1973 450SE intermittent problem - runs rough, loses power, blue smoke - PeachParts Mercedes-Benz Forum

a5a1234 · #1 09-13-2007, 02:02 AM

Hi All. May I describe an intermittent problem I have had with my 1973 450SE (M117.983)?

Most of the time the car runs "clean", but a few times lately (cold start or hot, it doesn't seem to matter) I will get a sudden loss of power, the engine runs roughly and I get "clouds" of thick blue smoke. It seems to start when the car is idling, but once it starts, it continues whether I drive slowly, or more quickly (although I think the problem has always started when I have been idling, or just taking off from a stop). It can happen after 2-3 minutes of idle, or 2-3 minutes of running (so I wonder if the cold start valve might have something to do with it?). I wonder if the idle is set too low, at about 500 rpm? The problem sometimes goes away after running for longer, at traffic speeds, although sometimes I limp home (with my head down) and hope no-body recognizes me! The last two times I have driven the car it has not happened at all - but I have made a point of driving it a little harder. I have looked through the forum, and I understand that this engine's D-Jetronic fuel injection is very sensitive to manifold vacuum pressure (hence my driving it harder, to make more vacuum, or so I think), because I wonder if it might be running rich. Initially I wondered if this might be valve seals. But would that be an intermittent problem? If the engine were running extra-rich for some reason, would that cause the symptoms I describe? What should I look at? Do you need further information?

Please forgive me if my question is phrased badly, or I use the wrong terms, or have wrong ideas - I am quite new to this.

Thanks in advance for any advice you can offer.
Joe

MunichTaxi · #2 09-15-2007, 12:14 AM

I'd do a compression test, to confirm you don't have weak cylinder(s).

Have you replace the spark plugs, spark plug wires, distrutor cap and set the timing?

ozawa · #3 09-16-2007, 02:44 AM

blue smoke is a sign of oil burning while black smoke indicates a rich mixture. there could be a couple of things going on here but i would pull the plugs and "read" them first. if they are sooty and black, but not oily then i'd say either you're running too rich or your plugs are too cold. if the plugs are black and oily, then i'd say you're burning oil. have you been having to add oil?

check your transmission fluid too for if you have a bad transmission modulator, you will suck ATF into the intake and burn it.

the D-Jet cars are very sensitive to vacuum leaks so make sure all vacuum hoses, intake plumbing, breather hoses, etc. are sound.

when i first got my '73 450se, i had the same symptoms you had , but without the smoke. after checking all of the basics and giving the car a tune-up, turned out to be a bad throttle switch.

i'd make sure all of the basics are in order and in a good state of tune.

a5a1234 · #4 09-19-2007, 12:58 AM

Thank you MunichTaxi and ozawa, for your help with my problem. I pulled all the plugs, and found seven of the eight to be black and sooty. The eighth was black and oily. I cleaned all the plugs, and then cleaned the rotor and contact points in the distributor cap (one of which was heavily covered in a dark deposit, two more had light coverings of dark deposit). The spark-plug leads are all fairly new, so hopefully they won't be a problem. The car started OK and sounds as if it is running more smoothly (wishful thinking?) However, I noticed some dark smoke (not thick) from the tail almost immediately. This is different from the previous problem, since the blue smoke only seemed to come after the car was warmed up, and seemed to go away if I drove the car a bit hard. I didn't take the car out for a run this evening, as it was late by the time I finished the work described above. I hope to try it out tomorrow evening.

I have been reading that temperature sensor II can cause symptoms similar to those I describe, so I checked the resistance of both sensor I and sensor II at room temperature. Sensor I (air temperature sensor, I understand) was reading zero ohms at RT - I guess it is 'toast', but apparently it is not so important (is this correct?). Temperature sensor II (coolant temp sensor, I understand) was reading only 250 ohms at 20 degrees. Should this be higher (say in the 2,000-3,000 ohm range)? Should I replace it?

Is there a test I can do at home for the manifold vacuum sensor?

A visual inspection of the vacuum hoses showed no obvious flaws (except for the red vacuum line that runs from the small vessel near the firewall through the firewall with the green and yellow lines. That is obviously cracked and must leak.

I don't have the equipment to do a compression test at home, so will have to wait to have that done professionally. Also, I don't know how to set timing (I am new to this, as I said earlier). But I am willing to get help with the things I can't do myself.

I'll report again after I have had a chance to take the car for a test-drive.

Thank again,
Joe

JimFreeh · #5 09-19-2007, 07:27 AM

Might have a loose valve guide in the head.

Jim

GaryF · #6 09-19-2007, 07:55 AM

My understanding is the temperature sensors go open circuit when they fail?
Are you loosing any Engine Oil or ATF ?

bwostosh · #7 09-19-2007, 12:41 PM

Until the compression and timing are correct, we can only guess
at temp sensors, trigger points, map sensor, etc.

Tomguy · #8 09-19-2007, 09:20 PM

Ok, your engine's injection system is very vacuum-sensitive.

The first thing you NEED to do is replace any rubber bits that may be in question:
- The MAP line (from the manifold to the pressure sensor on the driver's fender). It's VERY long in the 116 and can have one tiny hole that will make your engine run like someone is pouring gas down the throttle body.
- The brake booster line. Check and make sure your booster doesn't leak the vacuum, and the line is probably good but check anyway
- The central locking system. It's manifold-powered, any leaks here can make you run rich. Especially if the leak is so bad your tank doesn't ever get negative pressure stored in it.
- The injector seals. 8 lower seals (the upper ones aren't even seals). Vacuum leaks here will, obviously, result in much higher manifold pressures.
- Manifold plenum seals - the seal halves between the upper and lower manifolds, a bit more of a pain to do (you need to buy the 8 AND intake manifold seals, pull the whole upper manifold) but leaks here will again cause rich running.

After ensuring that, and making sure your distributor retard (and advance, if applicable) lines and unit don't leak either, make sure to set your dwell to spec (point gap), then your timing (not the other way around). And then make sure ALL 8 are firing by pulling off wires at the CAP (the engine will run with decent power on 7 cylinders, so you may not know). Pull 1, your idle should slow a bit, replace it. Pull the next, same thing. If you're not sure if it changed, try a few more times. That could be an issue as well. IF you're still not running correctly after that, then let us know exactly what! But this should all be verified with certainty before going farther.

**NOTE: When pulling wires off at the cap, please use A) Common sense, and B) Long nose, insulated needle-nose pliers with an insulated glove and DO NOT touch the chassis with a body part (like your leg). Failure to do these may result in a shock, which makes your arm jolt. Which is why I suggest long nose pliers so if you do jerk back, your hand doesn't hit the fan.

a5a1234 · #9 09-20-2007, 01:35 AM

Thank you all very much for your advice! It looks like I have a bit of a list of things to get to. This will take me a little while to complete, so I will get cracking, and report back. I will start by getting a MityVac vacuum gauge for diagnosing. I will need to get the compression measured by a professional, I think (since I don't have the equipment).

When I measure the vacuum in the MAP line, I am guessing that I should insert a T-piece between the line and the MAP sensor and read the vacuum from the line while idling. Can anyone tell me what the value should be?

Also, how do I check if the brake booster is losing vacuum (not the vacuum line, but the booster itself, as suggested by Tomguy)?

Thanks again to all who have offered advice.
Joe

Tomguy · #10 09-20-2007, 09:33 AM

Easy way to check the manifold vacuum that I use: The central lock's vac line. I plug the mityvac right into that (and, of course, don't use a "T" since I know the door locks leak). The worst my vacuum ever was, with no leaks, was 14" at idle. It should be in the 18" range. A small adjustment of the idle air speed screw can drastically affect your vacuum here.

Easy way to see if the booster itself leaks vac: Run the engine a bit, then shut it off. Go inside, grab lunch, go back out and open the line at the booster (17mm nut if memory serves). There should be a large release of air into the system. My booster holds vacuum for over a week, I haven't checked it for longer than that (either on purpose or while doing something else that required the removal of that line).

a5a1234 · #11 09-20-2007, 10:58 AM

Thanks very much Tomguy. I will report back on my progress.

Joe

a5a1234 · #12 09-22-2007, 12:22 AM

Hello all. An update on my problem...

I have a vacuum gauge now, and I measured the manifold vacuum by disconnecting the yellow vacuum line that runs from the rear of the manifold from its "Y" connection near the firewall on the DS, and plugging in my gauge. Results definitely not to spec! At idle the vacuum was a steady -10" Hg. At 2000 rpm, the vacuum was a steady -12" Hg. The vacuum returned to zero within about two minutes of the engine being turned off.

I also measured the vacuum from the same yellow line, closer to the manifold (at the short rubber tube that connects the manifold to the one-way valve on the yellow line. Same result.

Tomguy, I did come back after two hours to loosen the nut connecting the vacuum line with the brake booster - no "hiss" at all. But if the system is only building 10" Hg of vacuum, perhaps the brake booster leaked via whatever fault is preventing a proper vacuum from forming? In other words, how can I tell if the vacuum problem is the booster leaking, or the lack of hiss is a symptom of the leak elsewhere?

Any suggestions for how to test the integrity of the MAP line?

The brake booster line is clear plastic - quite thick. Visually it seems OK.

After checking the MAP line (if it checks out OK), is there something else to check before I have to think about replacing the injector seals and intake manifold seals? Is there a way to test the integrity of these?

Many thanks for your patience and guidance.

Joe

Tomguy · #13 09-22-2007, 01:28 AM

If you want to "Check" the MAP line, remove it from the manifold and start sucking away with your vac pump. You should get it down to -15" and not see it drop more than 1" every 10s (Ideally, it shouldn't drop at all). If it does, it's either a leaky line or MAP.

Now, your way of measuring vac sounds like you checked it AFTER the check valve on the line going to the door locks (you removed the line at the "Y"/"T"). You need to remove it before the check valve and look there. I don't think you'd have -10" at idle and then keep -12" at throttle (you would realistically be looking at about -5" at throttle... less vac when you open the throttle, because you're letting air in. I could check to make sure on mine)

a5a1234 · #14 09-25-2007, 12:28 AM

Hello All. A further update on my progress. As suggested by Tomguy, I re-measured the vacuum - this time *between* the valve check for the central locking, and the manifold (about 8 inches from the manifold, at the end of a black rubber hose, where this disconnects from the valve check).

I started the engine, and measured the vacuum at a steady -11" Hg.

I increased the rpm to 2000, and held it there for about 4 minutes (until the engine had reached operating temperature). During this time the vacuum steadily increased to -18.5" Hg. The vacuum needle at this stage was steady.

I reduced the revs to idle (approx. 800 rpm). The vacuum fell to -14" Hg and the needle showed a slow oscillation around the -14" mark (I would say about +/- 0.25" Hg). The idle is not smooth (I am not sure how to describe it, but it is "lumpy", it seems to stumble a little; the rpm drop momentarily, then pick up again - forgive me if I cannot use the correct description here).

I then took a propane torch, and (without lighting it) let propane flow around the injectors and the intake manifold, and along the MAP line and sensor. I was listening for a change in rpm, which I did not detect. But I must confess that I was a little nervous with this technique, and was not as thorough as I perhaps should have been.

I hope to get the cylinder compression tested in the next ten days or so, and a friend has offered to loan me a timing light in the next couple of days. So more information should be available soon.

If the manifold vacuum reaches better than -18" Hg at 2000 rpm, is this to spec, and if so, does this mean that there are no major vacuum leaks? Does the lower vacuum at idle point to another problem?

As always, I will be grateful for any guidance you might offer.

Joe

a5a1234 · #15 09-30-2007, 02:06 AM

Hello All. An update on my intermittent problem...

I looked at the timing this evening with a stroboscope, at idle with the distributor vacuum connected. It appears to be 0 degrees (TDC). I checked the specification value in the service CD. If I am reading this correctly, an M117, model year 1973, should have the timing set to 5 degrees after TDC. Could someone please confirm this for me?

Secondly, I noticed with the stroboscope, that it would "miss" a flash periodically. Is this normal? Is this a concern?

Thirdly (please be patient with this very basic question!), could someone tell me which cylinder is cylinder #1 on this engine? I expected that it would be the front (i.e. radiator) cylinder on the driver's side, but my distributor cap is marked in such a way that the #1 cylinder is closest to the radiator on the passenger's side. I have tried to find out by an Internet search, but can't find specific enough information - numbering seems to vary between car makes, and I can't find what I need.

May I repeat my question from the last post? If the manifold vacuum is reaching better than -18" Hg at 2000 rpm, does this mean that the seals/vacuum lines are OK? If so, what could cause the vacuum to drop down at idle?

Thanks very much for any advice you can offer,
Joe