D-Jet leak down rate - PeachParts Mercedes-Benz Forum

Graham · #1 08-25-2009, 08:35 PM

CAR: 1972 350SL 4.5L (D-Jetronic) Chassis 2572

I put in a dash mounted fuel pressure gauge some time back. I have been using it lately to check the leak down rate of the fuel system. Reason - Trying to determine reason for rough running after hot start when car has heat-soaked for 15-45 min.

What I am finding, is that leak down is minimal for first 10 min or so. If I come back to car after say 1/2hr, it might be at 10-20psig. After an hour or more, it is almost at zero.

Clamped off the cold start injector - still leaked down
Clamped off fuel return line and cold start injector - still leaked down

This leaves leakage though injectors or back through feed line and pump.

On my car, there is no rubber hose on fuel feed line - except perhaps under car somewhere. I may have to find it!

Is it likely that injectors would leak down in 1hr? Or is most likely the check valve at the pump?

I read somewhere about a modification on early SL fuel pumps where a check valve was added. Anyone recall anything about that?

Tomguy · #2 08-25-2009, 09:29 PM

Warm leakdown is most likely the injectors. To be sure, you can use fuel hose about 3' long to bypass the fuel rails on each side. You can get about 6' to bypass them all, let the system pressure up (key on, off, on, off, on until you hit ~30PSI) and see how much it leaks down.

Graham · #3 08-25-2009, 11:38 PM

Quote:

Originally Posted by Tomguy

Warm leakdown is most likely the injectors. To be sure, you can use fuel hose about 3' long to bypass the fuel rails on each side. You can get about 6' to bypass them all, let the system pressure up (key on, off, on, off, on until you hit ~30PSI) and see how much it leaks down.

I guess you mean I should bypass the injectors and connect back to the fuel pressure regulator? That way I should still get the 30 psig regulated pressure and then see how long it takes to drop. Makes sense.

As an alternative, I have a spare set of injectors ready to go in, so maybe I just test them for leakage before installing, then if I still have excessive leakdown, it must be at the pump.

Reason for this, is that I sometimes have the well known hot start rough running symptoms on hot days. I figure if I can keep the pressure above the vaporization point of summer/winter gas while engine cools, that problem will go away. Need pressure to stay above about 25 psig for an hour or more.

I see that a check valve is available for the early SLs - That's what I am not sure about - I hear there was a mod to early SLs that concerned replacing one pump nozzle with a check valve, but I don't have the details.

meltedpanda · #4 08-26-2009, 10:02 AM

my money is on the injectors, I cured a ton of ills on my 72 with refurbished/new injectors, including hot start issue, rough idle, ect

230/8 · #5 08-26-2009, 10:31 AM

Book says the leak down rate is a drop of +/- 7 to 8 psi in five minutes (from the correct 29-30psi) indicating an acceptable rate of leak down is designed into the system. If it falls to zero in that five minutes you should look for a leak. Sounds like you have a system that may be within specification, but why not simply check it at the five minute mark to verify the values at the specified times for the test? This will rule out or rule in the existence of a problem. Are you having some sort of problem with the car or is your question merely to satisfy your curiosity?

Graham · #6 08-26-2009, 11:20 AM

Quote:

Originally Posted by 230/8

Book says the leak down rate is a drop of +/- 7 to 8 psi in five minutes (from the correct 29-30psi) indicating an acceptable rate of leak down is designed into the system. If it falls to zero in that five minutes you should look for a leak. Sounds like you have a system that may be within specification, but why not simply check it at the five minute mark to verify the values at the specified times for the test? This will rule out or rule in the existence of a problem. Are you having some sort of problem with the car or is your question merely to satisfy your curiosity?

My manual says that that when engine is turned off, pressure can drop to 1.7 (~25psig) bar. After another 5 minutes it says it can drop to 1.5 bar (~22psig). For some Bosch pumps, it is OK to drop to 1.2 bar (17.4psig) Doesn't go beyond that or mention how long that pressure should be retained.

I check the fuel pressure each time I drive - I think that the leak down is within the above specs. But even early on, MB were coming out with fixes for the hot-start problem (hot start relay in '74, fuel line cooler a bit later, leak down orifice around pressure regulator, etc), so maybe the specs are not perfect!

Reason for my interest in this, is that I (and many others!) have a minor hot start issue - It was much worse before I boosted the fuel pressure to about 31-32psig (when car is under way).

Under hot-start/heat soaked conditions, car does now start but runs rough for a short while. My idea is to try and retain fuel pressure in rail above vapor pressure of gasoline for longer period so as to prevent bubble formation. This is of course temperature dependent and may be hard to do if fuel gets much above 100C.

Car runs and starts fine unless it gets really hot (that's above about 28C here in Canada

). Problem only occurs if restart is between say 15 and 45 minutes after a hot stop/start run.

Graham · #7 08-26-2009, 02:43 PM

Quote:

Originally Posted by meltedpanda

my money is on the injectors, I cured a ton of ills on my 72 with refurbished/new injectors, including hot start issue, rough idle, ect

Melted and TomGuy - You were right!

Just finished testing:

- clamped off fuel feed to injector banks which luckily still kept the
pressure transducer "live".

System held pressure for well over an hour. Pump check valve is
doing it's job!

Removed clamps - Pressure went down to ~5psig in 5 minutes -
Seeing everything else had held pressure, it must have been injectors
leaking.

Tested spare set of injectors on bench - applied ~20 psig pressure using
brake bleeder. Each injector held full pressure for at least 5+ min - no sign of any liquid at nozzle.

Time to install the injectors! Then recheck the leakdown again.

Thanks for the input!

Tomguy · #8 08-26-2009, 10:14 PM

Glad to help! The issue isn't that there is no pressure when you go to start, the pump repressurizes before the engine cranks. It's the fact that there's too much gas in your engine for a warm engine to re-fire easily!

Graham · #9 08-26-2009, 10:40 PM

Quote:

Originally Posted by Tomguy

Glad to help! The issue isn't that there is no pressure when you go to start, the pump repressurizes before the engine cranks. It's the fact that there's too much gas in your engine for a warm engine to re-fire easily!

If the replacement injectors do the job, I hope to be able to SMELL the difference

Cheers

meltedpanda · #10 08-27-2009, 11:10 AM

MPG increase (perhaps) as well!

GGR · #11 08-27-2009, 02:35 PM

Hi all,

I have a Djet 3.5 engine in my W111 Coupe and have been facing similar problems. Hot start problems have been reduced the day when I bypassed the fuelpump relay that was giving me trouble. By letting the pump run a few seconds more than the normal 2-3 seconds, the engine is now starting more easily. But it has now started developing a cold start problem. Even by letting the pump run up to 10 seconds before cranking, the engine takes time to start and when it does runs roughly with some signs of backfiring which is a sign of leaness. Things then smoothen out and the engine runns really well. I did not note hot starts being more difficult though... Any idea of what I should check?

Thanks!

meltedpanda · #12 08-27-2009, 04:28 PM

well first check the injector that is not above the cylinders - the CSV

, cold start injector
If that is OK make sure you do not have any vacuum leaks

GGR · #13 08-27-2009, 04:51 PM

How to I check the CSV? I take it out, I disconnect the wires, leave it plugged on the fuel line, put on the contact and see if it drops?

Or may be it does not do its job, so on the contrary I should leave everything plugged in, take it out with a cold engine and see if it squirts when I put the contact? Or should I crank the engine (with the high tension lead disconnected from the coil) in case the signal is somehow sent to the "brain" by the starter motor?

Thanks, and sorry for hijacking the post, but I feel it's related.

meltedpanda · #14 08-27-2009, 09:25 PM

sorry CSI not CSV
the CSI recieves battery voltage when the starter is engaged and is grounded through the thermo time switch, when the engine temp is less than 95 F.,the CSI is grounded and fires
It will only work below 95 F and when the engine is cranked - maximum 5-12 seconds
Not operating - cold start issue, leaking , cold start issue with causes rich start ( black smoke)
So you really have two items to check, others can chime in but you can on a cold engine carefully allow the circuit to work while disconected from the fuel rails. This only tests the whole system , you must rule out the thermal switch, and to do that you must idependently gound the CSI
Obviously you are delaing with spraying gas, so be very careful

there are bench tests but I have never used them

Tomguy · #15 08-27-2009, 09:26 PM

Test the CSV: Connect a FI pressure gauge to the system. Unplug the 8 primary injectors. Pressurize the system by putting the key to run (once should build ~30PSI right away, if not, toggle the key until you get 31 PSI). Disconnect a terminal from the EFI pump. Crank the car.

If your pressure doesn't drop the CSV is not turning on at all, or is 100% blocked. Pressure should drop like a rock in this test, not slowly.