85 560 SL mystery for the "real" techs - PeachParts Mercedes-Benz Forum

psfred · #1 10-12-2004, 11:58 PM

My friend Hans has a 560 SL (I believe he said it was an 85) with a very strange problem.

It runs at 20 degrees ATDC at idle, and naturally is rather severely defficient in power and the exhaust manifolds are overheating.

Engine control and ignition computers have both been replaced with new ones, fuel distributor is now OK (more on that later), good ignition, etc. O2 sensor is OK, mixture appears to be OK, but with the timing that late it's rather hard to set.

The car itself has a strange history. Seems the PO had this problem, couldn't solve it, and had a carb installed, alonth with a different distributor. Supposedly ran perfectly for two weeks, the problem (lousy preformance) re-appeared. When Hans got it, the injection system had been re-installed, but with a used fuel distributor that was bad.

It will occasionally work fine cold, but once it gets warm, it goes back to very late ignition timing. EHA current will go to 30 mA when cold, but will never go over 2mA on acceleration once warmed up.

Temperature sensors are, I belive, OK.

According to the two most recent owners, the engine/transmission have not been separated, ruling out the possiblity that the flywheel was installed a bolt off (possible, if difficult).

I suspect the crank position sensor myself, as that is the only thing I can think of that would delay the spark that much. Normally, they work or are inert, but I'm wondering if the signal from the pickup is strange (large skew on upswing rather than a nearly vertical volgate rise) resulting in delayed igntion triggering. Rpm signal from the computer appears to be correct.

Any ideas? MB Hotline is as stumped as Hans, no one has ever heard of this particluar condition.

Peter

mpolli · #2 10-14-2004, 12:26 PM

Hi psfred,

I can't believe this has gone over a day with 55 reads and no response. Well you are asking for "real tech's" but you didn't say what kind so I will chime in since I am a "real" electronic bench tech with 6+ years experience troubleshooting industrial control systems to component level. So there is my resume. I have no experience with your car. My nearest experience is with my 86 Supra which has a L-Jetronic based system. First question sounds dumb but here goes: Don't you set the timing by turning the distributor?

Anyway, in terms of the crank sensor, do you know if it is Hall or VRM type? If you suspect the signal coming out you could look at it with a scope. But the trouble could also be in th box that reads the signal. Could be a bad solder joint or component on that board. There is the approach of just change out the cam sensor but that might be quite a bit of work. As a bench tech my approach might be different than yours. Another thread mentioned that Mercedes uses lead free solder which causes bad solder joints. On the one hand this is not actually true, lead free solder can make perfectly good solder joints. On the other hand cold solder joints do exist and I just fixed one in my 98 C230. And bad connections of all types, and expecially cold solder joints can be very temperature sensitive. In my world connections were the source of all evil and solder joints are one type of connection. Just some ideas to maybe get this going somewhere.

Mike

porkface · #3 10-14-2004, 09:33 PM

ok, i'll bite. even though i'm in over my head. i can't see the car.

first thing i would do is pull the valve covers and 0 the crank pulley. cams lined up? dist pointing to 1? if the auxillary shaft slipped a tooth or 2, or the dist is off a tooth, best i can do. otherwise, i'd bet on the ezl unit or the flywheel has been off. keep us posted, chuck.

Gilly · #4 10-14-2004, 09:48 PM

I'm not sure but I don't think there was an 85 560SL, so double check what car it is.
I's double check also what you're saying about the TDC sensor (it is NOT a crank position sensor, it only senses TDC). It should be located on the front cover, not the flywheel on a 117. So causes for the timing being off, (besides an electronic fault) would include an incorrect front cover or harmonic balancer.
Of course one possiblity of a engine that runs worse as it warms up is a bad coil.

psfred · #5 10-14-2004, 11:10 PM

Gilly:

All I really know about the car is that it's black, it's a 560 SL, and Hans has been pulling what hair he has left out (although he's now to the point he HAS to find out what the problem is just out of curiosity).

When we talked about it, I asked him what could delay the spark if everything was working correctly otherwise (new engine computer, new ignition switching unit). I was thinking slow coil, bad switching unit (but that has been replaced with a new one), bad crank sensor. The crank sensor is supposed to be on the flywheel (I'm sorry this is such a fishing expedition, I'll check in and get better info this week), and the Hotline asked if the flywheel had been removed since getting it back on one bolt out of position would be just about exactly 20 degrees.

If, however, the sensor is on the balancer, I'd bet the balancer has slipped (but how could that show late timing?).

I'll tell him you suspect the coil (as I do -- too much resistance could cause the spark to be late due to slow field collaspe -- wouldn't it be worse at high rpm?).

To answer the distributor question, timing is not set on these engines with complete electronic control, it's done by the computer. The dizzy just gets the spark to the correct cylinder.

Test equipment is working properly, shows the right data on Hans' 560 SEC.

Thanks!

Peter

Gilly · #6 10-15-2004, 08:34 AM

Oh Boy, got out the disc to double check my suspicions, and I am all wet, it is a crank position sensor, not a TDC sensor, and it's in the bell housing. Assuming it is a 86 or newer 560SL, sometimes people will put a newer/bigger engine badge on the trunk, so be careful as far as what the car really is. Since he is indicating the sensor is on the bellhousing, I think it's safe to assume it's a 560.
On the coil, the typical problem I've seen is they break down (short out) at high temps, the insulation inside turns to goo and the engine runs like crahp once it's hot under the hood. High or low rpms, runs really poor, need to scope the ignition to see it, but in the past I've always managed to come up with a "test coil" to see if it fixes it or not.
Gilly

PS If I would have been correct about the engine having a TDC sensor running off the harmonic balancer, the reason it would screw up the timing is the same as if the flywheel would have been installed 1 bolt off, the harmonic balancer has a small magnetic "plug" installed in the edge of it, and this is what the tdc sensor senses. If the harmonic balancer had slipped, or the sensor changed position for any reason, the timing is directly affected. The harmonic balancer is on the crankshaft with a woodruff key to index it to the crankshaft. You can actually buy woodruff keys with offsets to change the ignition timing. I think mainly Hans has to decide if he's got a timing problem (or other ignition problem) or a mixture problem. It should be easy to determine if the timing is correct or not. No offense intended.
PPS If Hans has a 560SEC, have him try swapping the coil over to the SL as a test!

stevebfl · #7 10-15-2004, 09:01 AM

No 116/117 motor has ever had a sensor on the balancer that affected timing.

The only engines that had crank sensors for engine control were the 86 on 116/117. They measuer the flywheel. The front sensors were for a diagnostic tool.

Ignition is mapped on the 86- engines with vacuum, temp, and coding being the only variants I can think of. An overheating engine would get its timing retarded.

Gilly · #8 10-15-2004, 09:15 AM

Wow! That's interesting, what did they use 87-89 then?

Gilly

stevebfl · #9 10-15-2004, 09:22 AM

No 116/117 motor has ever had a sensor on the balancer that affected timing.

The only engines that had crank sensors for engine control were the 86 on 116/117. They measuer the flywheel. The front sensors were for a diagnostic tool.

Ignition is mapped on the 86- engines with vacuum, temp, and coding being the only variants I can think of. An overheating engine would get its timing retarded.

mpolli · #10 10-15-2004, 01:10 PM

Maybe a bad temp sensor fooling the computer into thinking the engine is over heated? Is it the same temp sender that the instrument cluster uses?

Mike

psfred · #11 10-15-2004, 07:43 PM

Thanks guys!

I wondered about the ignition retard function as I know the 560 does not have a knock sensor (usually good for about 10 degrees or so retard when signalling knock). However, all of the systems I've seen advance the timing to raise the engine rpm when too hot, retarding the timing usually makes them run hotter, not cooler.

Engine sometimes runs pretty good cold (especially if just worked on, to fool the technician), then goes "flat" hot. Low vac, very hot exhaust manifolds, late timing, sluggish performance. Timing does increase with rpm, though. Computer has been replaced (removed when the carb was installed, used one aquired when FI restored), so it's unlikely to be the computer.

Setting the mixture would likely be difficult at 20 degrees ignition retard due to lousy combustion, I would think.

Timing has been verified as 20 degrees late with engine warm.

I'll check in with Hans next week (more parts, like usual -- I think I keep their electric bill paid with all the stuff I buy!) as see if he has tested the temp sensors. I believe he has, but I'll suggest it anyway.

I personally don't put much reliance on what the PO says about never having the engine and tranny separated, unless he actually did all his own work anything is possible, including a shop fixing an error and neglecting to inform the owner.

I'm not very familiar with the crank position sensor systems -- is it possible for the pickup to be slow responding to the rotating flywheel tabs? A gradual voltage rise would account for the late timing, but I've only seen working or dead Hall effect pickups, never one that only sort of worked.

Would bent tabs on the flywheel cause the same problem?

Peter

mpolli · #12 10-15-2004, 09:15 PM

In terms of the flywheel being off 1 bolt, the only problem with that theory is that it runs good when cool. In terms of the sensor, I know quite a bit about different types of sensors but I don't know what is on that car. I am guessing that it is Hall which would mean that there is a magnet that it is sensing. Are you sure it is sensing "tabs"? There would have to be only 1 tab then.

Mike

psfred · #13 10-15-2004, 09:23 PM

I think Hans said there are four sheet metal tabs that go past the sensor -- he's had a number of them apart over the years.

I don't know what the timing is cold, but it could still be the Hall effect if it moves when it warms up.

The flywheel will fit one hole off with the bolts all loose...

needless to say, he doesn't want to pull the engine just to check this...

Peter

psfred · #14 10-15-2004, 09:43 PM

Mike:

Four pulses per rotation per the manual, there are four tabs or magnets on the flywheel.

I don't know that the timing is correct cold, only that it runs better!

Peter

stevebfl · #15 10-16-2004, 03:19 PM

Gilly: 86- means from 86 on. That means from 86 to the end of production which contains the 87-89 116/117s and the 90 and 91 models also.

Anyone ever check the ignition portion of the temp sensor? Is the coding plug in place?