Starting Issue, 1974 280 SE

Dear All,

I am sorry, I do not speak German.

My name is Carl from Melbourne, Australia. I have a 1974 280 SE that has been causing a LOT of problems and it has come to to the stage where I do not know what to do next!

The situation is that the car is very difficult to start from cold. When hot, and left for no more than 30 minutes, it starts first crank. After 30 minutes, it's as if it's cold again and is difficult to start. Approximately 1 in 100 starts, the car will start first time from cold and run well.

To start the car I have to crank it for maybe 30 seconds. I am scared to crank for longer than this! It will then fire for a few seconds then die. Cranking again, it will require 4-6 attempts before it fires, then settles down to a 1,200 rpm cold idle. The warm-up phase is uneventful and when up to operating temperature, the engines performs well. When hot, the only slight issue I can detect is if I "Snap" the throttle wide open from rest, there is a slight hesitation, a more moderate application of throttle causes no issue.

The car has 113,067kms on it, and is in excellent condition.

The initial jobs I performed on the car is the following:

Check and adjust valve clearances.
Oil and filter change
Replaced the following:
Spark plugs
Ignition leads
Distributor cap
Rotor
Ignition points
D-Jetronic trigger points.
Air filter
Fuel filter
Fuel hoses at tank end
Fuel hoses in engine bay
Ignition amplifier (With working second hand unit from a car operating well)
Ignition coil.

After the above I then set the ignition timing, mixture and idle speed and the problem remained.

I then performed the following:

1. Ignition test. Confirmed plugs were firing during cranking.
2. Fuel pump test. Tests revealed 2.1Bar pressure when running (And during cranking). Residual pressure test revealed a drop to 1.0Bar after 48 hours. When ignition switched on after 47 hours, system primed and pressure returned to 2.1Bar. Fuel flow test revealed 1.6L / minute.
3. Injector test. Removal of injectors confirmed injectors firing during cranking. Cold start injector firing. Removal of individual injector wiring revealed the same drop in RPM per cylinder.
4. Trigger points. Removed distributor and checked for trigger signal at distributor, was correct.
5. Checked ignition points for gap again and replaced distributor and set timing. Checked ignition wiring from ignition amplifier.
6. Checked with multimeter:
-Throttle position sensor. Although it checked out fine, I had a new one, so I replaced it. Set correct throttle position sensor position.
-Checked readings for temperature sensor I and II. (Both within specification).
-Checked for continuity the wiring for above.
-Checked manifold air pressure sensor and wiring.
-Checked for continuity between the ECU wiring harness and the engine bay, removed all connections, cleaned and re-fitted.
7. Checked for vacuum leaks around inlet manifold, OK.
8. Checked auxiliary air valve and confirmed correct operation.
9. Checked battery during cranking and confirmed battery to be slightly under-voltage, replaced battery.
10. Replaced fuel pump and system relays in engine bay.
11. Experimented with ignition timing, from 5 degrees BTDC to 5 degrees ATDC.
12. Shorted out ballast resistors and tested starting engine.
13. Bypass fuel pump relay and run pump for two minutes before attempting starting.

After the above work and tests, there was no change.

Then I decided to remove again the ECU. I opened the ECU and re-soldered every solder joint. Looked for obvious signs of damage from heat, or a faulty electrolytic capacitor. Could not fault. Exchanged ECU with a friend, this ECU made no change in my car.

I then bought a Kent Moore 112D D-Jetronic tester, although this is not really doing anything a multimeter can perform, it does emulate the the ECU so this provided a third ECU to test along with the ability to check injection items quickly. This unit did not up any faults.

Experimented with idle mixture on start up, changed AFR from 10:1 to 15:1.

After the above, no change.

At this stage, my opinion is that the issue is clearly fuel related. The reason is because the car has spark, I've tested it several times, and when cranking, you can hear the odd cylinder firing. When the engine finally starts, it does not run on all 6 cylinders, it take a few seconds to run on all 6.

What I find most confusing, is that the engine runs really nicely when hot (Except for the slight miss if you "Snap" the throttle open.

The next stage for me is unfortunately to try "Shotgun" engineering. That's where you just keep fitting new parts until it's fixed. This is not my preferred method of repair, but I am at the end of the road.

I will start by replacing the fuel damper. After then I will replace the fuel pump. If that doesn't fix it, I will have no option but to replace the entire injection and fuel system until it runs properly.

Unless, does anyone have any ideas? Thank you for any suggestions.

Hi Carl,

welcome from down under! Of course English is no problem and you are welcome to write in that language. This seems to be the months of failing MB M110s. I am impressed how much you have done so far. I have read your history of pain (and maybe should do again carefully) and understand that you only have cold start problems and that you do have a strong spark while it will not start. I also read that you have thoroughly tested everything on your car. On the fly 4 things come to my mind:

1. How is fuel pressure on cold car when it does not start? Also 2.1 Bar?
2. Did you do my spray test by letting all 6 injectors spray into glasses with cold engine?
3. Story reminds me of a 350SLC. When really cold it would start but then run maximum 40 km/h. Warm it would go well till 160 km/h and then start to stutter a little bit. In the end it was the rpm based ignition timing adjustment in ignition distributor. Old fat had made it stick when cold. I could only see that when I hooked the car to my igntion tester after a cold night in my workhop.
4. What could also cause this is a leaking injector or cold start valve. It would more or less drown cylinder in fuel and it would take long starting to get rid of it. Should be more than 1 injector.

This is a tricky one! Cold start in Australia is probably not that cold!

Question - Do you have the starting problem when you first start car in morning? The subsequent starting problems after 30 min sound like the common hot start problems. They are caused by fuel that gets hot after sitting for 20-30 min under hot hood (bonnet) and then flashes to mixture of vapour and liquid through injectors causing overly lean mixture.

One simple test to add to your data bank would be to blow through the fuel return line to make sure that fuel is free to return to the tank from the fuel pressure regulator/damper. If that line was blocked, pump might overheat fuel because it has nowhere to go at low engine speeds.

Is the pump an original D-jet type pump? The flow seems a little low. Should be ~2L/min.

Volker - I too had the centrifugal weights in my distributor seize up. Since then, I always check timing at 3000rpm (which for my NA V-8 is about 27 BTDC.) The 6 cylinder would of course be different.

Hi Graham,

fuel flow from pump is correct with 1.5 l / min. Original spec for 0 580 464 005 is 750cm³ in 30 seconds. Anyhow if fuel flow is too low you would see too little fuel pressure when car does not start. That is why I asked the fuel pressure on cold car not willing to start.

And as you have also experienced: Rpm and vaccum advance of ignition distributor can hang depending on temperatures. That is why I always pray that you have to not only adjust ignition timing but you also have to verify rpm and vacuum advance as MB / Bosch describe it in their workshop manuals. With that 350SLC from 73 I got crazy. When I drve to my workshop to test with Bosch Compactest ignition tester it seemed to be okay. But when the car stood over night nothing worked. Finally I left it in my workshop and then saw the disaster on Compactest when cold starting the engine 2 days later.

Hi Carl,
a Warm welcome from me, too.
As Volker mentioned before, you did almost everything one could do to solve the problems.

This is really a great task for us, too...
Your engine looks as being new, not as rotten as many other car-owners have problems with.

Now, some more ideas from me:
- What about the engine / body grounds? I once owned a W115 that had thick, round wires soldered in at the connectors screwed to the body. This soldered connections had broken and failed from time to time.
- Did you replace any connectors of the wiring harness by new ones? There are connectors being sold with wrong dimensions that don't give reliable contact. Look at the compendium, wiring harness.
- Maybe the injection at your car doesn't get the starting voltage directly from the ignition lock or the starting relais, but from an extra contact at the starter. At Opel-cars (I know better) this function failed sometimes because of contacting problems inside the magnetic switch at the starter (Opel has Bosch-Starters installed, too...)

What you could do more to identify the problem?
- replace the temp sensors by fixed value resistors, both sensors about 300 Ohm would signal 'normal condition' to the ECU. Simple 1/4 Watt-resistors will do.
- when installing LEDs with dropping resistors directly at the injectors you can control their firing under normal operating condition. A resistor with 100 Ohm did the job for me, using a white LED. You must try out the direction of the LEDs- seems once wasn't cared about the direction of these connectors when mounting the harnesses.

My ideas, so far... I will send you a PM regarding a favor you could do to me...

Regards
Norbert

Volker schrieb: Hi Graham,

fuel flow from pump is correct with 1.5 l / min. Original spec for 0 580 464 005 is 750cm³ in 30 seconds. Anyhow if fuel flow is too low you would see too little fuel pressure when car does not start. That is why I asked the fuel pressure on cold car not willing to start.

Volker, you are correct that 1.5l/min is the RATED capacity at 12V and 2bar using n-heptane. Bosch say to use their numbers just as a general guide. In practice with gasoline and higher actual voltage, Djet pumps including my own have higher capacity. So that is why I said 2L/min. Back when Djets first came out, Bosch ran training programs for mechanics and in those the troubleshooting section said AT LEAST 1L/30sec. dl.dropboxusercontent.com/u/54783344/efi330.JPG

Fuel pressure is set by the fuel pressure regulator, not the pump capacity. Even if pump only put out 1L/min, pressure would be the same so long as fuel was being recycled via FPR.

Of course, during starting the engine only needs a small amount of fuel. Most is returned to tank. Low flow would not be a factor. But it could be an indication of pump condition. For example, a malfunctioning internal relief valve that is built into Djet pumps could be spilling fuel back to the suction. Unless car has a full time pressure read out, changes in fuel pressure may not be noticed.

By the way, we have found that the Nissan 280Z pump is essentially the same pump, but was made in Japan. Still available new on eBay at times: www.thezstore.com/page/TZS/PROD/11-3063 . Less expensive and direct fit compared with new Bosch replacement Djet pump.

By the way Carl. Despite this discussion of fuel pumps, I suspect Norbert is more on track that you may have an electrical problem in either the injection or ignition system. If it was my car, I would be tempted to put in a Pertronix points replacement and even perhaps bypass the ignition module (switchgear). They are pretty cheap. And then do as Volker suggested - pull fuel rails and watch injectors discharge into jars during a morning cold start.

Hi Carl,

I recommend you install a fuel hose into fuel line above injectors via T-distributor, Connect it to a manometer and let it hang out onto the windshield Then you can fix it there with tape and watch your fuel pressure during start and ride.

If it does never deviate, then there is no problem with your pump.

MBGraham schrieb: By the way, we have found that the Nissan 280Z pump is essentially the same pump, but was made in Japan. Still available new on eBay at times: www.thezstore.com/page/TZS/PROD/11-3063 . Less expensive and direct fit compared with new Bosch replacement Djet pump.

Graham, that is an interesting hint. I am not fond of the Bosch 0 580 464 999 as it is for 4 Bar and has a maximum pressure of 8 bar. This Nissan pump looks like the Bosch pump. Do you have flow and pressure data for it?

Regarding Pertronix: In my eyes it makes no sense that Carl changes more things than necessary. If he has a stron spark when his car does not start, Pertronix will not change it. No need to replace a functioning ignition module.

Now Carl we are curious whether you will notice a pressure deviation when car is not starting. If that is okay, injector spray test will let you know whether injectors are working well. Then finally you have to verify correct ignition timing. All that an engine need to start is: Air, spark and fuel.

Volker schrieb: Graham, that is an interesting hint. I am not fond of the Bosch 0 580 464 999 as it is for 4 Bar and has a maximum pressure of 8 bar. This Nissan pump looks like the Bosch pump. Do you have flow and pressure data for it?

No data, but I do have one new in box. It appears identical to Bosch. There are many in use on Mercedes and Volvos it seems. The company that made them was bought by Bosch at one time. They then discontinued manufacture. But there is still new stock available, mostly on eBay from Nissan dealers.

Regarding Pertronix: In my eyes it makes no sense that Carl changes more things than necessary. If he has a strong spark when his car does not start, Pertronix will not change it. No need to replace a functioning ignition module.

It is hard to discern whether a spark is strong or not. Pertronix eliminates problems caused by distributor shaft wobble at low rpm. Carl already swapped ignition module for a working one, so probably no need to bypass existing unless it has a wiring or other fault.

All that an engine need to start is: Air, spark and fuel.

If only it was that simple

Volker schrieb: Hi Carl,

welcome from down under! Of course English is no problem and you are welcome to write in that language. This seems to be the months of failing MB M110s. I am impressed how much you have done so far. I have read your history of pain (and maybe should do again carefully) and understand that you only have cold start problems and that you do have a strong spark while it will not start. I also read that you have thoroughly tested everything on your car. On the fly 4 things come to my mind:

1. How is fuel pressure on cold car when it does not start? Also 2.1 Bar?
2. Did you do my spray test by letting all 6 injectors spray into glasses with cold engine?
3. Story reminds me of a 350SLC. When really cold it would start but then run maximum 40 km/h. Warm it would go well till 160 km/h and then start to stutter a little bit. In the end it was the rpm based ignition timing adjustment in ignition distributor. Old fat had made it stick when cold. I could only see that when I hooked the car to my igntion tester after a cold night in my workhop.
4. What could also cause this is a leaking injector or cold start valve. It would more or less drown cylinder in fuel and it would take long starting to get rid of it. Should be more than 1 injector.

Hi Volker,

Thank you for the reply!

1. Yes, fuel pressure remains at 2.1 bar during cold cranking and at all other times.
2. Yes, I have done a spray test for all 7 injectors.
3. Yes, have plotted ignition advance curve, it is correct.
4. With regards to the leaking injector, i do not believe this to be the case. I did remove the plugs and inspected the top of the piston crowns after sitting for one night and could see no evidence of fuel dipping from the injectors. I also attempted to start the car with the ECU removed but with the ignition system connected figuring that the engine may fire a revolution or two with any fuel in the combustion chambers, it didn't.

Remember however, that once the engine has finally started, it runs fine. This is my biggest frustration, the fact that there are no running issues, fuel consumption is good and it consumes almost no oil. (Note, the is a small "Stutter" if I snap the throttle wide open from idle, but my friends 280 SE also does this, except his car starts first time .

MBGraham schrieb: This is a tricky one! Cold start in Australia is probably not that cold!

Question - Do you have the starting problem when you first start car in morning? The subsequent starting problems after 30 min sound like the common hot start problems. They are caused by fuel that gets hot after sitting for 20-30 min under hot hood (bonnet) and then flashes to mixture of vapour and liquid through injectors causing overly lean mixture.

One simple test to add to your data bank would be to blow through the fuel return line to make sure that fuel is free to return to the tank from the fuel pressure regulator/damper. If that line was blocked, pump might overheat fuel because it has nowhere to go at low engine speeds.

Is the pump an original D-jet type pump? The flow seems a little low. Should be ~2L/min.

Volker - I too had the centrifugal weights in my distributor seize up. Since then, I always check timing at 3000rpm (which for my NA V-8 is about 27 BTDC.) The 6 cylinder would of course be different.

Hi Graham,

Yes, the starting issue is occurring under all conditions except if the engine has been run for more than 30 seconds and restarted WITHIN 30 minutes. Any other condition will entail hard starting.

Thank you for the tip about the return line. I have not tested this function.

I think pump flow is OK according to factory data.

nordfisch schrieb: Hi Carl,
a Warm welcome from me, too.
As Volker mentioned before, you did almost everything one could do to solve the problems.

This is really a great task for us, too...
Your engine looks as being new, not as rotten as many other car-owners have problems with.

Now, some more ideas from me:
- What about the engine / body grounds? I once owned a W115 that had thick, round wires soldered in at the connectors screwed to the body. This soldered connections had broken and failed from time to time.
- Did you replace any connectors of the wiring harness by new ones? There are connectors being sold with wrong dimensions that don't give reliable contact. Look at the compendium, wiring harness.
- Maybe the injection at your car doesn't get the starting voltage directly from the ignition lock or the starting relais, but from an extra contact at the starter. At Opel-cars (I know better) this function failed sometimes because of contacting problems inside the magnetic switch at the starter (Opel has Bosch-Starters installed, too...)

What you could do more to identify the problem?
- replace the temp sensors by fixed value resistors, both sensors about 300 Ohm would signal 'normal condition' to the ECU. Simple 1/4 Watt-resistors will do.
- when installing LEDs with dropping resistors directly at the injectors you can control their firing under normal operating condition. A resistor with 100 Ohm did the job for me, using a white LED. You must try out the direction of the LEDs- seems once wasn't cared about the direction of these connectors when mounting the harnesses.

My ideas, so far... I will send you a PM regarding a favor you could do to me...

Regards
Norbert

Thank you Norbert,

I have the PM, please give me a few days to reply.

1. I have not tested to wiring loom grounds. Is there one specifically I should look at?

2. No, I have not changed any connectors, and in fact, I have removed and replaced the engine wiring loom, just to check.

3. Let me check the wiring and look for the starter contact if applicable.

4. That is a good idea to check the sensors with fixed value resistors, I will try this, and thank you also for the LED tip.

MBGraham schrieb:

Volker schrieb: Hi Graham,

fuel flow from pump is correct with 1.5 l / min. Original spec for 0 580 464 005 is 750cm³ in 30 seconds. Anyhow if fuel flow is too low you would see too little fuel pressure when car does not start. That is why I asked the fuel pressure on cold car not willing to start.

Volker, you are correct that 1.5l/min is the RATED capacity at 12V and 2bar using n-heptane. Bosch say to use their numbers just as a general guide. In practice with gasoline and higher actual voltage, Djet pumps including my own have higher capacity. So that is why I said 2L/min. Back when Djets first came out, Bosch ran training programs for mechanics and in those the troubleshooting section said AT LEAST 1L/30sec. dl.dropboxusercontent.com/u/54783344/efi330.JPG

Fuel pressure is set by the fuel pressure regulator, not the pump capacity. Even if pump only put out 1L/min, pressure would be the same so long as fuel was being recycled via FPR.

Of course, during starting the engine only needs a small amount of fuel. Most is returned to tank. Low flow would not be a factor. But it could be an indication of pump condition. For example, a malfunctioning internal relief valve that is built into Djet pumps could be spilling fuel back to the suction. Unless car has a full time pressure read out, changes in fuel pressure may not be noticed.

By the way, we have found that the Nissan 280Z pump is essentially the same pump, but was made in Japan. Still available new on eBay at times: www.thezstore.com/page/TZS/PROD/11-3063 . Less expensive and direct fit compared with new Bosch replacement Djet pump.

By the way Carl. Despite this discussion of fuel pumps, I suspect Norbert is more on track that you may have an electrical problem in either the injection or ignition system. If it was my car, I would be tempted to put in a Pertronix points replacement and even perhaps bypass the ignition module (switchgear). They are pretty cheap. And then do as Volker suggested - pull fuel rails and watch injectors discharge into jars during a morning cold start.

Graham thanks for the tip re: the Pertronix. The problem I have with such modifications is twofold:

1. The cars ran perfectly when they were new so I expect mine to be the same! I know this can be a pipe dream with a 42 year old car, but in my case, my car has only done 113,000kms.

2. I am afraid also that I am an originality freak, I know it's old, I just expect it to work!

Volker schrieb: Hi Carl,

I recommend you install a fuel hose into fuel line above injectors via T-distributor, Connect it to a manometer and let it hang out onto the windshield Then you can fix it there with tape and watch your fuel pressure during start and ride.

If it does never deviate, then there is no problem with your pump.

Volker I have done this test and the pressure does not deviate a fraction from 2.1bar.

Thanks to everyone!

My plans is the following:

1. Re-do injector test as per Volker's procedure.
2. Check wiring loom ground connections.
3. Check wiring loom in relation to the starter.
4. Substitute fixed value resistors in place of temperature sensor I + II
5. Try a third ECU.
6. Check return fuel line and damper in supply line.

If the above turns up nothing:

7. Set car on fire and move on.

Hi Carl,

if fuel pressure in fuel line is always 2.1 Bar, there is no need to check 6. Fuel pressure would increase if return line was blocked. And if fuel pump would not deliver enough fuel it would sink. I had the case where it would sink to 1.8 Bar after stopping fuel pump and then sink slowly. It turned out that I had a leaking injector.

However don't do 7. even if I understand your frustration.

I thought I had posted a reply, but somehow it did not show up:
Briefly:
- I would check return line regardless - blockage in swirl pot is a common problem especially if car has not been driven regularly
- Check ground resistance from ECU plug ground (with ecu disconnected) to body nearby. That should check total ground circuit. Then check each circuit from ECU plug as outlined in MB shop manuals. Maybe do cold and then warm. Problem is sometimes IN harness.
- Check all connectors. Sometimes terminals get pushed back inside.

carl888 schrieb: ...
6. Check return fuel line and damper in supply line.
...

Hi,
just noticed this.

The damper is situated in the return line, not in the supply line.
I still don't know why (my car doesn't have a damper at all), but that's how it has to be.
Might be of relevance???


In diesem Beitrag ist für Gäste nicht alles sichtbar. Um alles zu sehen, muss man registriert und angemeldet sein.

Regards
Norbert

Hi Norbert. Seems not all manufacturers installed the damper. I believe it is there to even out pulsations caused by injectors firing as well as from roller positive displacement pump.

To add to confusion, we also have a damper between the fuel tank and the pump suction. For unknown reasons, Mercedes call it a "damper cage" . Maybe it's a translation problem? This one is of different design, but likely serves a similar purpose.

Never heard of damper causing any running problem, but some have rusted and leaked.

Hi Graham,
my car (Opel Diplomat E 2,8E, late model) doesn't have any damper installed originally.
Some users highly recommend the installation of a damper because the engine running 'smoother' with a damper installed. I don't intend to install one - my car runs smooth without the damper and another part installed means another part could go defective...

The early cars of my car-model had installed the damper near the tank you call 'damper cage' , 'Ausgleichsbehälter' in German, translation would be 'reservoirs'.
This reservoirs-damper ist known to make problems. It has 4 hose connections - both the fuel-line-side and the suction-side-line fuel flows through, divided by a diaphragm. This diaphragm can get leaky - the reservoirs now being a bypass between fuel- and suction-line. No fuel pressure anymore, engine running no more.

Regards
Norbert