EDS/ELR Idle Issues - OM603
 

This is an '86 300SDL OM603

I've been chasing an intermittent idle "miss" for some time now. The car will intermittently and randomly "shake" at idle, especially when in gear. Very similar to the behavior you get with a miss or "skip", although I haven't been able to track down where.

Imagine my total shock when I disconnect the ELR plug on the solenoid on the IP and the engine suddenly runs SMOOTH. AS. GLASS. In complete and utter disbelief, I jumped in the cab and raised the RPM up to 650ish where the ELR usually maintains and the car continued running completely smooth. No shake whatsoever!

So what on earth would be causing the ELR/EDS to make the idle shaky? I assume it is hunting? A related issue is occasional idle RPM stuck ~1000 RPM, then slowly dropping back to normal. I've caught it in the act and unplugged the ELR solenoid with the idle speed instantly dropping. Plug back in, and right back up to ~1000 RPM.

EDS computer going flaky? More bad caps/solder joints? Crankshaft position sensor? Am I just going crazy? :confused:

Believe it or not the EDS computer does have a diagnostic system. It is read with a blink code reader.

I would make a blink reader and pull codes...

Link to EDS test procedure
http://w124-zone.com/downloads/MB%20CD/W124/w124CD1/Program/Engine/602_603/07.1-1120.pdf
Your engine starts on page 22

They call for a special impulse counter but you can make one yourself.

Need a 1 k resistor, LED, pushbutton switch and an enclosure (pill bottle or other small container).
And 3 leads with banana plugs, red, yellow, black
Red lead to one side of resistor
Other side of resistor to LED anode (lead opposite the notch on the case or the longer lead)
LED cathode (lead next to the notch on the case or the short lead) to the yellow lead and one side of the pushbutton
Black lead to the other side of the switch
Solder/heat shrink it all up, install in a pill bottle.

You plug into the diagnostic socket according to the procedure, hit the button, and count the flashes, then look up the code.

Could be a lot of things. The place to start is to read the blink codes. Do you know how to do that?

I have the FSM procedure on how to pull codes from the EDS computer. I will do that in the near future. Thank you Jay Bob for the quick rundown on making a blink tester. I have all the stuff to whip one up in my office at work.

Following pressure from another forum member, I pulled the EDS computer to check for bad caps. Glad I did, not one of the electrolytic caps on the board has a reading greater than .01. Two of the "phenolic" electrolytics have cracked open and leaked on the board.

My intention is to replace the caps, then pull codes. Given my experience with this car, every single electronic module has had bad capacitors and/or bad solder joints, and replacing/resoldering them has restored functionality to every one of them. I have a feeling it will probably address the EDS computer as well. I still plan to pull codes to see if there's anything meaningful stored. With bad filter/reference caps, any codes that are stored currently are likely meaningless.

Out of curiosity, what else does the EDS do on these cars? I'm aware of it doing idle stabilization, tach output, and EGR control (emissions), but does it do anything else or is that pretty well it?

In that web location there is a whole trove of info on the engine. There is another document in that folder that explains the function of the electronic diesel system.

This site is for 124s but the 86-up 126 and the 124 share a lot of systems.

Mercedes Benz Model 124 - OM602, OM603 Maintenance Manuals

On some models, it provides the engine speed signal to Klima, so it can affect your a/c.

When I bought my '87 wagon years ago, it had that idle shake. I disconnected the ELR two-pin connector on the back of the IP, idle was very smooth but too low, so I manually adjusted the idle back up to ~750 rpm using the set screw on the back of the IP, and never looked back.

As far as I could tell, the key contribution of EDS/ELR is increasing the idle speed when the AC compressor kicks in and when the engine is cold.

I forgot about the Klima tach reference. I'm now seriously wondering if the EDS is part of the reason I get rapid cycling of the compressor at very specific engine speeds.

For Example: Doing 70 MPH I run ~2800 RPM and the compressor will cycle on and off every 2 seconds or so. If I drop to ~2700 RPM, or speed up to ~2900 RPM, the cycling stops. With the bad caps I found last night, I have to wonder if the tach signal being sent out is noisy enough that the Klima is affected by it, but not the tach.

We'll see if new caps helps out any of these issues. I fully expect to find fault codes for the EGR system when I pull codes. The diaphragm in the EGR valve is bad, so I have vacuum disabled to the whole system. I don't intend to hook the EGR back up either, so let's not go there. Other than that, the engine controls are factory-stock.

Max - Interesting to hear of another person with the idle shake from the ELR/EDS. I did quite a bit of searching before posting this thread and didn't find anything from anyone else. Hopefully this thread has a happy resolution and helps someone else in the future. If not, my plan is to do exactly what you did - set the idle mechanically and forget about the ELR functionality.

New capacitors came in today so I recapped the EDS computer this evening. Only 1 of the capacitors was any good, the bulk of them didn't even return a capacitance reading when checked with a Philips capacitor tester.

The difference of before/after is PROFOUND. With the EDS solenoid plugged in, the car idles just as smooth as it does with it unplugged now. More interesting than that is the throttle response when returning to idle. Prior to the recap job, the RPM would typically drop to ~800 RPM, then slowly reduce back to 650 or so where the idle speed was set. Occasionally the RPM would stick ~900-1000 RPM and nothing you did would get it back. Unplugging the EDS solenoid would drop it immediately. Post-recap, the idle speed drops immediately to the idle setpoint. None of this weird coasting business, and so far, no hangups at 1000ish RPM either.

The lack of idle shake is a major accomplishment in my opinion. I have ~25 PSI compression difference between cylinder balance, so the engine will never be 100% smooth, but without the EDS enabled, the idle was smooth enough that you didn't notice the car idling. With EDS enabled, on a good day there was a very slight wobble. On a bad day, it would rock the car enough to spill your coffee (not exaggerating, ask my lap). I can only assume it was reacting to that very slight imbalance and buildup of tolerances sending it wildly out of control. Sitting at a stop light would usually have the car rocking by the time the light turned green. It doesn't do that any longer.

I still plan to pull codes and will report back with longer term results. I'm interested to see what the A/C does on the highway on the next warm day.

I don't suppose you could share a list of caps?

There are 5 Electrolytic capacitors on the EDS computer boards. On the 126 the EDS computer is located in the passenger kick compartment and requires no tools to remove. Not sure about the 201 and 124.

4 phillips screws in the lid removes the boards which are clipped together. Looking inside the 2 boards, you can see 2 heatsinks that are clamped to some transistors on one of the boards. You will want to remove the 2 phillips screws holding the OTHER board to the main heatsink/plug assembly. You can then lift it off and pivot backwards on the ribbon cable. There are plastic clips that hold the 2 cards together, they simply unsnap.

Once you have the cards opened up like a book, the locations of the capacitors are obvious and the capacitors are clearly marked. On the board you didn't unscrew from the main heatsink/plug assembly, there are 4 capacitors: 100µF @ 50V Radial, 100µF @ 10V Axial, 22µF @ 40V Radial, 7µF @ 63V Radial.

The board you unscrewed and pivoted out of the way has a single electrolytic capacitor 22µF @ 16V Axial.

The circuit cards are double-sided, so be careful desoldering and resoldering. I used 63V parts for all my replacements. 7µF is also not a standard value, so I substituted a 6.8µF capacitor for it.

Out of all of my capacitors, only the 22µF radial was good (I replaced it anyway). The next highest reading I got out of any of mine was .006µF In Germany, they'd refer to that as "kaput".

I'll keep this post updated on my experience with lack of "jiggliness" at idle. Hopefully this solves that ANNOYING issue.

Disconnected mine, it ran at 500rpm, maybe 490+ something. The idle improved a lot, besides from the low cylinder. It isn't randomly twitching and tweaking constantly at idle.

I wonder how a could start will it affect it.

Thanks, that's really helpful. I'm sometimes getting stuck at 1000RPM with a slow descent to idle speed. I may be able to tackle this before winter sets in.

It happens to me as well on occasions as well.

So we're all on the same page with regards to the idle-descent problem, here's how mine was behaving prior to new caps (for reference, I have my idle set ~650 RPM):

When operating "normally", a quick blip of the throttle would have the idle return to ~750 RPM, pause for a second or so, then drop back to ~650 RPM where the idle is set.

When "misbehaving", a quick blip of the throttle could have the RPM drop to ~1000 RPM and quickly (3-5 seconds) taper back to 650, or it could have the RPM hang at ~900-1000. Sometimes it would slowly (we're talking 15-30 seconds) return to idle, or on occasion, it would just stick indefinitely (several minutes). No amount of throttle manipulation would get it to quit what it was doing, but unplugging the ELR solenoid would immediately drop to ~550 RPM where the mechanical governor is set. The "stuck at 1000 RPM" idle was truly a random occurrence, generally at traffic lights or drive-thrus where a small amount of throttle was required to move the vehicle, then remaining stationary. It's so random, that I've never fully worked out exactly what has to happen to make it stick. I know that slow speed and gentle throttle movements were required. Generally when it would stick, the RPM was less than 2000 RPM prior to the sticking event.

With the capacitors replaced in the ELR, the 'delay' returning to idle appears to be gone. When blipping the throttle, the idle speed drops back to ~650 RPM immediately, and without the pause at the higher RPM. Whether or not capacitors have solved the "stuck at 1000 RPM idle" issue remains to be seen. I typically had that behavior 2-3 times a week and while irritating, didn't hold a candle to the car rocking (that was REALLY irritating). If I make it a week or two without a "stuck at 1000 RPM" idle, we can assume the capacitors were a fix.

I had a laundry list of errands to run this afternoon, both personal and for work. I deliberately went through town to hit all the stop/go areas and slow traffic zones (2 school zones). My car rocking idle has not returned, but more importantly:

1: My WOT A/C cutout switch is actually functioning now. The switch was good, but the A/C would rarely cut out. Now when it does, that car's like a rocket. Very different behavior compared to it's usual sluggy self with the A/C on!

2: You can actually tell the ELR/EDS is actively correcting idle speed when the A/C compressor cycles. There's a mild "bump" when the compressor cycles on or off and I no longer get RPM droop with it on.

3: This one is important: THE KICKDOWN SWITCH WORKS NOW!!! Since I've had the car, the kickdown switch has always been flaky. A hotline to Ms. Cleo would have been helpful to know whether or not it would work, there was no rhyme or reason to its behavior. Over ~2500 RPM, it would not work at all, regardless of load, you'd have to manually downshift using the gear lever. Now it's working 100% of the time I hit it. I even took it on the highway and it'll downshift at highway speeds! It's NEVER done that! It's worth noting that on the '86, the downshift has an RPM sensitive circuit in the KLIMA relay. I don't think the later cars have that, this one does and the KLIMA I have installed has the relay in it for that function.

Still not hot enough to test the A/C cycling on the highway, but it wasn't doing it when I went for my highway run this afternoon.

What a great thread, tremendous contribution here. Should become a DIY thread!

My KLIMA relay would never work, so I ended up removing that and replacing with a work-around relay that restored AC compressor function. I also had the rough idle from EDS/ELR. I really need to try your fix and see if I can restore the whole system to operation. My engine wiring harness is looking rather poorly, with a big of insulation flaking off because it has gotten brittle and cracked when flexed, so I'd attributed my issues to that. Now I need to try this fix.

My original KLIMA relay had the contacts welded on the compressor relay, which ultimately fried the compressor when it ran out of refrigerant (it was UGLY). The replacement KLIMA relay has been working, but the kickdown has never worked properly and the A/C cutout microswitch wouldn't disconnect the compressor, even though the wiring and switch were good. Both functions work now. It's hot today, so I plan on testing the A/C on the highway after getting off work. Hopefully no more rapid-cycle.

It is worth noting that the ELR/EDS will not fix a rough/lumpy idle caused by mechanical problems. If the engine runs rough with the EDS solenoid disconnected, it's gonna run rough with it connected as well. If it smoothes out with the EDS solenoid unplugged, it's likely to be issues in the EDS computer contributing to the shake.

My idle is not perfectly smooth and never has been, but now that the EDS computer is serviced, any roughness I have is completely ignorable, I've been at a couple stop lights where I thought the thing died, but was still running fine, just without the familiar stomach-jiggling shake!

EDS re-cap took care of the rapid cycling A/C compressor on the highway too. Was 87˚ today, and on a 20 minute highway trip, it only cycled once the fan was ramped all the way down to maintain temp in the cabin.

Still no recurrence of the 1000 RPM idle.

Day 5 and still no high stuck idle. The things that started working with the re-cap job are still working, and the car-rockingly rough idle still hasn't returned. I've been doing a LOT of in-town driving to test that thing out, so far it seems to have solved several mysteries.

Pictures
 

It's been a week now and none of my problems have returned and no stuck high idle. I'm calling the recap of the EDS board a success.

For those folks out there who need a photo, I pulled the EDS module out and took some quick shots to reinforce what I said earlier in this thread.

Photo 1: When you pull the EDS computer, make sure you note which way the latching clip (red arrow) faces. The boards inside can go back into the plastic box either direction, but if you get it backwards during reassembly, you won't be able to reattach the wiring harness since it'll be "upside down".

Photo 2: Once you remove the 4 screws that hold the module together, pull the boards out of the plastic box and arrange as shown with the latching clip (blue arrow) at the top right. Remove the 2 screws (red arrows), and squeeze the 2 plastic clips (purple arrows) to release the boards.

Photo 3: At the bottom end of the board pair, you have 2 plastic clips. Separate them at the bulges (red arrows). They should pop apart with a small effort. At this point you should be able to fully separate the boards and lay them open like a book.

Photo 4: With the boards laid open, you will see the 5 electrolytic capacitors you need to replace. They are marked with their values, but they are also given here for reference. I used 63V capacitors for all of my replacements because they were the easiest (and cheapest) to find.

These boards are double-sided circuit traces, so take your time de-soldering. If you get impatient and start ripping and tearing, you can destroy the circuit card. You may need to wick the solder out of the hole, then use the soldering iron to maneuver the old leads out of the holes.

The polarity of the capacitors is marked on the silkscreen of the circuit card. The positive end of each capacitor is marked with a "+" symbol. The capacitor itself will have the negative end marked, usually with a white stripe or with a series of "-" symbols.

As is usually the case, reassembly is the reverse process of disassembly.

Again, that's for a very useful thread.

That's the idea of a forum isn't it? People sharing/transferring knowledge? :beerchug:

One drawback...reading this thread has caused the same symptoms in my SDL. "Pretty sure"


I was hoping for dry linkage lol.


I'll grab the electronics and get to work. Thanks again!


PS...I just love this kind of fix. Kind of like my clock repair. Thought it'd be big bucks to get a new clock. Couple bucks in capacitors and it keeps perfect time.


Thank you Diseasel!

I figured I'd follow up since it's been about a month since I started this thread:

Still not one instance of the high idle issue since re-capping the EDS computer. Not one. Considering I was having it happen 2-3x per week, I'm reasonably confident in saying that repairing the computer solved that issue.

A/C still has not done it's rapid-cycle behavior on the highway. It cuts in and out as you'd expect it to based on temperature in the cabin. It stays engaged until the fan cuts way down and it reaches setpoint, then begins to cycle. Currently not really an issue since the weather is nice enough for windows-down and/or Economy mode.

Kickdown switch has been working like a champ! Gone are the days of repeatedly mashing the pedal and praying for a downshift. Poke it and let the transmission do the work.

Idle improvement is still there too. I occasionally get a rough idle, which I think I tracked down to an air leak in one of the fuel lines to the lift pump. Since replacing it, the idle has been much improved for the last few days. Even with the leak, I still was not having the car-rocking idle issues I had before the EDS computer was re-capped.

Through and through - improvements all the way around in terms of refinement.

Where is this connector ? Thanks

And Does 350SDL has ERL adjustment knob? Just curious.

The ELR plug is on the back of the injection pump. It's a red solenoid with a plug on top. If you unplug it and the engine speed suddenly drops, you know the ELR is doing something.

Not sure if the 350 has the ELR adjustment knob or not. If you do, it's on the firewall behind the fusebox. It's numbered 1-7

Shamelessly bumping this thread. 1) It's a good read and 2) it references Ms. Cleo in post #17.

Unplug one wire engine started running even worst, two wires out its stop running at all. I guess my ECU works good.

Your engine doesn't though. With the EDS solenoid unplugged, the idle should be ~500 RPM. If the engine won't idle that low, you have some severe mechanical problems or the IP is seriously out of adjustment.

The EDS solenoid on the back of the IP is a 2-pin plug. Not sure how you unplugged one wire or another, but it's unimportant - if you unplugged EITHER of the pins on the EDS solenoid, it should have had the same effect - low RPM.

Oops! A Mistake!
 

Just a 2018 update to this thread, I made a mistake on one of the capacitor values! The 7µF is actually a 1µF!

When I originally recapped my ECU, the top was split open on most of my caps. They were those crappy brown "UK" caps that ALWAYS die (seriously, if you see one in any of your car's electronics, just replace it, they're crap!). Because the font is extremely hard to read, I assumed it was a 7µF, especially due to the size of the capacitor.

Flash forward to today, and I was reworking an ECU for a friend's SDL and noticed his was a 1µF of a different brand. Same board, same layout, same design.

I've fought an intermittent idle hunt that's been significantly worse since I tuned the injection pump....at an idle it's like someone rhythmically goosing the throttle, and it isn't always present, but will show up at some point every time the car is driven. Since changing that capacitor to a 1µF, it idles rock steady, no hunt at all!

This thread is old enough that I can't go back up and edit the original post, so hopefully future readers will make it this far and notice the update.

I have been dealing with 603 powered vehicles for many years and have had this issue on several cars and could never figure out what it was. This is great info.

This is MY SDL in question BTW guys.! :D



Late timing cover ordered. No 17 head receiving full rebuild. Guides, OE valves, springs, stem seals. This is very exciting news on my end. :D

EDS/ELR
 

Where is this solenoid and cm box on a 93 w140? Would these idle symptoms be the same ? My ac blows cold and works great but initial start up is rough idle.

The main symptom this box was giving is a lope at idle, like someone fiddling with the rack lever and goosing it in a rhythmic cycle. A rough idle when cold is unlikely to be the EDS system, you can easily enough test the theory. Unplug the 2 pin connector on the big red solenoid on the back of the injection pump and start the car. Idle speed will be ~500-550. If your roughness continues, the EDS system has nothing to do with it and you have mechanical, glow plug, injection, or compression issues.

Ok, Would smoke be in the mix with all this? #2 had a pc leak, I installed a new one but it's still not smooth,better though. Also saw lots of black liquid when I did a diesel purge. Its somewhat better but still not clear in the bottle, have to continue. I pulled the ETR pin, ng car barely ran if at all. While upon inspection I noticed oil and some coolant residue around #1 intake manifold port,coolant thermostat neck area. It's losing oil, and it's making its way down the head and block, that could be why? Also turbo at intake has oil in it, a small line. The impeller is not freewheeling more on the tight side with no play. Time for a compression test?. Sad because everything works Hvac is very nice both heat and ac. It's a florida w140 and needs very little body wise to be mint. Interior is almost perfect. If motor is ng Id like to do a 606 st with matching trans.. It would be the ultimate concrete living room.

I read page 2 of this thread bought the caps, went to replace them and found the 1µF instead of 7µF... then I went on to read page 3. Lesson learned lol. Fortunately I had a 1µF cap in my stash. No worries.

I figured that would happen to someone at some point. By the time I found the issue, the thread was old enough that it wouldn't let me change the original posting. :mad:

No worries, I really appreciate the write-up and now my idle is dead steady all the time.
:drink:

Eric is a good dude, fixed my box right up. Did my SD's also because he is just that sadistic!

I know this is an ancient thread but I was happy to see someone went through recapping the EDS. I fix 1970s stereos as a hobby so the EDS was fairly simple to do, though it is a double sided PCB so you have to be careful. I was amused to see that the caps Mercedes was using in the 80s were the same as B&O and McIntosh in their stereos, of course they're all on my list of junk-replace immediately before they fail.

That being said I replaced the 5 capacitors in mine, a few had drifted quite a bit but none were completely dead. The 4 on the bottom board and 1 on the top board.

Unfortunately the recap does not seem to have entirely fixed my idle issue. Before the capacitor replacement the ELR adjustment did absolutely nothing, now it does actually adjust idle. However once I start driving if the revs are 2k or higher when they drop back down fast the engine drops down to 500 or so RPM and comes back up to 530 or so. Only when coming off the gas slowly does the idle set by the ELR stick. It also was still occasionally sticking at a higher (800-900) RPM idle both before and after the work

I'm wondering what other components go into this idle adjust, maybe I just have a relay that's acting up or something.

Would you repair others?? I have a few.


How stiff is the throttle linkage? I had a 603 in a w140 that would actually go higher than throttle because of a stiff cruise control actuator.

I wouldn't necessarily be opposed to it, it's easy work, just 5 capacitors and some solder connectors to reflow. My only concern would be I don't have much of a way to test the board for function. I would assume back in the day Bosch had a test jig for these to test all the functions.

I'll have to take a look at the throttle linkage. I haven't done the cruise control amp yet as I am relatively new to these vehicles and afraid to tear things too far apart.

We generally have a shop so most of the mechanical work for us due to a lack of mechanical experience and time on my end. just have the skill set to replace capacitors and find bad solder joints.

Not cruise amp - the cruise actuator. On mine it was much stiffer and it affected the dynamic return, and actually held the rack allowing overshoot upwards too.

See this video which was an example:

https://youtu.be/MXt9yiNTHdY?si=6iEbX5j5jJwILkSB

But it would also allow an undershoot as well. Yet you can see that in other cases, the ELR allowed a soft landing.

Interestingly that looks very similar to what mine is doing. It'll start up and idle right where I want it but then after revving it'll land wherever it wants. Though mine also still likes to come down below 530 and come back up to where it's supposed to be. May have to take a video at some point.