Mercedes-Benz Forum banner

Fuel Tank Strainer, Fuel Tank Insides, Swirl Pot, Return Line Venturi

49K views 92 replies 22 participants last post by  Ruindr 
#1 ·
.






Greg in Oz said:
I agree that it will likely need the fuel tank cleaned. The symptom of fuel starvation only as the fuel level in the tank drops to around a 25% is typical of a dirty tank in an MB.

The photos above help show how the MB tank has a swirl pot surrounding the strainer where the fuel is drawn from the tank. The fuel return line enters the swirl pot perimeter via a venturi. The venturi allows the return flow to draw fuel from the bottom of the tank into the swirl pot allowing the swirl pot to remain full and overflowing, even when the tank is almost empty. In fact, when this system functions correctly the car should continue to run until the tank is almost completely empty with only the fuel in the swirl pot remaining. A full swirl pot prevents air being drawn into the strainer.

If the tank is contaminated with sediment, rust or the like, the venturi at the base of the swirl pot becomes blocked. When this happens fuel can only enter the swirl pot from the opening at the top. This works fine if the tank is above half full. At lower fuel levels the swirl pot will begin to empty allowing air to be drawn into the strainer causing the engine to suffer from fuel starvation. This will become worse as the fuel level drops and as the fuel sloshes in the tank with the movement of the car. When your car suffers from fuel starvation, pull over and remove the tank filler cap. With the engine running give the car a shake. The engine will likely hesitate and you should be able to hear air being drawn into the strainer (via a vortex caused by the swirling fuel in the pot).

This problem can be difficult to diagnose as the strainer can be removed and will usually appear clean. The issue is not one of the strainer (or fuel filter) being blocked. If it was the fuel starvation would occur regardless of the amount of fuel in the tank. The key to diagnosing the blocked venturi is the symptom of the fuel starvation occuring only when the tank is less than around 20 to 30% full.

You may be wondering how I know this. It was only through having to diagnose the very same problem with my 500SLC after it had been stored for several years (with deteriorating fuel in the tank). I only solved the problem after removing the tank and cleaning its internals thoroughly with a pressure washer (followed by a flush with methylated spirits and then airing in a sunny loaction for several days to ensure it was completey dry).





Added 07/16/2012
 
See less See more
7
#2 ·
Hah, I suspected that guys problem was something akin to this, but I never imagined anything that convoluted. On the one hand, it seems like a good idea engineering wise--to get the last drop of fuel, on the other hand, given the filth one can expect to accumulate at the bottom of a tank, it's not such a good idea.

I was always taught never to buy gas from a station that runs out--ever, you will get the crap from the bottom of their tank. And I was taught to fill the car when down to the last 1/4 tank.
 
#4 ·
Cleaning out blocked swirl cup

My SL sat for a few years in Florida and after having replaced the fuel pump, filter and accumulator I am still experiencing fuel starvation at 25% fuel level. If I understand correctly, this is due to blockage of the fuel inlet at the bottom of the swirl cup. Is there a way open that blockage without removing the tank? Say, through the fuel sender opening at the top of the tank?
 
#5 ·
My SL sat for a few years in Florida and after having replaced the fuel pump, filter and accumulator I am still experiencing fuel starvation at 25% fuel level. If I understand correctly, this is due to blockage of the fuel inlet at the bottom of the swirl cup. Is there a way open that blockage without removing the tank? Say, through the fuel sender opening at the top of the tank?
Get underneath the rear of the car and follow the big ass fuel hose from the fuel cluster to the tank above.

That fuel line at the tank end is where the tank strainer is. You might get lucky and unscrew the hose WITH the fuel strainer. If not....the strainer comes out with a 46MM socket, crows foot, blessing from the Pope.
 
#7 ·
#9 · (Edited)
.


Ummmm....post #1?

As mentioned, I can't see from those pics just how the venturi is configured.

In a typical venturi pump, the motive fluid (fuel return in this case) would pass straight through while pulling the secondary fluid through a side nozzle. I don't see that in pics, but maybe it is just me.



It could also be that the fuel return just passes through an oversize hole in the swirl pot and draws in fluid through the annulus.

Or something else?

Also wondered if plastic like Porsches. Could determine type of cleaning fluids to use.

This is better picture of venturi. It looks like the fuel may enter through an opening just above the fuel return line? Venturi nozzle built into casting?



This is a swirl pot off a different car, but it is of annular type and external. Jet from return line draws fuel from tank before it enters swirl pot

 
#10 ·
We were able to get a length of steel rod from a welding supply store and manipulate it enough to get it through the fuel sender opening into the opening. Since ours was gunked with fuel tank liner/epoxy, we used a torch to heat up the steel and jam it in through the opening. Mind you, we had the tank off the car while doing this, and it took a long time. After a bit of time, we took a hose syphon tool and hooked it up where the fuel return line is and pumped water through it to see if water would enter via venturi and it did. Whether it allows enough in or not, I'm not sure.

But you'll most likely want to take the tank out to do this. I couldn't imagine doing it while its still in the trunk. Also, is the strainer is clogged up it could cause your problems. Either of those senario I would think the wisest move is to remove the tank and get it cleaned. If something has clogged up your strainer or venturi, then there is more debris/rust in the tank that is just waiting to makes its way in there IMO.
 
#11 ·
After a bit of time, we took a hose syphon tool and hooked it up where the fuel return line is and pumped water through it to see if water would enter via venturi and it did. Whether it allows enough in or not, I'm not sure.
Is it possible to determine if the venturi is working - in other words, is it sucking fuel from bottom of tank outside of swirl pot. And if not, what to do about it. That is reason I was asking questions about how exactly the venturi is configured. Anyone got an old tank ;)

When the car is running, there is more fuel leaving through the strainer than there is returning to swirl pot. When tank level is low, the venturi would have to suck in enough fuel from tank so that swirl pot stays almost full. (I don't think is has to, but it would be better if it did)

From what I can see, the swirl pot performs two functions. One is as a baffle to prevent cavitation in pump suction line when levels are low and fuel could be slopping around in tank. The second, also with low levels, is to cool hot return fuel by mixing/contacting with cooler fuel in tank to help with hot starts.
 
#12 ·
IMO If you get below a quarter of a tank or so and don't start having fuel starvation issues it's working. As I've regularly run until the reserve light comes on, I will assume mine is working properly.:thumbsup:
 
#19 · (Edited)
nobby, MBGraham, isthisdave, hinero7 and abiby - thank you all for chiming in.

MBGraham is right: I was interested in a way to clear a possible blockage of venturi-supplied fuel port at the bottom of the cup.

I also hear nobby's "get the strainer replaced and then see" - The strainer is one item I have not yet replaced. But I still believe that if I am getting a reliable supply at above 25% fuel then it is not the strainer.

Wondering if it is possible to apply compressed air through the return line in trying to blow out the blockage? And is it possible for the return line to get clogged? The engine runs great. Would it run the same if the return line was blocked? Say, I removed the fuel quantity sender and looked inside the tank while the engine was running - would I be able to see the fuel return stream?
 
#21 ·
"Removing tank and doing a complete clean up - new strainer, tank flush etc might be a worthwhile investment even if you have to get it done professionally."

Agreed, but unfortunately, the tank removal not an option at this time. The car runs fine for 250 miles on a full tank without starvation symptoms. That's with a 500SEL motor and 14 mpg avg.

I do plan on installing a new strainer though.
 
#22 ·
removing stuck strainer, cleaning gum with tank in car, clearing swirl pot venturi

dragonflyer: "Wondering if it is possible to apply compressed air through the return line in trying to blow out the blockage? And is it possible for the return line to get clogged? The engine runs great. Would it run the same if the return line was blocked?"

[this got long..]
Clogged return line :
The car won't run if the return line is blocked. Mine was completely blocked and fuel pressure was way too high (~100 psi with aftermarket pump). When I disconnected the return line and fed it to a coffee can, the engine started/ran. Someone said running a stiff wire into the return line at the tank cleared it, and this worked for me (MUCH gunk on the wire) - then I used a vacuum setup to pull fuel and gum through the line. This car sat for a few years and had much gum in tank (froze fuel pump).

Cleaning out gum with tank in car :
A couple months ago , I drained the gas, then dumped three 16 oz cans of carb cleaner in the tank to break down the gum - not much drained out. I hooked up the fuel lines, added ~1/2 gallon gas, and let it set in the hot sun for a couple days (once in a while bouncing the car around to slosh the cleaner). I pulled the line from the pump and fuel started dribbling out - by the time I got back to it, the gas had evaporated and left a mess of dried gum in the drain pan.

Removing stuck strainer :
Today I pulled the fuel line from strainer - last time the strainer came out with the hose - this time it came apart so I drilled it and used a #8 easy out [cheap set on ebay] to remove the strainer. I have no 46mm socket and couldn't have gotten the off-center strainer out with it anyway.

Second cleaning and clearing venturi :
I dumped another can of carb cleaner in tank with strainer removed and little drained out. I added ~1/2 gallon of gas and it started dribbling (maybe a quart). Dumped in another can of carb cleaner, disconnected the return line at tank, blew in compressed air and MUCH gas came out along with about a tablespoon of rust - fuel was dark with dissolved gum. Added 16 oz of gas (measured with empty carb cleaner can), then blew it out into a clean pan - got about 12 oz of cleaner gas but still had rust particles. Repeated this, got about 12 oz of clean fuel with a bit of rust, hooked up the fuel lines, and will let the filter do its job. Added about 1.5 gallons of gas (well below top of swirl pot).

Result :
I energized the fuel pump (shorted 7 and 8 on relay socket), it ran dry for a few seconds, then sounded "hit/miss" as fuel pumped though the lines, made bubbling sounds in tank until return line was full, then sounded normal after pressure stabilized.

I don't think rust alone was clogging the return line/venturi. Nothing I had dissolved the gum except carb cleaner - maybe 2-3 ounces of gum came out of tank. Would be good to flush tank with carb cleaner whether on or off car, as I don't think water would do much (aside from moving loosened rust particles, but seems it would rust again...). Eventually I'll move the car into garage and pull tank - maybe rig up something to pump/recirculate carb cleaner around in the tank.

Battery was dead so couldn't start the car - hopefully I didn't damage the fuel injection with gummy gas.
 
#24 ·
You're welcome. Much to learn even with basics on these cars. Couldn't do it without these forums.

I'm not sure what you mean by "with tank hose disconnected and plugged" (the 14 mm hose from tank to fuel pump? or ??? just to keep solvent in tank?).

After battery charge and priming, mine started easily - is right where I left it a few months ago (sounds like one cylinder has slight misfire [guessing a dirty injector]), but doesn't die with only 1 (less than 5) gallon(s) of gas in tank.

I'm still confused on why so much gas (10-12 oz) came out when I blew air into return line - it didn't take much air flow to do this. It looks like the inlet hole is in bottom of swirl pot, but maybe the tank is lower (a small "well" outside of swirl pot) and the venturi pulls in fuel from the well. After seeing the venturi diagram and the alternate swirl pot photo (on pg 1 of this thread), I wonder if there's an opening in the bottom of the venturi (labeled 'suction' in diagram).

Some tips if not done so:
1) While pulling the strainer, etc down, change the 14 mm braided fuel lines (I wasn't prepared and 14mm line can be hard to find locally - cheapest I saw was here, but I don't know quality Shopping Cart ). Bad to find the short hose is leaking at 3 am (soon the crack clogged up with rust particles [today is dried rust residue on outside of hose]).

2) Start clearing lines at the tank (I started up front and pulled/likely deposited gum through the long line from the tank to engine).

I'd do this quick check before attempting to pull carb cleaner through return line:
Remove the short rubber hose on the return line near the tank (enters at left front/drivers side of tank) - then connect a piece of rubber fuel hose to line from engine and (secure it) into container - energize fuel pump relay (or on/off key a few times to get 2 second flows) and fuel should go into the container [this checks everything from tank to engine and back to tank] - you can check flow rate at this point if you jumper the relay socket pins.

Then connect rubber fuel hose to return line going into tank - you should be able to blow/compress air into it.. if not, it's clogged (then try jamming a wire into it - I used ~2mm soft steel wire with a mashed 'hook' on the end to lessen chance of hangup). This thread details from beginning of clog and has some good info from posters Fuel Distributor leak caused by clogged return line 1983 380sl - PeachParts Mercedes ShopForum

3) you may want to make a "trough" to catch/direct gas/solvent when draining the tank with the strainer out - it ran and dripped everywhere - the large pan caught all, but gummy gas makes a mess.

Photos show what was in the tank. First is from last cleaning (that was a clean galvanized pan), and second is what I kept pouring into a large coffee can (minus a few leaves) - is 1/8-1/4" of gunk.
I can imagine what is still in the tank. It'll probably need a good spraying with solvent. The best it could be is like the galvanized pan after pouring out the gas/solvent/deposits, which isn't clean at all until I wipe it down.
 

Attachments

#25 ·
Yikes, looking at the first photo - that is what the bottom of my tank looks like when viewed from the fuel quantity sender hole at the top of the tank. I can also get a partial view of the swirl cup and the strainer.

Yes, I did mean to plug the 14mm line from the tank to prevent the carb cleaner from getting to the pump. I already replaced that line with new from the dealer. It was longer than it needed to be, so when I trimmed it, the remaining piece made a perfect bridge for a 90-degree connection to the pump.

What I want to do today is to get a few spray cans of carb cleaner and connect them to a piece of clear plastic vacuum line which I'll try shoving directly into the swirl cup through the fuel level sender hole. I'll suck out the fuel first and give it a shot with just two cans of cleaner. If that doesn't work, I'll try the 2.89 quart can of carb cleaner ($30 at my local NAPA).
 
#26 ·
Good info about the large $30 can of carb cleaner. With a pump hooked up, one could wash the gunk from edges towards the swirl pot (there must be a low spot in the tank there) and pull out the sediment as it builds up (with suction or ???).

Key with gum and carb cleaner is to let it soak a while (better if warm/hot) so gum dissolves. It was hard to scrub off the bottom of the drain pan once it hardened (unless I let it soak), which is how it would be in the tank.

Is your return line open?
Do you know if the problem is mostly rust, gum, or both?
 
#27 · (Edited)
It was mostly chunky flakes. What I observed today was that once the fuel level dropped below the swirl cup the fuel stream from the 14mm hose would gradually decline, becoming a 1/8-inch trickle at .5 inches from the bottom. It would not just stop. Also, when I flushed out the pump, out of it came that flaky debris, which leads me to believe the tank to swirl cup passage is blocked by a loose debris which does allow some fuel to get though and may become dislodged with compressed air or a larger quantity of carb cleaner.

Today, I did not evacuate all the fuel and instead, when the level got to be .5 inches from the tank bottom, I plugged the 14mm hose and using 2-foot long .250 od aluminum tube, taped the plastic squirt tube that comes with a can of carb cleaner to it and shot two cans into a swirl pot from the top. I let it sit for a few hours in 90 degree heat and then unplugged the 14mm hose. Less then two ounces came out, leading me to believe the carb cleaner dispersed into the tank from the swirl cub through the venturi. So, yes, need more carb cleaner and leave it soaking for a few days. A project for another day.

In the process of reinstalling the lines I damaged the little 5mm nipple that comes out of damper and connects to the accumulator. Ended up JBWelding it in until I find a replacement. Those suckers are not cheap. AutohouseAZ sells them for $152. So I may just look a used one up on Ebay.
 
#28 ·
I feel your pain (and will be feeling again). I've pulled the tank previously and had it cleaned, only to discover after re-assembly that the return line was restricted. Rather than fight trying to rod it out in-car, I'll be removing the tank and clearing it on the bench. This also gives me an opportunity to replace the fuel pump without having a gallon or two of old gas run down my arm.
Good times!
 
#30 ·
Sorry to hear about the broken accumulator nipple. Could be it was weakened from rust.

I'm surprised about all the rust I read about in tanks. I've heard to keep tanks full because moisture won't condense inside the tank (but then chance gum if fuel gets old). If tank supply/return lines rusted out enough to leak, that's a lot of oxidation - something caused it and maybe did same to sheet metal inside the tank.

I'd suspect your strainer as much as the swirl pot inlet. Strainers clog from the bottom up. Can you pull it to check it? Even if replaced (with the 14 mm fuel hose), rust could build up on it quickly, making tank drain slow when fuel level is low. But even with my strainer removed, only a few ounces drained when I poured 32 oz of cleaner into a dry tank - when I blew air into the return line, much more came out from venturi action (I also got a LOT of rust). You have advantage of seeing into the tank (does cleaner pool up in a "well" near the swirl pot inlet?).

Dry rust moves easily (unlike gum), as I see in dried drain pans. Let the tank dry out, and since you can access through the sending unit opening, can you get a tube hooked up to a vacuum cleaner (like 1/4" air hose line, maybe with a wire/coat hanger/small dowel rod attached to direct it) and pull rust out? Large flakes might clog the vacuum tube, but just keep pulling them out and cleaning tube end (or start with a larger diameter, or tap the clogged end in the tank to break up flakes). NOTE - maybe want to ensure fumes are not in tank - not sure if some vacuum motors spark.

Flakes could build up at the swirl pot inlet, but they'd likely break up easily (when dry). Blowing air into return line will tell you if the venturi is clogged. I couldn't see inside my tank, but if you run a wire (14 gauge solid copper electric wire should work) I'm curious if it will go through the hole in the venturi (depends on size of venturi hole), then through the inlet hole into swirl tank (depends on alignment of venturi and SP inlet) - once you get rust loosened, high pressure air (with gas tank cap removed) should blast away rust particles.

Just thought.. if you could fill up a line with solvent (like 6 ft of rubber hose, clamped to the UNCLOGGED return line going into tank), maybe by siphoning when gas/solvent in tank (I like paint thinner since it's less flammable), then hit it with 80-100 psi, you'd have a mini power washer shooting through the venturi at the SP inlet (go easy with air for starters).

Even with tanks removed, there's no way to easily get to that swirl pot inlet to clean it good (someone mentioned using a cable hooked to a drill like an auger). Not even sure how easy it would be if you cut the tank open like someone did.
 
#31 ·
Just documenting this in case it of some use to someone. Hope I don't need it myself!

Finally found an explanation on how the return line venturi is configured. According to leadguy, the swirlpot end of the return line just pokes into an opening in the side entrance of the swirl chamber(something like an annular venturi, mentioned previously) There is probably a reduction in diameter so as to form a jet. If so, size of that would be useful for those wanting to pass a wire through the return line.

It shouldn't be that difficult to feed a wire into the opening around the return "nozzle" from the swirl pot side using the level nozzle. But tank would probably best be taken out so any loose crud can be washed out. Otherwise hole will probably just get blocked again.

http://www.benzworld.org/forums/3588096-post22.html

If above is wrong, please feel free to correct.
 
#32 ·
Yup, those flakes is exactly what came out of my pump when I flushed it. Shame - the tank looks so shiny on the outside )

I already got the 46 mm socket and plan on removing the strainer, but now I am inclined to remove the entire tank and C-section it. No amount of carb cleaner will get rid of that rust.
 
#33 ·
Sorry I haven't commented on this thread. Sorta forgot about it.

MBGraham: I never actually checked to see if the venturi was sucking gas from the bottom of the tank (in addition to the return line). What I noticed with my simple set up was i pushed water through return line and it gushed out of the strainer hole. I was happy with that and excited as I spent many hours fiddling and trying to clean it. And to be honest, I didn't think I was making any progress. When I saw liquid pass through, it was Eureka! Lets get it dried up, primed/painted and installed. But, I would imagine that the pressure differential would encourage the gas to flow in from the gas tank itself; otherwise the car would of had cavitation problems and we haven't had any so far.

Dragonflyer: I don't know how much patience you have, but a vinegar bath will take out the rust pretty well. The tank on our 450sl was filthy, I posted some pictures of the rust that came out. We bought about 10 gallons of vinegar and let it sit for 4-5 days while slushing it around/flipping it over every so often. It took the rust right out and was pretty remarkable. The one thing that will happen after doing that is it will flash rust almost immediately. Since we planned on using a gas tank liner that wasn't too big of a deal for us.

However I will state this, if you do plan on using an epoxy liner, its important to ensure that any holes that the liner can make its way into should be blown out well.

Overall we went from almost concluding that we just had to drop a lot of money on a new gas tank to being pretty satisfied with the results from the mild acid bath.
 
#34 ·
Thanks hinero7, either way the tank has to come out. I've been inquiring and around LA (not to be confused with Louisiana) there are plenty of shops that will rebuilt the fuel tank for a few hundred bucks, which some say is too much, but to me is well worth the time I would have spent on it, not knowing what I was doing.
 
#35 ·
FWIW...I chased the fuel system dragon for a while with my long stored '79. Had the tank dipped..pressure cleaned..solvent, etc...Every time would end up with the filter at the FD being plugged...I'm familiar with the coatings that are applied to the inside, but my indie..and a couple other hot rod buddies said to just get a CLEAN tank and be done with it..I bought a used tank ..put everything back together...and problem was solved....This was after Larry at CIS rebuilt the FD and the WUR, new strainer, filter and pump...One of those problems that is not worth the worry every time you take the car out...
 
#36 · (Edited)
Interesting to read since my car has been down with what I assume is a rusty gas tank since late last year. I removed the fuel filter since it was acting fuel starved and it was clean on the engine side... looked like coffee on the other. I didn't bother with the strainer in the tank. The car would run just fine for a few minutes then stall out and not restart until a while later when I assume whatever rust was blocking a filter/line/screen settled out.

Is it safe to assume rust is the cause? If so I'd be more inclined to find a place that can rebuild the tank than try and DIY, then repeat it again once the repair fails. Wish I could just buy a reproduction and be done with it.
 
#37 ·
ozark1 - As my nick would imply, I deal with aircraft. A lot. there is this stuff made by PRC - the fuel tank sealer. We just call it the PRC. It it used to seal the integral wing fuel tanks in all general aviation aircraft. It is good for 30-50 years, or more. If properly prepped, the steel MB fuel tank can be coated with this stuff long enough for my grandkids to learn to drive. Looking into it now.
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top