'73 450SE - should I replace intake manifold seals?

[Tomguy]

How thin is your oil? Does it run off the dipstick like water? You may be burning oil at this point if your oil is super thin, which won't help you diagnose by looking at smoke. Plus the fact that if the system was running super-rich and is now OK, it has to blow the carbon out of the exhaust first to really be able to tell.

[Graham]

Originally Posted by a5a1234

I have snapped two pictures - forgive the blurring, as I tried to get close enough to show the details. I hope these images are good enough to give an idea of how sooty the plugs are.

Still blowing what looks like grey/dark smoke.

I'll keep at it and report progress.

My plugs looked like that when the trigger points were malfunctioning. I wouldn't put in new plugs until you have the mixture problems resolved - just clean those ones up. I had a set that looked that bad, but now they run a nice golden brown color.

Maybe time to do an injector flow test - set up a series of small glass jars below the injectors and see how much you collect as well as watch for consistent pulsing and pattern. Remove coil wire while you are doing it!

But the smoke may be the problem - Could be from valve cover vent or engine. Have you done a compression test?

[a5a1234]

Thanks everyone.

I checked the oil today - it was thinner, so I did an oil change. The old oil certainly smelled of gasoline!

I cleaned the PCV valve with some carb cleaner.
I cleaned the multi-pin connector leading to the ignition points under the distributor (contact cleaner).

I checked the radiator for bubbles while running the engine cold - no bubbles!

The injector flow test was done two weeks ago when the injectors were reconditioned. All good now - no leaks, good spray pattern, good flow.

Because of the state of the plugs, I "leaned" out the idle mixture by turning the ECU screw to the left a few turns (I know I can't tell properly until I get an exhaust gas analyzer - I'll try to rent one locally).

When I fired her up, the smoking was definitely less :-) It may well be that the thin oil was burning a bit. It looks quite different now. I would not be embarrassed to take her out onto the public roads, which I plan to do tomorrow.

Will report back after a test drive.

Thanks once again for all the advice!
Joe

[a5a1234]

Success! Well... a great deal better anyway. Thanks so much to everyone who helped out with advice. I took the car out today (three times, with cool-off in between, to test). Some smoke initially, as the system cleaned itself out - then no smoke, smooth running, good power.

In summary, here is what I think the problem was, and the solution.

Intermittent smoking (initally) then more constant smoking at idle. The intermittent smoking was accompanied by running very rough. Also a problem with low vacuum (less than 15" Hg).

I think this intermittent problem was caused by injectors getting stuck open, and dumping fuel. I think I caused the injector problem by cleaning the fuel system with an injector cleaner (Techron).

After replacing all the vacuum hoses (except the air intake plenum seals), and still finding a problem, I checked the fuel pressure. Pressure was OK, but wouldn't hold. Isolated the problem to two leaking injectors. Took all injectors to have them reconditioned. All good - no leaks, good spray pattern, good flow now. Replaced injector seals.

Running OK, but milder, "constant" smoke problem. Changed the oil (old oil smelled of gasoline, and was less viscous than expected). Now no smoking (after those initial few minutes today) and running well.

Thanks once again to everyone who contributes to this site in general, but also specifially to GGR, Graham and Tomguy who helped me solve this problem.

Next issue is replacing the transmission modulator. Will have to post separately with a picture - I can't see where the vacuum line connects to the modulator I bought!

Joe

[Graham]

Posting copy of PMs here so others can read and comment. This is what I have learned from discussions with a couple of long time MB mechanics

Originally Posted by a5a1234

Hi Graham,

You may remember my recent saga with the 1973 450SE. As you will recall, I finally got it running reasonably well once I had the injectors reconditioned. I now have access to an exhaust gas analyzer. I found a neighbourhood service station that has the equipment, and we have arranged for me to bring my car, and the tech will operate the equipment, and I will adjust the car (all for a fee, of course!). You mentioned that you have a correct procedure for doing this. I would be most grateful if you could share.

Thank you,
Joe

Joe,
I have the procedure, but two mechanics gave me slightly different numbers to aim at.

First, you need to have the timing set. It should be at 5-8 deg BTDC with vacuum retard disconnected and plugged. This should give you about 27-30 BTDC at 3000 rpm.

To set MPS, car should be warmed up and you should pull the plug on the throttle position switch. You need to do this, because the ECU has a separate circuit for idle mixture control - pulling the TPS plug disables this. Put the air cleaner back on so you have the air temp sensor working.

Set the knob on the ECU to middle of range - I think it is 11 clicks either way.

With the O2 analyzer warmed up and working, adjust the 4mm screw on the MPS so that you get about 1-1.5% CO ( If you want better mileage, you could go lower and if you want more power, then higher - range that has been suggested by mechanics is 0.5-3.5%. Give engine a little burst in between adjustments. Move in very small increments - say 10-15 deg.

Now, reconnect TPS plug and recheck CO reading. Set it (USING ECU IDLE KNOB) for somewhere in 1.0-2.0% range. Chose something where car idles evenly - probably 1-1.5%.
1 click makes a significant difference.

The MB Technical Data Manual says in group
07.4:

For a '73 the specs for the 117 engine are

Full load 3rd gear shifter position S 3,000 rpm = 2% - 5%
Upper partial load 3rd gear shifter position S, 2,500 rpm 300 mm Hg vacuum 0.1% - 0.5%
Lower partial load 3rd gear shifter position S, 1,500 rpm 300 mm Hg vacuum = 0.2% - 0.5%
Idle neutral, idle speed, oil at operating temp = 0.5% - 2.0%

All % CO
North American specs. Euro are slightly different.

To check these numbers, you would need a dyno, or a meter that would work while driving the car. Plus a vacuum gauge hooked up.

My wide band AFR gauge allows me to see how car is running 100% of time. It cost me $200 plus $50 to install the bung in the exhaust for the )2 sensor. I could easily have spent that going back and forth to shop for testing!

PS: I have not said whether to turn screws CW or CCW - I forget which is which and it is confusing depending on how you look at it. Somewhere on-line it does say. I just try it one way and see how meter responds!

Hopefully the garage has a good 5-gas analyzer or something equivalent and that they have a way to ensure that they get an exhaust sample not mixed with outside air.

I am going to post this in Vintage forum just in case someone spots an error.

Good Luck!

[a5a1234]

Thanks very much Graham. I hope to get to the analyzer next week.

Joe

[Graham]

Originally Posted by a5a1234

Thanks very much Graham. I hope to get to the analyzer next week.

Joe

No problem - I added a few words in red in above post.

[a5a1234]

I took the car to the local garage today, to use the exhaust gas analyzer. The first thing we noticed was that the car was running *very* rich at idle - over 6% CO! So I adjusted that back to about 1.6%. Then we worked on adjusting the CO level at 3000 rpm using the MAP. That level was also quite high, but we brought it back to 3.65%. Had to go back and readjust the mix at idle (to about 1.6%), but that was straightforward. The car certainly runs better - it feels more responsive. Also, the "rich" smell in the exhaust has gone. I'll pull the plugs to see how they are looking now. Thanks again for all the advice with this.

However... there is a problem with idle speed if the car has been sitting. It runs really well when going, but if I stop for any reason and restart - after a few minutes or after a couple of hours - the car sometimes (not always) idles low at 500 rpm or less, and runs rough for a little while - seconds, not minutes. Once I get going, the idle settles down, and is fine when stopped at traffic lights, for example. As I said, the engine seems to run well when driving, it's just the initial idle that isn't right. Any suggestions? I mentioned earlier that I think the ball valve for the fuel return on the fuel pump is stuck. Could that cause this problem? The car seems to be less likely to idle poorly if I turn the ignition key to position 2, but don't start the engine, then turn back to position 1, then again to position 2 and start the engine (I did this to make sure the fuel pressure in the fuel ring was normal, given the possible ball valve problem).

Thanks all,
Joe

[Graham]

Originally Posted by a5a1234

I took the car to the local garage today, to use the exhaust gas analyzer. The first thing we noticed was that the car was running *very* rich at idle - over 6% CO! So I adjusted that back to about 1.6%. Then we worked on adjusting the CO level at 3000 rpm using the MAP. That level was also quite high, but we brought it back to 3.65%. Had to go back and readjust the mix at idle (to about 1.6%), but that was straightforward. The car certainly runs better - it feels more responsive. Also, the "rich" smell in the exhaust has gone. I'll pull the plugs to see how they are looking now. Thanks again for all the advice with this.

However... there is a problem with idle speed if the car has been sitting. It runs really well when going, but if I stop for any reason and restart - after a few minutes or after a couple of hours - the car sometimes (not always) idles low at 500 rpm or less, and runs rough for a little while - seconds, not minutes. Once I get going, the idle settles down, and is fine when stopped at traffic lights, for example. As I said, the engine seems to run well when driving, it's just the initial idle that isn't right. Any suggestions? I mentioned earlier that I think the ball valve for the fuel return on the fuel pump is stuck. Could that cause this problem? The car seems to be less likely to idle poorly if I turn the ignition key to position 2, but don't start the engine, then turn back to position 1, then again to position 2 and start the engine (I did this to make sure the fuel pressure in the fuel ring was normal, given the possible ball valve problem).

Thanks all,
Joe

Joe,
If your car is like mine, you are probably being affected by vapor formation when engine heat-soaks after a stop. We get this when gas stations switch to winter fuel.

You can try increasing the idle speed on cold starts up to say 1200rpm. Then on hot start you "may" be OK. But if the engine has gone lean due to insufficient fuel, then you either live with it, or make a few small modifications.

I have started a write up on hot starting. (I have been living with this for years and decided to document some of my findings!)

http://home.cogeco.ca/~mbz/vapor locking.pdf

[a5a1234]

Thanks Graham. You may have hit the problem exactly. I started the car this morning (cold start) and it started and idled without any issues, so it looks as if vapor lock may be the problem. From reading your document, I understand the condition better now, and can see how having a faulty ball valve would contribute to making the situation worse (lower pressure once the pump stops, while fuel is still hot, and so easier for vapor to form). In the service manual procedure 07.4-120, the following advice is given:

c) Check ball valve in pressure connection of fuel pump by switching-on ignition while disconnecting (pinching) fuel hose of fuel feed line in front of ring line (arrow) the moment the fuel pump stops.

If the fuel pressure at pressure gauge is not reduced, renew fuel pump.

Do I really need to go that route? The fuel pump was replaced by the PO a few years back (with a reconditioned pump, apparently. Might that be a problem?). Any advice on trying to repair the ball valve in the pressure connection of the fuel pump?

Thanks again,
Joe

[Graham]

Originally Posted by a5a1234

Thanks Graham. You may have hit the problem exactly. I started the car this morning (cold start) and it started and idled without any issues, so it looks as if vapor lock may be the problem. From reading your document, I understand the condition better now, and can see how having a faulty ball valve would contribute to making the situation worse (lower pressure once the pump stops, while fuel is still hot, and so easier for vapor to form). In the service manual procedure 07.4-120, the following advice is given:

c) Check ball valve in pressure connection of fuel pump by switching-on ignition while disconnecting (pinching) fuel hose of fuel feed line in front of ring line (arrow) the moment the fuel pump stops.

If the fuel pressure at pressure gauge is not reduced, renew fuel pump.

Do I really need to go that route? The fuel pump was replaced by the PO a few years back (with a reconditioned pump, apparently. Might that be a problem?). Any advice on trying to repair the ball valve in the pressure connection of the fuel pump?

Hi Joe,

Have you done a leak down test? i.e. Does your fuel rail pressure drop down more than the amount specified in the manual (2.0->1.7bar when you switch off and then down to 1.5 after 5min.)?

If your leak down is faster than the above spec, then have you done the various (hard to understand) tests to determine if the leakage is through the fuel pressure regulator, the injectors or back through the pump check valve?

If you have determined the leakage IS back via the pump check valve, then it should be possible to fix that. If your pump is like the one on my '72 SL, it should have an integral check valve on the discharge like the one in the picture below. The check valve screws in the side and can be removed. Maybe it just needs cleaning. But I think I have seen replacement check valves advertised.

If it is an aftermarket pump of different design you would need to have a look at it. It should be possible to add a check valve either between the pump and the filter or after the filter.

It would be just as well to check the flow that your pump is putting out. If it is not providing the specified flow (2l/min), then it might need replacement. My pump puts out about 1.85 L/min and seems to work OK, albeit with a minor hot start problem and then only when we have winter fuel.

I wouldn't buy anything if your pump capacity and the leak down is close to spec. On my car, I see leakdown to about 20psig if I stop at Timmies for a coffee and read the paper Minor hotstart problem is par for course when we have winter fuel and it can be addressed by some of the methods in the pdf I linked to. Slightly high leakage through a set of old injectors is also not unknown. Sometimes we have to just work around these things

[a5a1234]

Hi Graham. Thanks once again for your input. I did the fuel pressure tests as specified in 07.4-120 of the service manual (a couple of weeks ago). Pressure was solid at 28 psi when the pump was running, but dropped to about 5 psi within a few seconds of the pump being switched off. These tests enabled me to determine that the injectors were leaking (and I subsequently had them reconditioned), and also that the test for the ball-valve was positive (i.e. leaking ball valve).

I have attached a picture of my fuel pump setup. I don't see any valve (even on the other side, not visible in this photo). I know the pump was replaced a few years ago. Maybe it was not a Bosch OEM type?

Joe

[Graham]

Originally Posted by a5a1234

Hi Graham. Thanks once again for your input. I did the fuel pressure tests as specified in 07.4-120 of the service manual (a couple of weeks ago). Pressure was solid at 28 psi when the pump was running, but dropped to about 5 psi within a few seconds of the pump being switched off. These tests enabled me to determine that the injectors were leaking (and I subsequently had them reconditioned), and also that the test for the ball-valve was positive (i.e. leaking ball valve).

I have attached a picture of my fuel pump setup. I don't see any valve (even on the other side, not visible in this photo). I know the pump was replaced a few years ago. Maybe it was not a Bosch OEM type?

Joe

Joe,
First of all, no need to keep thanking me. Your questions prompt me to do research for my own car too which uses the same fuel system and pump.

Your pump looks like the original. By the way, these are hard to find and list at about US$850 and sell for about 1/2 that on-line. Not a cheap pump!



The hose barb connection that comes off the side of the pump (and then connects to the filter) is actually a check valve and can be removed. I have had mine off and there is a spring and ball inside.

I would try and clean the check valve first.

Replacement check valves listed on most sites are for later pumps and don't have the hose barb. This is the one most sites list, but it is not the right part:



But, ******** Arizona do list one for EFI cars (but say it is in short supply)

http://tinyurl.com/yareydw


You are not alone - even BMWs have this problem, and one guy had a cheap fix:

http://www.digest.net/bmw/archive/v5/msg03061.html