W124 electrical fun: Two problems
 

Problem #1: I've been studying wiring diagrams and disassembling things left and right to try to figure this out, but am hoping maybe someone has had a similar issue who can direct me ... I have some, but not all, switch lights out in the climate control area (not the CC lighting itseld), as well as a couple other areas. They all starting occasionally going out together, started working again together in between, and have now not been working AT ALL for a week. The dead lights are: Shiftgate, window switches, rear dome switch, antenna switch, rear headrest drop switch, exterior light switch (I actually don't remember this lighting up before, so it may be independent), center and right vent switches. It's just the illimunination that doesn't work ... all the switches function fine. From what I can tell on the wiring diagram, they all have connection to the X6 two-post terminal ("58d" circuit). I have tried cleaning all the involved switches (didn't think that would help but worth a try, and I have pulled out the CC, stereo and shiftgate wood to study the wires. I didn't see anything out of place, loose, bent, shredded, melted, etc. I'm pretty sure I found the X6 connection and both screws and all the wires going to them look fine. I'm guessing it's probably one bad ground ... but I have no idea how to find it in that mass of ground wires. Any ideas on how to proceed? I can try replacing the shiftgate bulb, in case it was a massive coincidence ... but I'm pretty sure it's not bulbs (unless one out bulb would cause a circuit failure). This is both annoying and a bit problematic when driving at night. I know where the buttons are and where the shifter needs to go, but it's still confusing not being able to see all those things.

My second problem, which I half figured out: I have one brake light that is always on, albeit dimly, when the exterior lights are on. For a while, it was actually going out when I hit the brakes. I discovered when I run a ground wire directly from the chassi to the metal strip the brake bulb is on, the light goes out (as it should). I noticed the part of that strip that connects to the ground wiring tab is corroded and not touching the ground tab. My terminology might be off, so I attached a picture with the ground flaw in question circled. Should I try to sodder the connection? I hesitate to do that, because it wasn't soddered in the first place. My alternative would be replace the tail lamp backing, or at least that one strip. Soddering would be easier, but I don't want to create a worse problem if there's more to that connection than I know about. Sorry I don't have a closer picture. My camera is "in the shop" so I only have a cellphone camera.

Update: Even though my wiring diagram does not include it, I figured out based on another part of the FSM that the power to the x6 terminal comes from the N40 relay, which is behind the instrument cluster (ugh ...). Once I searched that, it looks like Jeremy5848 had pretty much the same issue in this thread: http://www.peachparts.com/shopforum/diesel-discussion/231045-another-w124-broken-wiring-problem.html. Only different is I don't have lighting to the heated window switch, which makes sense.
I'm going to try a spare N40 relay I had sitting in a box (score another win for random parts hoarding) and check to see if pin 30 is getting voltage. Anything else I should try while I have the cluster out? It's not my favorite job so I'd prefer not to have to do it more than once.:o

Here is what I found: There is 12v at pin 1 (power supply) of N40 with the lights switched on. There is also 12v at pin 4 (to X6). There is nothing at pin 2, the 58D (to X6/1) circuit. I tried my spare relay and it did not work either. Out of curiosity I tested the spare relay by itself ... when I connected pins 1 and 3 to 12v and ground, I found 12v at pin 4, but not pin 2. Do I have TWO bad relays here or am I understanding the wiring wrong? There should be 12v coming out of pin 2 (58D) and pin 4 (58d) when 12v is applied to pin 1, correct? One thing I found odd was that even with the relay unplugged, the CC and most of the switches (all the ones that were not part of the problem) still lit up with the switch on. I guess they are also powered from something else.
I tried putting a wire directly from pin 1 to pin 2, which was clearly not a great idea because it immediately blew fuse 9. Before the fuse blew, I got a brief lightup from the shiftgate bulb though, so putting juice there does get it to the switches in question.
Should I try buying a new relay? This seems like an odd failure to have in not just the one relay but also in the spare, so I'm wondering if I'm just testing wrong.

A brand new N40 control unit didn't help. 0V is still coming out of 58D. I pretty much give up for now ... whatever it is is way above my head to find. It doesn't make any sense that it would be the rheostat because the cluster and several other lights work (and dim/brighten with the switch). I did try supplying a known good ground to N40 and it didn't help. Since it's blowing fuses when I bypass N40, but not with N40 in place, my guess would be it's blowing something in N40. The fact that I don't understand electricity, at all, probably doesn't help.
Any ideas about the taillamp issue I mentioned in the first post? Should I try to sodder it? It's occurred to me it could be related. Though when I jumped the taillight ground it didn't help the other lights. I want to fix the taillamp either way.

If someone on the board has a link to the ETM for Bodhi's car I would be glad to take a look. This is the kind of thing I do at work all the time but I am lost without a diagram.

Here is a link: http://www.w124-zone.com/downloads/MB%20CD/W124/w124CD1/Program/ETM/ETM.pdf. My car has a build date of Dec. 86. Section 130 has the pages covering this circuit. I've been trying really hard to learn to read these things, and I have a good feel for what belongs in what circuit and roughly where problems might be, but I have a hard time understanding it when I actually sit down in the car with a multimeter looking at all the wires. I really had high hopes the N40 block would do it because the lack of voltage coming out of it explains all the lights that are out. Unfortunately now I need to find out an explanation for the lack of voltage from that pin of N40, since it's not just a bad N40.:confused:

I took a brief look at the link. Unplug N40 and with power off (ignition key out), measure resistance from pin 2 of N40 socket to ground and make sure there is no short, i.e. it should not read zero ohms. If you do. that can explain the blown fuse. If you get zero ohms, find and fix the short before testing another N40.

From what I see, your issue is not with 58d, but 58D. 58d heads over into the dash, which you say is working...so that circuit is fine. If you have it all apart, you can start testing the circuits for continuity (before they hit the 58D switch). So stick one pin on a ground and then test all the grounds for the rest of the switches...then stick one pin on the output of each 58D and check for continuity...unfortunately a huge PITA but that will tell you whether you have any open circuits or shorts in the wiring. Unfortunately all of the wires coming out of that switch appear to be gray/blue? Maybe also check for continuity on the gray/red wire in between the N40 module and the 58D switch

When I connect the ground of the multimeter to the console ground and stick the other pin in pin 2 of the N40 harness, I got 0.00 ... but I also get 0.00 on pretty much everything I stick the pin into to, including the lamps that work. I tried the instrument cluster ground instead and got the same. My multimeter is low on battery so I'll replace those just to make sure I'm getting accurate readings. I'm pretty sure there's at least one short though, given the fact that pin 1 to pin 2 blows fuse 9 immediately. I should have thought of that before trying the new N40 unit and wasted it.
If all the switches read zero ohms, how do I narrow down where the short is? I haven't found anything that doesn't read zero yet.:( It doesn't help that X6/1 is pretty much inaccessible without taking the console out.

You have to set your multimeter to the lowest resistance range, i.e 1 ohm if your meter has one. When in the lowest range, short the meter leads together and note the reading. If it does not read zero, that's ok, just write down the reading. Let's say it's 0.2 ohms when you short the meter leads together, that is your zero calibration. Now measure pin 2 of the N40 socket to ground. If you get 0.2 ohms, you have a short. Find and fix that short by unplugging the loads on pin 2 one by one while watching the meter reading till the 0.2 ohm reading goes away, then you found the short..

I was using the 20 ohm setting, but I will check and see if there's a lower setting, though. I did not think to calibrate the meter.:o
I wish it were a little easier to access the X6/1 connector, because I think that would be a good (central) place to remove each load. Of course all the wires are the same color there but if I found one that made a difference, I could trace its path from there.

Is it possible to remove the loads by unplugging the connector at each load? If you still have a short, then you have to get to the x6/1 junction to isolate the short.

I should be able to do that by just unplugging each switch and taking the light bulb out of the shifter holder. Though I was thinking, if the short is in the wire on the way to a particular switch, wouldn't there still be am extra load on it after taking that switch out (from whatever circuit is crossing into that switch's circuit)?

The short can be anywhere. disconnect the easiest things first, keep going till you find it per the multimeter reading.

Thanks! Jeremy pointed out to me that he had a problem with the wiring to the rear door switch lighting once because of the bent wiring at the B pillar ... which kind of sets off a light bulb (no pun intended) in my head because I have just been fiddling with the rear door window regulator, and having problems with the power supply to it. The switch lighting wire runs through the same place in the B pillar as the regulator wiring, so I am going to start there for a possible short. That would make sense timing-wise, too, because the lights stopped working just about the same time I was working on that door. I'll move along from there if I can't find a problem. Hopefully I will at least learn a lot about wiring from this.:rolleyes:

Two things I found today ... there is actually a little resistance on the circuit at the No. 2 pin of the N40 harness ... about .6 ohms. Is it normal for the resistance reading to kind of bounce all around at first? When I first touch the probe to the socket, it usual jumps to a higher reading, then goes down until it settles around 1 ohm (the reading across leads on the multimeter was .4). The lowest setting I have is 200 ohms. Yesterday I was using the 20K ohms, which I thought was 20.:rolleyes:
Also, I started digging into the rear door lining and found the wires indeed are damaged. The blue wire poking out is the wire from the front switch to the rear switch, but I'm guessing the lighting wire is boogered up too ... I didn't have time to cut the casing apart before work. I also pulled out the puddle lamp lighting, which was damaged even though the puddle lights work. I will fix all the wiring in both rear doors and then see if it fixes the circuit.

B-pillar
 

The effect you're seeing with the ohmmeter may depend as much on the meter as on the circuitry. It sounds as if there's a capacitor somewhere that is absorbing electrons until it gets full. Was N40 still connected when you made this measurement? I don't see anything else in the wiring that would cause such an effect; there's nothing there but incandescent bulbs, LEDs, and resistors.

Here are three pictures of the B-pillar cover and the clip that holds it to the frame of the car at the top. The clip's spring action locks into that rectangular hole in the frame of the car. Pulling straight out should release it. It's also possible that the clip can be unhooked from the B-pillar. Put the clip back on the cover before reinstalling the cover.

Putting the cover back in place isn't too bad except that you have to align the mechanism that adjusts the position of the shoulder strap of the seat belt. It usually takes me several tries.

The B-pillar pictures were taken on my 1995 E300D but IIRC my '87 300D was the same.

Jeremy


http://i323.photobucket.com/albums/n...llar_20780.jpg


http://i323.photobucket.com/albums/n...llar_20781.jpg


http://i323.photobucket.com/albums/n...llar_20782.jpg

WOW the 124 electrical system is way more complex than the 123!!
 

Anyway I finally got a chance to download the diagram and take a look.

I'm fairly new to Mercedes and all my time so far has been on 123s but I think I see what is going on here. The 123 electrical manual spans about a dozen pages and it is crazy simple compared to a 124. I hate to think what's going on in my wife's 2012 Honda Pilot :eek:

Issue #1, no dash lights
I think the N40 relay is like a 'rheostat repeater'. It takes a 12 V reference signal on pin 4 (58d) from the dash rheostat. The items on 58d (mostly on the left side of page 130/3) are directly off the rheostat.

The reference signal on pin 4 has 12 V on it when you have your lights on because you are getting good voltage there, and you said your main cluster lights were working.

Pin 2 (58D) is the output. N40 takes power from pin 1 and electronically regenerates a signal on pin 2 that is a copy of what it is seeing on pin 4. I suspect they did this because there are too many bulbs for the dash rheostat to handle by itself (unlike in the 123, everything is off the rheostat, but I have about 10% of the bulbs you do). They wanted to keep the critical bulbs direct off the rheostat (instruments), but the less important items got relegated to N40.

When you tried to tie Pin 1 (12 V from fuse 9) to Pin 2 (the remainder of the lighting circuit that is not working) you blew fuse 9. That leads me to believe, as you do, that you have a short in your lighting circuit.

Unfortunately it is extremely difficult to use a multimeter to find a short in a circuit with incandescent bulbs. The cold resistance of an incandescent filament is nearly zero; as the filament heats up, the resistance increases to anywhere from several ohms (brake lights) to several hundred ohms (dash lights), inversely proportional to the wattage of the bulb.

Here is a trick to try:
Go to the parts store (or look around in your parts stash) for an automotive bulb socket with wire leads and a 12 V bulb. You want a bulb that is normally used for brake or turn signals. Wire the bulb socket across pin 1 and pin 2 of your N40 socket. This puts the bulb in series with the remainder of the lighting circuit. If you have a downstream short, the test bulb will glow brightly. The advantage is that the bulb limits the current, so you won't blow the F9 fuse while you track down the short, and it gives you a visual indication of what is going on.

Unfortunately you are going to have to bite the bullet and pull your console so you can lift wires at X6/1 one by one to find the shorted one. When you remove the shorted wire, the test bulb will go dim or all the way dark.

That damaged wire pack in your B pillar is certainly suspicious. There is a wire from the output of N40 in that wire pack that goes to the window control switch in the left rear door. (Was this the car that got hit in the left side?)

Once you get the short fixed, you can temporarily run without N40 by jumping pins 1 and 2, the lights will just be at full brightness. If you smoked your N40, try cracking the case, strip out the guts, and solder a wire between pins 1 and 2 for a good temporary fix. The fuse is 8 Amps, so #16 wire is plenty sufficient.

Issue #2, weird tail lights
That is a classic lost ground situation. Fix the bad ground connection and you will be all set.
What is happening is that your left side brake and tail lights share a common ground point. Normally the power for the left side tail lights flows like this:
fuse -> head light switch -> tail light -> ground.

And your brake lights go: fuse -> brake pedal switch -> brake light -> ground.

In both these circuits, the ground is the metal frame of the tail light assembly. There is a wire that brings that frame back to the actual ground on the body. I think you are on the right track that the wire is loose and probably needs to be reattached.

When you lose the left side ground, the power tries to get to ground any way it can. Now the power goes:
fuse -> light switch -> left tail light -> shared ground that isn't really grounded -> left brake light bulb -> brake light circuit (which is at 0 V since you aren't on the brake) -> right brake light bulb -> right side ground.

Your bulbs glow dimly because you have effectively put 3 bulbs in series. The brake bulbs are higher wattage so they glow brighter than the tail bulbs.

When you step on the brake in this condition, you bring the brake light circuit to 12 V. This causes the voltage across the tail light bulb to go to zero, because both sides of the tail light bulb now have 12 V on them. This makes the tail light go out when you step on the brake while the brake lights illuminate normally.

Good advice from jay_bob. Listen to him! The light bulb should be easier to find a short but a DVM will work if you have a discerning eye. The switch illumination bulbs are around 19 ohms ea and when you have a few in parallel, can drop to a few ohms. One of the loads is a heated window, which is the highest load of the bunch and the lowest resistance. Did you try shutting the heated window switch off and see if the resistance changed?

Thanks guys!!! That description is very helpful, jay_bob. This is indeed the car that got hit on that door. The damage to the wires looks more like it occurred over time, but I'm guessing the fact that that door has been opened and closed A LOT since the accident (by me, the adjuster, the body shop, etc.) that it exacerbated the breaks in the already brittle wires. I also played around with the window a lot. It had not worked in several years, and I had the body shop put a good used regulator in ... after that it worked but not every time I pressed the button, and it was dependant on the door's position. This explains that, as well as perhaps the lighting problem. I attached a pic of the wires. Wow! Three were totally severed, and one hanging by a thread. I cut out the damaged parts and stripped the ends. Now I can't decide whether to solder or use butt connectors. I'm not great at soldering, but I'm worried the butt connectors will get pulled apart over time ... this area of the wires is under a lot of repeated stress, as is obvious from the damage. Also, I will need to put in a new segment of wire on each one, because I had to cut out a little of each and I don't think there's enough extra lengths to just put them straight together. That would be two connectors on each wire. But soldering is messy and unhealthy. Thoughts? I'm willing to make the extra effort to solder if it will be a better repair. Also, is it going to be a problem that the replacement wires I have are AWG not metric? They look about the same size but I'm always worried about possibly using too small or too large wire. Maybe I'm overthinking it.
http://compressionignition.files.wor...2116.jpg?w=450

Cleaned up and ready to be put back together:
http://compressionignition.files.wor...2118.jpg?w=450

Good prep job. The electrons don't care whether the gauge is Metric or SAE as long as it's at least as big as the old wire. Crimp connectors are not as reliable as a proper soldered connection but if you are uncomfortable with soldering, use crimp connectors and again, the electrons won't care. Just make sure the wires are well insulated. If you can get shrink tubing to go over the crimp connectors it will help to hold everything together and prevent short circuits. If you have enough wire, consider staggering the connectors along a couple inches of the cable so that the connectors do not form a big lump at one spot. The cable will also be easier to thread back into the accordion.

I soldered mine the first time I fixed it. I tried to get away without adding any new wire, my mistake. The second time I removed a section of the old fragile wire and added some new wire.

You are going about it correctly, but as Jeremy said solder them back together.

The accident may have made the problem worse, but it certainly didn't cause the problem. That is a known issue on the rear door of the 124.

I have not had a problem with the puddles lights so far. I hope it is just the window wiring.

I am going to offer a different advice. Do not solder it if there will be movement in the harness because it will eventually break at the solder joint from the flexing. Use crimp connectors, add extra length wires if necessary (will need extra set of connectors), put heat shrink over each connector and it will be a lasting repair.

Good point. I should have added that when I redid mine I used a long enough piece of wire that my 2 solder joints were in the pillar and in the door. I only left new wire in the area that would flex when the door was opened and closed.

My rear doors are opened about a dozen times a day, so I'll find out soon enough if my repair holds up.

Soldering is ok as long as you strain relief it properly. With all the wires you have to do and being inexperienced with soldering, crimp connectors will give better, guaranteed results. Just be sure to use the right size crimp barrel connectors that the wire fit into tightly before crimping, and squeeze that crimper nice and tight. You want a strong mechanical and electrical connection.

Don't forget to put the heat shrink on first before crimping. I make that mistake often lol.

Good job tracking this down. When splicing, I always lap the wires over each other and solder them down. I also use glue filled heat shrink tubing over each joint and then wrap the whole repair in cloth tape.

The glue filled heat shrink provides a water tight seal, especially good in potentially wet areas.

It's also a good idea as suggested above, to add a little extra wire if you come up short. Get it, short? :)

Anyway, one thing that's a blessing when soldering is a tool called Helping Hands:

Helping Hands

I usually strip everything off, leaving just the bar and the two alligator clips. This allows me to hold wires in place to solder them no matter where they are.

Those helping hands will help a lot, but it may not fit into tight quarters. If you have to hold the wires with your hand, wear leather gloves! To avoid cold solder joints require experience and to know what a cold solder joint look like. I'd still suggest crimp connectors. Strip the sheathing of the wire bundle further back so you have more working room. You can also cut and stagger the wires in the bundle so you don't end up with a big lump.

Well, I finished up the wiring (that was a lot of work in a tight space!) and the good new is, the window works, and the fuse doesn't blow anymore when I jump pin 1 to pin 2 on the N40 unit. Here's the bad: My tach is totally dead, and the fan clutch is on as soon as the car is started. Everything is plugged in the instrument cluster correctly, so now I'm really worried I blew something else in the process of testing. Any immediate thoughts on the connection? I did not have any tach problems before. The other gauges work (not sure about speedo ... did not take it out of the garage).
Today is my last night of work, and I am extremely sad not to have Bodhi with me for the last commute. I tried really hard to finish everything, but I'm not going to drive it if I've blown the OVP or something.
Edit: On second observation I don't think the fan clutch is doing anything abnormal. Think it just sounded different because I cleaned out my garage and it's more cavernous in here. So, just the tach ... I doublechecked the wiring and did a little searching and if I had to guess I'd say I somehow touched a hot wire to something and created a ground problem.:(

The tach signal runs through the EGR electronics; no EGR, no tach. The EGR electronics (in the compartment behind the battery) gets its power from the OVP relay, also behind the battery. Check the fuse (10 Amps) built into it.

If the OVP relay is damaged or the fuse is blown, several other things won't work: Electronic Diesel System (EDS), governs idle speed and EGR valve; ABS; A/C compressor (the Klima relay needs the engine speed signal, which it can't get if EDS is out due to lack of power from OVP). So many things are interconnected!

If OVP works correctly (+12V at pins 2 and 4) then make sure no wires are disconnected at the tach. It needs +12V from Fuse 5 and the engine speed signal from EDS. These two come into the tach case on the back through a 3-pin Molex-style plug (middle pin missing and not used). Ground is through the case -- there is a strip of metal along the bottom of the middle (speedo) section that carries ground between the left and right thirds. Make sure it hasn't been bypassed by incorrect assembly of the instrument cluster.

The EDS electronics seem themselves to be fairly bulletproof but they do need +12V from OVP and the engine speed signal from the speed sensor on the flywheel.

As always, look first at that part of the car that you most recently 'disturbed;' your problem is probably there rather than at the far end of the car.

Congrats on fixing the B-pillar wiring problem (welcome to the club).

Jeremy

Thanks! I've had the EGR plugged for years, but just at the hose. My A/C has been inoperable for a decade, at least, so I can't really test that. The idle speed seems fine ... obviously I can't tell exactly where it is, but it sounds good. I'll doublecheck the fuse ... I was messing around with the fuse box and other fuses so maybe I bumped that one ... I'm pretty sure the front dome lamp was working, though. I have lots of time now so at least I have something to keep me busy.
At least the wiring bit is settled ... it came out kind of lumpy but hopefully it will hold up. I'll probably do the other side fairly soon, since I'm guessing it has been similarly beat up ... might as well do it while the whole process is fresh in my memory and all my wiring equipment is spread across the garage.:cool:
I do still need to find an N40 control unit, or fix one of the ones I have. The lights still don't work with N40 plugged in, but do work with power jumped across the harness. Wish I'd done all this before I bought the brand new N40, but there are worse ways to learn I guess.:cool:

Do you mean pins 2 and 4 on the OVP? How does one go about checking them with the OVP relay in place?
Also, is it worth anything to check the pins in the plug at the back of the tach? Should the one that supplies voltage be hot at all times? I'm hesitant to touch anything behind the cluster with the battery connected, as I don't want to short anything else out. Is it OK to pull the harness off the tach with the engine running to check for a signal from the EDS (if it's even readable with a multimeter?).

On a side note, oddly, when I went to check things out in the car this morning, I found all the lights work normally now. I.e., the N40 unit that still didn't work yesterday, even though the circuit was apparently fixed, now works. I had assumed the short circuit damaged it. I'm extremely confused now, but I guess I shouldn't question it when something magically works.;)

Magic goes both ways
 

Ah, you caught me! Testing the OVP in place without special wiring harnesses is difficult. Get some insulated thin solid wire. Strip an inch or so of insulation from one end, bend it onto a loop just big enough to fit over one of the pins. Do that for pins #2 and #4 then plug the OVP relay back into its socket. Bring the wires out to where you can get at them and now you can measure voltage between each wire and battery ground. When finished, remove the wires and put OVP back.

Pins 2 and 4 should both have +12 Volts when the key is in "Run or Start," engine running or not. If no voltage, and assuming all of the factory wiring and the OVP fuse are both good, then the OVP relay coil has failed. Also, with the key OFF and in your pocket, your ohmmeter should get zero Ohms between pins #2 and #4 (because they are both the output of the OVP relay).

Safety note: OVP pin #1 is hot at all times. Be careful or disconnect the battery or both.

The tach's power lead (IIRC a "+" stamped into the metal case) is hot in "Run or Start." The signal lead gets a positive-going 10 Volt square wave that increases in frequency with engine speed. No telling what a voltmeter might think of it. Try setting the meter to 20 Volts range and see if you get a reading. Try both AC and DC and yes, it's safe to unplug the tach lead with the engine running so you can probe that lead. If you get a signal when the engine is running and it stops when you stop the engine then there is a chance that the EDS electronic box is working.

No telling why N40 now works. Maybe it has an internal circuit breaker that needs time to cool off and reset. Maybe it has a "smoke recovery" ability (electrical things work because they have smoke in them; when the smoke comes out they stop working). Maybe it was just irritated at you. Remember that things that start working by themselves can also stop by themselves; in the meantime "drive safely." .

Jeremy

I am a little worried that it will magically stop working, especially if it stops working in a way that causes yet another problem (I have a pretty powerful fear of things catching on fire in general). I'll be pretty carefully keeping an eye on everything once I get back to driving it. I suppose I could drive it without the tachometer now, but as I mentioned I don't really want to ignore it if it's a problem in the EDS, OVP or a short that could affect something else. I'll try the wire trick on the OVP and check for any kind of signal at the tach harness.

Awesome notes on the b pillar wires. MY left rear window is finniky as well.

B-pillar wiring
 

The broken wires in my B-pillar looked just like pictures in the thread. It's amazing how they can almost tie themselves in knots as they gradually break apart. I've always found it strange that the broken wires usually show up in the B-pillar although most door openings are at the driver's door. Possible explanations:
a) Different subcontractors for the front and rear parts of the passenger compartment;
b) Decision by M-B or sub to use coarser (less expensive) stranded wire for the rear doors since they don't get as much use;
c) what do you think?

Jeremy

Quick question re: the OVP: I noticed it rattles when I pull it out. I've seen other threads on here that suggest a rattling OVP is broken. Is that at all a reliable assumption? I have a hard time imagining it failed in total coincidence to the work I did behind the instrument cluster, though ... seems like the occam's razor is something in the cluster region. It's too late to go out and test more, but I'm just sitting here pondering it and was wondering if that rattle means a thing. 10A fuse looked fine but I tried a different one anyone, no change.

I can't really come up with a reason the rear door wiring would have more problems than the driver's door. Maybe people tend to open the rear doors all the way more (to put things in the car), whereas only open the driver's door partially, enough to get in and out? Maybe the wiring in the front shifts about more, and doesn't get as much wear on one single point? I was impressed how most of the wiring looked fine, other than that one very specific spot.

Rattles are not good. More specifically, a very slight rattle could be the relay contacts, which is OK. A louder rattle that you can hear and even feel, like something has come loose, is not good.

With the engine running I get 14v at the power lead and about 6 v DC on the the input lead ... it drops down slightly when I rev a little bit. I switched to AC and got a reading of 15.8, also dropped when I revved. I forgot to check as I shut down the engine. The fact that there's anything registering on that wire, and it changes with engine speed, leads me toward the theory of it being the ground in the cluster, rather than an input problem. Does that thinking make sense?

Tach problem
 

A bad ground is certainly possible. Notice in the following blow-up picture the path across the speedometer that the electrons must take to get from the tach (and the clock) back to the 3-gauge cluster, where the ground wire is located. If the ground strip does not properly contact the contacts on either side then the tach and clock won't work. BTW, does the clock work?

Your voltmeter measurements sound like there is indeed a signal coming from the EDS module, trying to get to the tach. As you rev the engine the frequency increases; your voltmeter is less sensitive at higher frequencies so it reads less, as you discovered.

Jeremy


http://www.peachparts.com/shopforum/...round_5159.jpg

Unfortunately the clock hasn't worked in a few years. Since I have one on my aftermarket stereo, I didn't bother to fix the clock. Is the ground strip connected to the rheostat? Maybe when the wires hit together while I was working around N40, it damaged the ground through the rheostat. I know there's nothing I can do about it now, but I wish I'd been smart enough not to pull the back of that plug off with the battery connected. It seems obvious now, but all I can do is try to fix whatever is damaged.:rolleyes:
So I have a gameplan before I take it all apart: If the ground strip is damaged, can it be replaced? Repaired? I'd really rather not replace the whole cluster or speedo/odo, because I want the original odo in there if possible.

You can easily swap in your original odo. Jeremy has at least one write up on it.

Cool, I will search for writeups. Guess I can fix that clock too while I've got everything apart.:)

Not saying this is the answer, but ....
 

For the light bulb problem on the tail light, and other electrical gremlins that might come along in the future, change all your fuses from the ones that pit and build up resistance, to the new copper ones.

Ground strip
 

I don't think a rheostat failure can damage the ground strip. Although the rheostat is installed close to one end of the ground strip, there is no electrical connection. Rheostat failures usually involve corrosion/wear/damage such that the rheostat stops conducting current so that the lights don't work.

I'll bet your ground strip is just dirty or a little corroded. Clean it with metal polish or very fine steel wool. If the ground strip is irreparably damaged (unlikely), you can use a short wire with a loop on each end to make a "jumper" that will be a good substitute for the ground strip. Notice in the photo that the two ends of the ground strip, the places that the red arrows (Red Arrows in the UK) point to, have holes in them. Two of the screws that hold the cluster together go through those holes; they also make sure the ground strip makes contact on both ends. No screws, poor contact. (Mercedes left that out of the owner's manual? Sorry.) The jumper wire can go anywhere on the metal back of the tach/clock but it has to go on the correct ground point (same place the strip ground goes) on the 3-gauge section. It may take some wiggling of the the jumper in between the two sections.

Fixing the clock takes a careful pair of hands. You have to remove the tachometer mechanism from the metal back plate in order to then remove the clock. Once the clock is out you need to unsolder and remove two 10 microfarad capacitors and replace them with new ones. I just did that job but didn't take any pictures, shame on me! The soldering is the easy part. I would be happy to guide you through the process and let me know if you need any parts.

Jeremy

Fuses
 

Yes, absolutely.

BB87, good to see the progress. If the tach was working before you did the wire bundle work, double check your work for mis-wiring, shorts, knocking connectors loose etc.

Did you end up soldering or crimping?

I disassembled the cluster today and found the ground strip, which looked intact. It was a little "dirty," so I cleaned it up as Jeremy suggested with fine steel wool and metal polish. Other than that it looks fine. I made sure the screws were holding it tight to the contacts when I put the cluster back togeterh. The cleaning did not help, it still doesn't work at all. I've racked my brain to think of wiring that might be out of place, I re-checked the door wiring (could that even have an effect on the tachometer?). Since there's a signal and 12v going into the tach, and the ground is OK (I'm assuming the ground connection going into the 3-gauge area is OK, because that all works), does that point to something being wrong with the tach itself? I haven't tried testing the output of the OVP as Jeremy described so I can still try that, but it seems to me if the signal is reaching the tach, the OVP or EDS would not be a culprit.
Funola, I ended up using the crimps, with insulated heat shrink and then electric taped the bundles together. They looked pretty lumpy and I'm not sure how they'll hold up, but when I rechecked they are all still intact.

Tach
 

Hmmmm . . . door wiring would not affect the tach as they are on different circuits. If you are getting (apparently) a tach signal from EDS then EDS, OVP and related items are good. As a quick check on EDS, pull the idle-speed adjusting resistor plug (on the inner firewall immediately in front of the brake master cylinder and to the right of the left shock tower). It is round, black, and has numbered positions. Don't have a photo, sorry. Note the position, then pull it out while the engine idles. The idle speed should change (lower). Plug it back in. If it works, EDS works, OVP works, the tach signal has to be there.

Failure of the tach itself is possible if you eliminate everything else. If you know someone with an oscilloscope, the tach signal, available at connector X29/4, the EDS test point, pin #2, under the hood (right side, inner firewall, next to battery -- see photo, ignore ammeter) should be a 10 Volt positive-going square wave. Pin #1 is common (not "ground"). This is the same signal that the tach gets, it's just easier to get at without pulling the tach.

Jeremy



http://i323.photobucket.com/albums/n...oseup_7127.jpg

My dad said he has an oscilliscope that he made himself, but has been sitting on a shelf since the 60s. Guess I should find something I know works to test it on to see if it still works. He said he also has a signal generator, so I guess if that works, I can use that to test it.
The ELR seems to work fine. It was set at 4. When I pulled it out, the idle stayed the same. However, when I turned it to 2 and press it in, the idle dropped. Then when I pulled it out, the idle went up. When I turned it to 6 and pressed in, the idle went up further. Pulled it out, idle dropped a bit. When I press it back in at 4, it stays the same as it is when the knob is out. Does that sound right? 4 seems to be the "default" speed for lack of a better term.
I located the plug near the battery in the photo, though the pins don't seem to be numbered. I'll check in the FSM to see if there's a chart.
I took some pictures earlier of the tach and ground strip in case they are of any interest (for this discussion or for people who find the thread in the future). I could have sworn I took "after" pictures of the cleaned up strip, but apparently not. It looked nice and shiny. There were two marks on the contact area on the back of the tach, shown in the pictures, that wouldn't entirely clean off. But I can't imagine they would cause a total lack of ground, as the part around the screw is clean.
http://compressionignition.files.wor...2181.jpg?w=450
http://compressionignition.files.wor...2182.jpg?w=450
http://compressionignition.files.wor...2183.jpg?w=450
http://compressionignition.files.wor...2184.jpg?w=450
I did notice the solders when the power/signal leads come in look a little "brown," but solid
http://compressionignition.files.wor...2185.jpg?w=450
http://compressionignition.files.wor...2186.jpg?w=450
http://compressionignition.files.wor...2189.jpg?w=450
http://compressionignition.files.wor...2187.jpg?w=450

Clock removal
 

You have 8 pictures; I'll call them Picture 1 through Picture 8. They may be enough for a DIY clock repair article.

1. Several places need to be unsoldered.
Picture 5, the red and green power/signal leads for the tach, where they go into the connector pins in the metal back plate. Note which wire goes where (the picture helps).
Picture 8, brown ground wire.
Picture 7, upper center, the clock ground pin (do this last, when you're ready to take the clock off of the metal plate).

2. The clock-setting mechanism (plastic) is held by one screw through the back plate. Notice also that it has a little "hand" that holds the faceplate where a 3rd screw would go, but there isn't one. Remember this "hand." Remove the screw and you should be able to twist the mechanism to one side and remove it. If it is tight, wait until you have loosened the other screws.

3. The faceplate is held by 2 screws. Remove them with a small screwdriver. Make sure you use a tight-fitting screwdriver with a big enough handle to get a decent grip. These 2 screws are fairly tight; you don't want to strip the slot. The faceplate stays with the gauges -- you don't need to remove the hands from the gauge. This eliminates the risk of damaging hands, scratching faceplate, losing track of which hand goes where, etc.

4. The tach is held to the back plate with 3 screws in a triangular configuration. There are also 2 larger screws side by side and recessed. Remove the 3 screws, leave the 2 recessed screws.

5. The clock is held by 2 screws (the same size as you removed to take off the clock-setting mechanism), two plastic pins, a power terminal, and the soldered ground pin from Picture 7. Remove the 2 screws. Clamp the metal back plate in a vise or have a helper hold it. With one hand heat the ground terminal so that the solder melts; with the other hand wiggle the clock free of the solder connection. It should pull straight away from the metal back plate. The 2 plastic pins and the power terminal aren't fastened to the metal plate, they just stick though it and act as "locator pins."

6. You should now have the clock and the tach off of the backing plate. The two instruments are loosely bound by the faceplate and everything is flopping around. Handle gently.

7. The clock should have two 100 microfarad, 16 Volt electrolytic capacitors side by side along one edge of the little pc board. I can't see them in your pictures. These capacitors are what go bad. They look like tin cans. The ones that I just removed from a clock say "100/16 FRAKO +" where 100/16 is the rating, FRAKO is the brand name, and "+" marks the positive lead (important!). Buy replacements at Radio Shack or an electronic surplus store for a dollar or two each. Try to get as close to 100 microfarads as you can; higher voltage (for example, 20V or 50V) is OK but not lower.

8. Unsolder the old capacitors, noting where the positive lead is. The new caps may have the negative lead marked; as long as you keep track you'll be OK. The pc board may be marked with a "+" next to one of the solder holes.

9. Solder in the new capacitors. Cut the leads almost flush with the pc board. Make sure you didn't spill hot solder on the board where it will create a short circuit.

10. Clean the solder out of the little brass piece on the metal backing plate so the clock's ground pin will go back in. You may have to use a probe or a tiny drill bit. Work slowly and carefully.

11. When the hole is clean and clear, you can fit the clock back in place and the tach will follow it. Put the screws (clock 2, tach 3) back in loosely. Wiggle the clock-setting mechanism back in place and then the single screw.

12. Make sure everything fits properly, then tighten the screws but leave a little slack.

13. Wiggle the faceplate until the "hand" (remember step 2?) and the two screw holes are all lined up. Put the 2 screws in.

14. If everything is lined up, tighten all of the screws.

15. Solder the three wires and the clock ground pin. Apply power to the clock power terminal and ground to the backing plate and see if the clock runs. You may have to wait a few seconds before it starts.

That's pretty much all it is. Note: later model 124s have a slightly different clock design with only one capacitor; they seem to be more reliable. I just tested a 1989 clock and it runs fine without being repaired.

Jeremy