Engine Shaking at Idle

Hi, I am having a problem with my 1974 450SL Mercedes, which is a D-Jetronic V8. Any suggestions on what I can check next? Today I will check the flow of the injectors into small jars and also resistances at the ECU connector, but I am running out of ideas.

Thanks! Andy

Problem

When I turn the car on the cold engine is shaking enough to shake the car. As it warms up the revs drop and the shaking becomes worse and the engine seems quite unhappy.

General Observations

G1. Acceleration and power appear to be OK.

G2. Shaking is only at idle.

G3. The revs are stable. At cold they are at 1,200 to 1,300 and they gradually drop to 700.

G4. Some smoke out of the tailpipe when first turned on but it quickly dissipates and it is hard to tell the color. Car has always done that if not regularly driven.

G5. Engine sounds OK and revving is smooth.

Air

A1. I have rebuilt the AAV and tested it with water at various temperatures. I think the fact that the revs drop as it warms up and I don't get surging indicate to me that it is working.

A2. Vacuum at the manifold at warm idle is 15inHg. I am at 2,500 ft ASL.

A3. I have set the position of the throttle position sensor with a feeler gauge so that the first three degrees of pedal press don't open the throttle plate per the factory service manual. Visually checked the throttle plate is closed at idle.

A4. The throttle body has been cleaned including the small hole for the vacuum takeoff.

Ignition

I1. I replaced the points with Pertronix a couple of years ago and I have never had a problem with it. I put my scope on the output of the Pertronix and it shows a square waveform at a stable frequency as expected. There are glitches above 20V according to my scope but they are not continuous.

I2. Switchgear is not bypassed.

I3. Coil is original Bosch and I have tried a spare Bosch coil (same part number) and no improvement.

I4. Rotor, cap and wires were all replaced a couple of years ago. I've tested with brand new spark plugs gapped to 0.032" does not give an improvement.

I5. I don't know if the spark is consistent because I haven't been able to find a spark tester that will fit with my wires and with plugs with the push on caps removed (which is how I use them).

I6. I have put in a brand new reman alternator. No improvement.

I7. Battery voltage is fine - all cells are functional - 12.6V output. Battery is only a couple of years old.

I8. I have not changed the voltage regulator.

I9. I have double-checked the ignition wires are all correctly located and fully connected to the spark plugs and cap.

I10. Timing at idle is 8 deg BTDC. I found this gives the best acceleration and I have not changed it in two years. I've tried 5 deg ATDC to 30 deg BTDC and no change in the shaking.

I11. Distributor weights and vacuum plate have been cleaned and lubricated.

I12. The vacuum pod on the distributor appears to have a leak however there is no leak in the switchover valve and AC is turned off during these tests.

Injection

J1. Fuel pump is outputting one litre in 30 seconds. I bypassed the filter and it still outputs one litre in 30 seconds. Voltage to the pump has a 1.4V PTP ripple on it - not sure if this is normal or not. Voltage was measured at the fuel pump relay.

J2. Fuel pressure is 30psi. During injection there is a high frequency variation of the fuel pressure from approx. 26 to 31 psi. I have found a 350SL owner who has the same thing but with no engine shaking.

J3. If I force the engine into an over-run state by pushing on the throttle linkage and then releasing it, while the revs drop the fuel pressure oscillations stop and the engine shaking stops. I know that injection has stopped in this state, so clearly fuel pressure oscillations are caused by injection.

J4. I have rebuilt the trigger points with new blocks. Dwells were measured at (degrees): 136, 134, 121 and 132.

J5. Throttle Position Switch has been cleaned and tested.

J6. It appears to be running very rich. I don't have a CO or AFR meter at the moment. The knob on the ECU is six clicks from fully counter-clockwise so I take that to be six clicks from fully lean. Setting it fully counter-clockwise does not stall the engine or make it any worse running than it currently is.

J7. The MPS I am testing with is not original to the car but it has it's plastic cover intact.

J8. All my testing is at cold idle and occasionally warm idle and all of the plugs are carbon fouled within a few days of doing this.

J9. Soot coming out of the tailpipe.

J10. I have put Techron and premium fuel (91 octane) into the tank and drove the car up and down the freeway at 75MPH and no improvement.

J11. The engine wiring harness is brand new.

J12. I have cleaned the metal grounding point and eyelets for the wiring harness near cyl #4.

J13. Injectors have been cleaned in an ultrasonic cleaner with Techron. They received new seals and pintle caps two years ago.

J14. I have tested the T2 coolant temp sensor at various temperatures and it matches the specifications.

J15. Fuel pressure leak down test was successful - meets specs in factory service manual.

J16. I put my stethoscope on each injector in turn and they they were all clicking away without pauses and seemingly at the same rate.

J17. Fuel pressure on return line between damper and tank is zero psi.

J18. I got the engine hot and connected my scope to the tach signal. It fluctuated around 40 - 41 Hz. I then unplugged each injector, measured the new frequency and plugged it back in. Results in cylinder order (Hz): 37.8, 37, 37.2, 38, 38.4, 36.2, 37, 39.

J19. Injectors each output 7cc3 per minute during cranking with 30psi (2 bar) of fuel rail pressure

Mechanical

M1. Compression of each cylinder in order is (psi): 116, 127, 124, 118, 122, 117, 112, 112. Total compression range 1.02 ATM..

M2. There may be oil in cylinders #3 and #4 - I have to investigate this further.

M3. Valve lash has not been checked.

Hi Andy,

oops so much information. Well 2 things as a quick comment:

1. Did you verify ignition timing at idle, 1500, 3000 rpm? Plus vacuum retard?
2. How is exhaust CO and HC in warm idle?

It possible please use SI measures when specifying pressures, temperatures, volumes. Either your ignition timing seems wrong or you have a bad fuel mixture. It is normal that your car would tend to run better in cold and higher revs and start to go worse as soon as it gets to warm idle.

1. Not yet, only idle and I set it there because that gave best performance, but changing it at idle didn't help with the shaking at idle.

2. I don't have a meter. I don't want to modify my vehicle by cutting holes in the exhaust - do you have a recommendation?

However even if I set the ECU knob to the leanest setting it doesn't help. What other adjustment options do I have?

Thanks! Andy

Hi Andy,

measuring would be best: Use a digital ignition test gun like Equus 5568. Espeicially after you disassembled you distributor. Or even better an engine and 4-gas exhaust tester. Bosch, Sun ... does not matter. They can analyze ignition primary and secondary voltage via oscilloscope plus measure CO and HC.

Thanks I will look into that.

Meanwhile is the voltage ripple to the fuel pump OK?

Are the fuel pressure fluctuations during injection OK?

I would like to at least rule out the electrical system, the fuel pressure regulator and the fuel damper.

Thanks! Andy

Hi Andy,

oscillations in fuel pressure come from opening and closing injectors. I cannot comment on your oscillations as I neither know psi nor your gauge. But it is no surprise that you have no oscillations in over-run. Injectors are switched off in over-run.

In bar the pressure is set to 2.0 bar per the spec and the oscillations are 1.79 bar to 2.13 bar, approx.

In general are oscillations in this range in the fuel pressure expected in a D-Jet system?

What are your thoughts on the ripple to the fuel pump? Can such a ripple affect the fuel flow and therefore affect injection in a D-jet system?

I've updated my original post to add item J18 which I missed off.

Thanks, Andy

Hi,

I believe you see more the shaking of the engine than such high ripples on fuel pressure. It is not normal.

OK, thanks. I don't want to just replace expensive parts in the hopes it will fix it so is the problem caused by the fuel pressure regulator or the damper (in the engine bay)? How can I tell which it is and how can those components be bench tested?

Thanks, Andy

Hi Andy,
what about trying to connect the pressure regulator directly to the return line?
Does this have influence to the pressure oscillations?

Regards
Norbert

edit: There is no need to bench-test regulator and damper. The regulator is part of the fuel-pressure-line, the damper is in the return line. There is a hose between them that can get clogged because of the low angle it is bent to.  

Hi Norbert! Good to hear from you!

So I should try passing them through my ultrasonic cleaner? Fuel hose was all replaced three years ago.

Thanks, Andy
 

Hi Andy,
I don't think a cleaning will 'repair' a defective regulator or damper.
Just remove the damper to find out if this has has an effect.
<There is no absolute need for the damper. There are D-Jetronic-cars without this device.>

Regards
Norbert

 

Hi,

 I recommend that you do not loose yourself in such details as long as you obviously have a very wrong mixture or ignition timing.

OK. I am struggling to find a low cost CO/HC meter that I can just put onto or into the tailpipe. I guess very old units on ebay are no longer calibrated... seems everything now is O2 sensors that have to go near the engine and at an angle.

I have a scope so I am wondering what can be done with that and perhaps some simple electronics?

The problem I see for the mixture is that I don't have any adjustment options. Only six clicks left on the ECU knob. What else could I change to reduce richness?

Thanks, Andy

Hi Andy,

hectic actions will not help. You have to take a systematic approach. After you took ignition distributor apart it is a must to verify rpm and vacuum. Next you want to know your mixture status. If it is too lean or too rich, you have to try to compensate with rheostat. if it does not help, you have to search further. My checklis t is a guidance for you.

Thanks but I have already systematically gone through a large number of checks and I am running out of things to check. For example I know that a faulty T2 sensor can cause richness so I have checked that. Can you please look through my list and see if you can spot anything I have missed off that can cause rich running.

Note that everything on that list I did because it was already shaking and running rich.

Thanks! Andy

 

First tell me how rich please.

Hi Andy,
you tell us: "J7. The MPS I am testing with is not original to the car but it has it's plastic cover intact. "

Does this mean there is a different parts #. installed?

Any MPS will provide basic function to any car, but a false sensor will result in a false mixture.

Tell us the # of the item installed...

Regards
Norbert

Bosch 0 280 100 100

Not original to the car. Never adjusted. No vacuum leaks. Resistances OK.

Andy

Hi,

that is most probably the correct one - if it is not a late 450 SL. Can u make a photo of the black cap and also tell me the Bosch no of ECU. There are 2 possible combinations for 450 SL. But you most probably saw that in appendix A.