1973 450SL starts on Cold Valve Squirt then stalls - PeachParts Mercedes-Benz Forum

mespe · #1 02-03-2012, 02:46 PM

Just picked up a 1973 450SL and I would like to get her running for a summer driver. She'll start, run a few seconds and stall. Throttle does nothing.
so my first guess is she's running on the squirt from the cold start valve.
What to check from here?

RadioTek · #2 02-03-2012, 03:40 PM

You have a Bosch D-Jetronic ignition / injection system. they are a bit fiddly to work on, they are very sensistive to vacuum leaks, and the parts can be a bit expensive. The good news is that they are not extremely complicated, and some "Quality Time" with available manuals can lead you to a solution. Whether that solution is a fix for the problem, or alcohol mixed with various flavoring agents can be a sticking point.

There is a set of breaker points beneath the distributor cap that needs to be properly adjusted, a piece of automotive equipment of near legendary status.

The place to start is with the usual tune-up suspects. Plugs, Points, timing, dwell, etc...

Do you have fuel to the injectors in sufficient quantity?

Do you have spark and is it occurring when it is supposed to?

Here is a link to an online copy of the service manual: Model 107

Here is a link to another forum that is pretty active that may be useful. Read the Sticky at the top titled "This Damned old car". http://www.benzworld.org/forums/r-c107-sl-slc-class/

Have fun, these are some fun cars to drive, they look classy, and you can actually work on them without a million dollar piece of computer equipment to open the hood.

Welcome to the madness.
Scott

mespe · #3 02-03-2012, 04:28 PM

here's a video of what she's doing,,,

1973 450SL as purchased - YouTube

mespe · #4 02-03-2012, 05:36 PM

Anyone know if the ECU case is supposed to be grounded? my SL has no passenger floor, and very little firewall. When I went to inspect the jetronic, it literally fell off the firewall.

mespe · #5 02-03-2012, 06:31 PM

I found a site that explains the jetronic system and helps troubleshoot it,,, here's the link

914 D-Jetronic Parts and Troubleshooting

It's for a porsche, but I tell you what, the 1973 450SL has a cylinderhead temperature sensor, that I think was part of the problem. It was dry corroded, a splash of PBBlaster and the car starts and idles for more than a few seconds now. maybe 10-20 seconds, so I think I'm making progress. 2/4/12- THIS TURNED OUT TO BE COINCIDENCE

RadioTek · #6 02-03-2012, 07:36 PM

The D-Jet system relies very heavily on 2 sensors: The IAt (Incoming Air Temperature - sensor 1) and CLT (Coolant Temp - Sensor 2) The most important one being sensor 2 (go figure). If you have corrosion issues with those, you might have a starting problem. I remember having left the IAT sensor (the one in the air cleaner intake) off and getting a 'no start' problem. The wiring and connectors should be checked out carefully. At almost 40 years old, there may be some corrosion issues at the connectors.

Good Luck.
Scott

rowdie · #7 02-03-2012, 08:46 PM

The ECU is not grounded through the chassis.

79Mercy · #8 02-03-2012, 10:50 PM

I had a customers car do this on me before, it was a '73 450SE to be exact. Turned out to be the trigger points for the EFI system.

Unplug the wire harness on the top of the motor up front, try to start the car and see if it acts the same way, if so i personally think its the trigger points.

mespe · #9 02-04-2012, 11:10 AM

I pulled one of the spark plugs and it was wet. Also looked into the distributor is almost totally disintigrated on the shaft side.

Another observation (or two) There are two vacuum ports on the throat, one above the butterfly and one below. each of these lines go to a vacuum solenoid on the passenger fender wall. i.e. the vacuum port above the butterfly goes to the solenoid on one port, and the other port goes to the vaccum port below the butterfly, this doesn't seem quite right.

Another observation is that distributor's vacuum advance is capped off, almost looks factory. Did Mercedes not use vacuum advance on some early 450SL's?

Or, are these things done to increase horsepower or was this done in an effort to figure out why the car won't start, by a previous owner/mechanic?

mespe · #10 02-04-2012, 11:14 AM

On a side note, although the body is rotted, the engine compartment is very clean for such an old car. Things like trigger pickup connector on the distributor, the little plastic tabs that hold the connector to the distributor are in place, and I really didn't notice any broken connectors or corroded connections.

I think the car was stored outdoors for a period of time and water built up on the floor board and rotted it away.

The odometer shows 020596 and the interior is really nice (minus the tear in the bolster of the driver's seat.)

mespe · #11 02-04-2012, 01:41 PM

After pulling a plug and seeing it wet, I started thinking, maybe not a good spark. So I looked at the points, yuck, they were blue, Once again PB Blaster and the paper stripe covering the glue from a USPS Priority mail box, (like waxed) and points clean. Then I checked the dwell, and it's between 30 and 34. I've got a spark, maybe not as good as I should, I do not know.

checked the trigger points, there was a little blue on the wires inside, so, PB Blaster once again. and blasted the connector, eventhough it looked OK, with no corrosion.

meltedpanda · #12 02-04-2012, 02:22 PM

Welcome to the world of DJET , seach my name in the SL section and you will get all you want to know. I suspect your spark is missing or weak, but there are a host of thisgs to check
I am sure Graham will check in soon, Rowdie is already here
in the meantime read this, and I have a host of other files if you PM me I can get some to you if needed, specific to many items on the Djet

System Overview:
D-Jet is a "multi-port," electronically-controlled system that permits greater precision in metering an optimum fuel / air mixture than do carburetors, resulting in generally increased power and fuel economy. In a sense, it's a lot like spreading the internal bits of a carburetor around the engine compartment. The air component of the mixture is drawn into the motor just the same way as in a carbureted motor: through a throttle body on the outboard end of the intake manifold. The throttle body contains a single large butterfly opened by pressing the gas pedal (which should really be called the "air pedal"). Rather than using the resulting airflow to suck gas into the mixture through metering jets in a carburetor, however, D-Jet injects an atomized spray of fuel into the airflow directly outboard of each intake valve. How much fuel is determined by a simple computer known as the Electronic Control Unit, or ECU,.

Fuel is supplied to the injectors at a constant, regulated pressure. Each injector contains a small solenoid which snaps the flow fully on or shuts it fully off. The ECU's job is to determine how much fuel is needed and then control how long the injectors are snapped on during each piston stroke (even at full throttle, the injectors are on only a small fraction of the time). The ECU will vary this period in accordance with the following inputs:

Throttle position and movement from the Throttle Switch, which is just outside the throttle body and is connected to the butterfly shaft.
Intake manifold pressure from the Pressure Sensor, which is the hand grenade-looking thing on the fender connected to the intake manifold by a rubber hose.
Coolant temperature from Temp Sensor #2, which is a small fitting screwed into the head on the right just below the thermostat housing.
Ambient air temperature from Temp Sensor #1, generally located in the air cleaner box.
An additional input is required so that the ECU will synchronize the fuel injectors with the piston strokes. It therefore receives signals from a set of dual Trigger Points in the base of the ignition distributor. Each of the two points controls two injectors (in our four-cylinder motors, anyway).
Two or three more components (depending on year model) are used to aid cold starting, acting much like an automatic choke would on a carburetor. These are not controlled by the ECU, nor do they provide signals to it. They are the Auxiliary Air Valve (the silver metal cannister below Temp Sensor #2), the Cold Start Injector (centrally located on the intake manifold), and, in later cars, its enabling element, the Thermo Time Switch.

Fuel Pressure:
An electric Fuel Pump and filter are located at the rear of the car, immediately outside the gas tank. The pump is powered through a relay, operates for a few seconds whenever the ignition is switched on, and continuously when the motor is cranking or running. It provides fuel at several times the required pressure to the Fuel Pressure Regulator. No (or inadequate) pressure to the regulator input may be caused by failure of the pump itself, the associated fuse, relay and wiring, or simply a plugged-up filter. Fuel additives like Methanol or octane boosters are reputed to be injurious to the pump.

The Fuel Pressure Regulator, on top of the fuel "rail" supplying the injectors, maintains the injector supply at a nominal 30 PSI (adjustable by turning the nut on top). It does this by returning "excess" fuel to the tank -- a partial obstruction in the return line will result in pressure to the injectors rising over the first several minutes of operation (one possible cause of good initial starting followed by increasing excessive richness). For modified, high-performance motors requiring more fuel, adjusting the regulator for higher fuel pressure is the principle way to get it.

Fuel pressure is readily measured by inserting a gauge between the fuel rail and the Cold Start Injector.

NOTE: Due to the pressures involved, all fuel hoses must be rated for fuel injection. Also, use only FI-type clamps -- regular hose clamps will cut into the hose material and cause potentially dangerous leakage in time.

Injectors and Trigger Points:
Just as in a carbureted motor, a little study of spark plug color and condition can tell a lot. If the plugs indicate fuel-related problems in pairs, the Trigger Points are suspect. These are easily removed from the distributor as a unit for inspection or replacement.

If only one plug indicates a fuel problem, the associated injector may be at fault. First, please see our article on D-Jet Injector Seals in the Archive.

If all the plugs show a problem, it's not likely that all four injectors are going bad at once. Carefully check the wiring, including the grounds grouped together at the rear of the intake manifold. Check for problems in the fuse box -- the ECU senses analog voltages; if there are voltage supply problems to begin with, generalized problems will be the result.

It is possible to check injectors for spray pattern and flow rate, but I won't encourage you to do it at home. If troubleshooting the rest of the system leaves you with buggy spraying as the only remaining possible failure, have the injectors checked by a professional.

Throttle Switch:
As our D-Jet cars approach thirty years of age, the Throttle Switch is a common source of trouble. Typical symptoms include hesitant acceleration and "bucking" at sustained speeds. Fortunately, it's very simple and can generally be repaired with careful cleaning of the internal contacts.

Under the cover is a circuit board with 22 segments, across which three wipers are swept as the throttle butterfly shaft moves. One segment and one wiper make contact and signal the ECU when the throttle is closed. Another segment and wiper make contact to signal wide open throttle. The middle wiper sweeps the remaining 20 segments as the throttle is opened. Each time it "climbs" a segment, it signals the ECU to fire all injectors one additional time. This is, essentially, what D-Jet uses in place of a carburetor's accelerator pump.

Check the adjustment of the switch on its shaft and the condition of its contacts by turning on the ignition (don't start the motor) and manually operating the throttle through its full range. You should be able to count exactly 20 clicks from the injectors as you slowly open the throttle.

Pressure Sensor:
Vacuum in the intake manifold (that is, pressure lower than that of the outside air) varies greatly (for any given engine RPM) with variations in throttle opening and with load. The Pressure Sensor compares relative air pressure inside and outside the manifold and provides an analogous voltage to the ECU. The lower the vacuum, the more the mixture is richened.

It's worth noting here that vacuum leaks, depending just where they are, can cause a D-Jet to run either lean or rich. This is not true for most other FI systems and not true for any carbureted system I know of, so even experienced mechanics are sometimes sorely baffled by this phenomenon.

The simplest mechanical check of a the Pressure Sensor is to unhook its connecting hose from the manifold and suck on it. No air should leak through the diaphragm in the sensor (of course, there are more graceful ways of applying vacuum, if you're both hesitant to do that and want to spend more money).

Temp Sensors #2 and #1:
Bosch numbered these two sensors in reverse order -- #2 does most of the work, while #1 is fine tuning. #2 causes the ECU to richen the mixture when the engine coolant is cold, progressively returning to normal as full operating temperature is approached. For this reason, operating a D-Jet car without a thermostat (or with the wrong one) has a large affect on both performance and economy.

Temp Sensor #1 causes the mixture to richen when the intake air is cold. A failure of this sensor or its wiring will affect performance and economy, but not major problems.

Cold Starting:
If you were starting a cold carbureted motor, you'd probably pump the gas pedal a few times while cranking, thereby making the accelerator pump squirt a little raw gas down the intake. This is completely useless if your Volvo has SU carbs (which don't have accelerator pumps), but I'll bet you do it anyway; I know I sometimes do. The D-Jet system automatically does this pumping for you.

This is the province of the Cold Start Injector (aka "the fifth injector" or Cold Start Valve). It receives power from the starter solenoid only when the motor is being cranked over. In earlier models, the power is routed through a third FI relay activated by the ECU. In actual practice, the Cold Start Injector simply fires for a few seconds whenever the motor is cranked, regardless of temperature.

In later cars, the third relay is absent and power is supplied directly from the solenoid. The circuit is completed by a ground routed through the Thermo Time Switch, which senses coolant temperature. If the coolant is hotter than 95 degrees F., the Thermo Time Switch opens and no ground is supplied. If you've ever flooded a warm motor by pumping the gas out of turn, this should make perfect sense to you.

If hard cold starting is the main trouble, check that the Cold Start Injector is receiving power while cranking. Check that the other pin on its connector is grounded (on later models, only when the coolant is cold). If that doesn't reveal the trouble, remove the injector, hold it tip-down in a glass jar and see if it sprays when a helper cranks the motor.

The fifth injector does have full regulated fuel pressure applied to it, warm or cold, and fuel leakage through it can cause rich running and poor fuel economy, although not usually enough to cause serious driveability problems. As long as you have the injector out and pointing into a jar, turn on the ignition and don't crank the motor. No fuel should drip.

A cold motor needs to idle slightly fast until the coolant and oil temperatures come up, requiring a little additional air and fuel. On a carbureted motor, this is usually accomplished by holding the throttle slightly open on a choke-operated cam. The D-Jet uses an Auxiliary Air Valve instead, which simply allows a small amount of air to bypass the throttle butterfly when cold, gradually reducing the amount as the coolant warms up. To the rest of the system, this is identical to holding the throttle slightly open; the ECU automatically supplies the needed matching fuel as dictated by pressure and temp senders.

There's nothing electrical about the Aux Air Valve on a Volvo D-Jet (unlike on air-cooled Porsches). Test it by unhooking the connecting hose from the intake manifold and drawing air through it. With a cold motor, air should flow through readily; when warm, it should be much harder (it's OK if you can get a little bit through).

Electrical Checks:
There are two choices here. If you suspect a particular component of electrical failure, you can verify that failure by disconnecting the wiring and measuring directly on the component. On the other hand (this is more effort), you can disconnect the ECU and measure through the entire harness from that end, thereby verifying both component and wiring all in one shot. Decaying wiring is a common D-Jet problem, so we recommend the second method.

Always begin by checking all connections (including grounds) and by verifying that there are no fuse problems. All resistance checks must be conducted with the ignition off.

Injectors:
Individually check ECU harness pins 3, 4, 5 and 6 to ground. Resistance should be less than 25 ohms, but not zero. NOTE: The ECU powers the injectors with a 3V signal lasting between 2 and 5 milliseconds. This signal cannot be measured with a meter; don't try.
Trigger Points:
ECU harness pin 12 to 21, then pin 12 to 22 = near 0 ohm continuity on one set, open on the other. Rotate engine 360 degrees and test again. Readings should be opposite.
Throttle Switch:
First check for 20 clicks as outlined above. If no clicks are heard, check wiring continuity between Throttle Switch connector and ECU harness pins 9, 14, 17 and 20 (one connector wire to each harness pin only, of course).
Pressure Sensor:
ECU harness pin 7 to 15 = 90 ohms.
ECU harness pin 8 to 10 = 350 ohms.
Temp Sensor #2:
ECU harness pin 23 to ground as follows...
Cold motor = roughly 2000 ohms.
Warm motor = well under 1000 ohms.
Temp Sensor #1:
ECU harness pin 1 to 13. At air temp of 85 degrees, F., about 200 ohms; resistance varies inversely with temperature.
If you're sure you're getting an abnormal reading, you've found the bad part (and/or wiring). If you get a normal reading, the wiring is definitely good and the component being checked is probably good. There are failures that won't show up by doing thse checks, particularly involving the Pressure Sensor -- but it's a lot less expensive to try doing them than just throwing new parts at the problem until it goes away.
In Conclusion:
Don't be intimidated by the D-Jet system. If problems arise, first rule out anything that would cause a similar problem in a carbureted car -- Electronic Fuel Injection is just a different (I would say better) way of getting your engine the air and fuel it needs to run, not a cure-all, and not the source of all evil, either.

Common sense will go a long way towards helping you troubleshoot EFI problems efficiently. If nothing fuel-related seems to be getting power, check fuses and relays (a manual with a wiring diagram specific to your model is more than helpful in this). If the car is tough to start but runs well when warm, look into the components that affect Cold Starting. Note that if a wire falls off a temp sensor, the ECU will go towards rich; high resistance equals cold temps, so an open circuit must mean you're in Siberia at the height of winter -- the ECU's trying to help you out.

Obviously, I can't cover every problem or symptom you might experience with your D-Jet car, but I hope I've given you a basic understanding of just what makes the thing squirt. If you understand this much, you know more than most pros to whom you might entrust your car (unless they're specialists in early Bosch EFI; the later systems work quite differently). Give troubleshooting it yourself a shot -- you'll get a lot of satisfaction from figuring out what's wrong, to say nothing of potentially saving quite a bit of money.

mespe · #13 02-04-2012, 04:18 PM

Thanks, meltedpanda, there's alot of good info in that post.
I checked fuel pressure at the cold start valve. it measured around 30psi after toggling the ignition switch 3 times. after about 15 minutes the pressure still held. I then tried starting the car without the cold start valve in place because the psi meter was hooked up, and not even a "want to try to start'. I depressed the accelerator and not much change (if any).

At this point I don't feel I need to check flow rate, as the car will not even idle for more than a few seconds. Maybe 10 seconds at most, and she dies.

On a side note, the tach is not functioning.

mespe · #14 02-04-2012, 07:28 PM

Quote: