I need to understand how these tanks work myself, thanks for posting up your research and experiences!

I solved the running out of fuel problems I had with a new siphon jet pump. But now the tank is eating floaters, I think there is a pressure problem that causes the tank to contract and contact the float.

Somebody else had a problem of the left sending unit getting crushed. The thought was the roll over valve for shut and no vent to outside to let in fresh air. (there is also a set of valves for testing leaks that could be at fault)


Sent from my iPhone using Tapatalk Pro

e53 Fuel Supply Detailed (fuel pump / siphon pump diagnostic help)
 

You said you pumped 12g from the right side. It only holds 5L before flooding to the left side. When you put it back in 5L fills the surge tank then the rest is shared left and right. I'm not sure the ratio but actually I think close to 1.

That would mean about 40L/ 2 or 20L per side.

You said you had 18.2 left? That is not a normal value for flood stage it should be 28-29L.

If left side was at flood stage that's 29L and 12g is about 45.2L. If you started at flood stage and the siphon moved 5L to the right you should have 23L left and right with 5L should read 1.5 ish.

Put the 45.2 back in and the first five gets you to flood stage and then the remaining 40L is split left and right.

28+20 left = 48 and
1.5+20 right = 21.5

Total of 69.5 L or 7/8 of a tank.

Think about this: you removed half a tank (12g of gas) putting it back in should add half a tank not end up at half a tank.

I currently have a problem where my left sending unit at random times adds a random amount of fuel between 10 and 30 L.

There is a tech bulletin describing the problem and the fix basically need to add some pressure to the resistance wiper. Hoping that's all it takes and I can add one more fuel system fix to my repair catalog.

Another good excuse to get an Arduino on-board! This could be monitored along with engine temp and crankcase vacuum. ;)

Is that the system that uses an Android interface you can program? Yes that would be awesome to program some examples.

I've been saying since I started this thread it's important to drive down to single digits DTE to confirm siphon operation but you can do most of the important monitoring with the test six.

Right side will stay exactly the same from 28 down to zero left tank unless siphon jet has a problem. If the right side drops down to zero when there is 1/4 tank the siphon jet needs repair for example.

If the left side ever reads over 29 when right side is at flood stage (1.4L typical) then the sender is reading high like mine is and you will run out of fuel because the gauge reads plenty of gas.

I had a similar problem of a high reading float when I was working on this issue. Mine was an improper install bh the tech where the float actually hung up on some hose in the tank.

That might be me lol! Anyway, could you elaborate or link to more info on those tank system components you mentioned?

Unfortunately I've not dug into those systems. There is the system that shuts off the outside vent save a known size orifice to measure the current on the air pump that pressurizes the tank to measure for leaks.

Then there is a master valve that shuts off the vent to the charcoal canister. If that valve is stuck shut you'll have the same problem. I don't know where the parts are but some might be in the left rear wheel well.

Open your gas cap after a long drive if you hear a crazy vacuum suck the problem is for sure the vent not opening


Sent from my iPhone using Tapatalk Pro

Irregular R/S Float Reading corrected.
 

Yep, I binded the float on the R/S too; on Saturday's install.
Corrected and will monitor.


G

So I finally got to repair my fuel gauge issues today!

Some big helpful findings!

First: the distance from the bottom of the fuel guide post to the center of the floating float at flood stage (left below 28L) is almost exactly 2 cm. Makes it extremely helpful when trying to calibrate the right fuel sender !

My fuel arm was way off from when I got my replacement complete unit. It displayed 5.6L or something at fkod stage. I went to fix it and went way off the rails. When I had the fuel pump out to figure out why I stalled no fuel a few weeks ago I discovered the pump was fine the left sender was reporting 33L of fuel when there was ZERO.

Since I had the right side out I bent the float arm based on product photos and got sorta close: reported 0.2L at flood stage which if the parts never failed would be near perfect (0.0 would be ideal).

The problem is that when things aren't working properly and you are using test six to monitor say the siphon jet, you NEED that 1.5L of warning to get to a gas station. 0.2 is "background noise".

So I had two goals today: 1) fix my left side sender (success I will make a new thread to show how to do that) and 2) calibrate my right side sender to get something like 1.5L at flood stage.

https://uploads.tapatalk-cdn.com/201...f1f43a5610.jpg

Bingo! Also success!