and hopeless fuel gauge lies...or some age related pipe fracture.
Click to expand...
No, I'm sure it's NOT a camber issue.
And NOT a gauge issue.
I got 28 litres out of the tank, after it showed the problem on relatively level road.
The fact that the draining from the delivery chamber (with the car perfectly level) was very slow, indicates a tank internal issue.
It's baffling!
 
Ok ....... open tank surgery it is!

And while I'm in there, I'll convert it to left side pick-up.

I'll open a fairly big hatch in the top panel at the left, install a baffle, pickup chamber, pickup tube with strainer, and bolt on hatch cover.
 
Bob,
On one of the many occasions that I could not start my SM IE, I found the problem to be no fuel at the injectors or the fuel lines (rather than the more usual ignition problems). I had recently changed the leaking rubber pipe on the outlet of the pump going onto the nylon fuel line heading up to the front of the car so I checked that and it was still OK. So then I opened the boot and took the cover off where the fuel lines go into the tank and found two nylon lines coming from behind the tank that were joined to the tubes from the tank by short rubber segments. The source line rubber pipe was still secured by ligarex straps so I thought it might be original and it looked ok from the top view. However, when I put my hands on it to start to disconnect it, it just shattered in my hand. It was quite solid, seemed to be quite porous and obviously had a largish hole in the bottom that allowed the pump to suck air when the pump was running (and for petrol to pour out onto the top of the tank when the pump stopped). I managed to find a Repco that had a short length of 12mm fuel hose in stock and after I replaced it, I still couldn't get fuel up to the rail until I disconnected the output of the pump and it was then able to successfully prime from the tank and after reconnection, onwards up to the fuel rail.

Cheers, Ken
 
it sounds like there must be baffles in the tank that have been closed up with rust/sediment .... Can you get a USB camera down in there first before you start chopping at it (fuel tanks make me wary ..... I'd prefer to keep well away from them with welding/cutting tools!
 
I'm ok with them, having made 2 for my 2CV modifications, and repaired several over the years.
I've already cut a small inspection hole, with a hole saw, after completely filling it with water.
I'll fill it again, before cutting the access hole, starting with the hole saw, at each corner of the rectangle.
Rectangle so I can insert a hatch retainer/stiffener inside, before fitting the lid.

I'm quite looking forward to having a LHD tank that works.
20210906_145912.jpg
 
The IE fuel tanks are notorious for problems delivering fuel when the tank drops below half a tank. I think it's partly because the designers were still experimenting given the DS was one of the first mass produced cars with electronic fuel injection.

The fuel pickup lives in a reservoir that holds a litre or two. The reservoir is filled by the fuel rail return pipe and some very small slots at the base of the walls.
IMG-20191001-WA0010.jpg


This image comes form here https://www-salvadsie-fr.translate....oN7xIYNg8Z13z7iZk37Vsq6SmKy0meiWCBVgxTmfJRWd0

A first check Bob is to take out the brass plug and check the pickup tube actually hits the bottom of the tank. Also, if the brass plug has the stack of filter discs on it, remove them. IE cars should not have the in tank filter that was fitted to carby cars.

I've read reports from people in Europe who have had massive problems after using epoxy coatings to seal old tanks. It blocks up the slots that allow fuel into the pickup reservoir. If you're going to cut open the tank then I'd look at fitting the swirl pot from a VL/VN commodore or pretty much any 90's fuel injected merc.

vr-fuel-swirl-pot-jpg.76181


swirl-pot-jpg.456718


They use the pressure of the returning fuel to force fresh fuel from the tank into the pickup area. Supposedly they'll run down to about 1L in the tank.
 
Thanks for the pic, it tells a thousand stories.
Yes, the pickup was low in the well over the brass plug.
No there was not a stack disc filter, just a strainer frame with no mesh.
When removing the tank from the car, it was clear that it has never been out before. All original fittings and rubber and the car assembly spec tag were intact.
20210906_154108.jpg

So it's not had any sealer in it.
Inside is remarkably clean.
A small amount of surface rust.
Not surprising because in the known history of the car since 1975, it had always been garaged.
Original leather is still 99%, car is rust free.
I guess, yes rust flakes had gradually closed up the fuel delivery slots.
The previous owner had reported annoying intermittent loss of power, although he only did about 2,000km in the last 10 years. And probably didn't let it go lower than half full.
20210906_154839.jpg
 
In one of my 2CV 60L tanks i installed baffles about 50mm high in this pattern.
20210906_160257.jpg

When it ran out, there was only about 1 teaspoon left. I didn't install a gauge.

On the first Raid that I took it on, I wanted to test it, and my range, so I deliberately ran it out several times. Of course we had 2x20L Jerry cans onboard.
First-time it ran out at 898km (slight headwind), second time, 903km (calm day)......
On the third time, at 900 I remarked to my passenger that I expected it to run out soon.
Within 500m it did!

Thereafter I could predict it very accurately from the mileage.
It still doesn't have a level gauge.
 
Roger Wilkinson said:
Don't worry Bob, I still love you.

Roger
Click to expand...
What's wrong Roger?
Are you upset because I want to "Fix" the car, and I'm hacking into sacred original steel lovingly caressed into shape by Andre Citroen himself?

Bad luck, I want to enjoy this car without worrying about keeping the tank topped up every 100km!

I want a DS, that is how it should have been designed.
 
Buttercup said:
What's wrong Roger?
Are you upset because I want to "Fix" the car, and I'm hacking into sacred original steel lovingly caressed into shape by Andre Citroen himself?

Bad luck, I want to enjoy this car without worrying about keeping the tank topped up every 100km!

I want a DS, that is how it should have been designed.
Click to expand...
Whem my ds had this issue we found the ryslen feed pipe in the tank had shrunk with age. Adding a little standard rubber fuel hose in the full to extend the reach Dom Ed the problem
 
PeterMol said:
Whem my ds had this issue we found the ryslen feed pipe in the tank had shrunk with age. Adding a little standard rubber fuel hose in the full to extend the reach Dom Ed the problem
Click to expand...
Sorry. Spell check. Added the hose in the sill
 
If there is sufficient height to work with and you are cutting into the tank, you could go the old SAAB 900 route with a pump inside a bucket in the tank and the bucket refilled by a small lift pump. Fuel can slowly flow into the bucket by itself should the pump fail and leak back out at a controlled rate. A remote surge tank is employed on the Jaguar XJS, but that it below the level of the tank and the DS has the fuel tank far too low for such a gravity-fed arrangement. Gravity is not your friend here.
 
The Bosch ie pump is fine, I don't see any reason to rethink it.
I just need to make sure that fuel will flow to it unimpeded, and without air.
Because Citroen didn't put any filter in the tank, but did fit a large capacity inline filter just before the pump, I'm not concerned with tank junk getting taken out with the fuel. In fact it's good to do that so the tank doesn't gradually fill up with crud.
I have decided to employ the interlocking Vee baffles, the same as I sketched above, cos I know they work and they are easily fitted through my access hatch.
The return flow from the ie pressure valve will stay where it is. I see no need for a swirl pot. Petrol doesn't foam like diesel.
 
You must log in or register to reply here.
Top