Gotcha. I see the markings on the gears made of harder plastic but I can't find any markings on the softer gears that I am replacing. Those gears seem to be blistered in a waxy substance that when scratched off, would remove a thin layer underneath. The gears are so malleable that I could split one in half with just my fingernail.

JamesDean,

Can you confirm that I bought the correct capacitors for the speedometer and clock? I'm not an expert by any means and the kids at Fry's knew less than I did. I saw uF and MF ratings and it confused the heck out of me. The box on one of them says 47MFD but on the cap, it says 47u@50V. The others seem to match your description.

Eaton,

MFD is another way of saying microfarads or uF. Those all look good to me. I don't remember the voltage specifications off hand but as long as your new cap's voltage is equal to OR higher than the old ones you'll be good!

Thanks for the quick reply! I feel confident that I have the correct parts now.

Hopefully by the end of the week I will have a full, in-depth, video series describing the odometer gear and capacitor replacement. The filming has been done and I am waiting for it to be edited.

560SEC speedometer repair
 

Hi folks just read this chatline and signed up for my 20 cents worth..

Just pulled the cluster from my Australian compliance 1989 560SEC - it was reading 100kph at a real roadspeed of 90kph (satnav check) - no wonder everyone was passing me on the freeway. The cluster has never been changed from new. Replacing the 2 capacitors made only a small difference. I've read a few blogs saying you cannot adjust the speed reading of these Mercedes digital speedometers.

I'm a bit of an electronics nut, so downloaded the data sheet for the drive chip (UAF2115), and found that you CAN adjust the speed reading on these instruments. The 750ohm resistor next to the chip on my PCB (see photo attached) is used to calibrate the meter reading on the dial. I removed it, extending some leads out to find the room to attach a 68ohm resister plus a 20ohm multi-turn precision trimmer. I can now adjust my speed reading by about +/- 15%.
http://www.peachparts.com/shopforum/attachments/tech-help/131236d1438234522-how-w126-odometer-repair-clock-capacitor-replace-560sec-speedo-1.jpg
I'm not convinced this is related to the REAL fault as I still have a linearity problem in this speedo so have more work to do (plus I have to design and 3D print a replacement odometer drive pinion (mine has split).

I'll make another post once I've cleaned up everything and am happy I've got a final solution.

My apologies for a very crude post - maybe I should just stick to electronics!!

I'm going from memory here on a good amount of this so bear with me.

I am fairly certain that resistor is used to configure the galvanometer current. Changing it will cause some linearity problems. I recall that from trying to make a 560 speedo head work for a 300SD.

OdometerGears.com sells replacement gears as does Garagetastic but 3D printing does sound fun so its all in how you wanna spend your time.

Aside from electronics, speed error could be introduced by wheel/tire sizes that are dramatically different from stock sizes. Are you tire/wheels stock or have you changed out?

Hi Dean
I'm new to this chat-line thing and still learning the ropes, but thanks for your comments. You make some good points I had not addressed.
Wheels and tyres are definitely NOT standard issue. Wheels are 16" Momo brand competition types with rims, 7" wide front and 8" on rear. Fitted tyres with rolling radius close to original Mercedes specification. Rear tyres were within +0.2% of original spec and they are currently 1/2 worn, so that wasn't the problem. My speedo error was +10% at 100kph (65mph). This was measured against my Uniden "Guardian Angel" GPS speed minder (designed to act as a speedometer), and very accurate on straight roads.
I had previously checked my odometer against highway markers and it came up spot on for distance, so I assumed it was not a sender problem.

You are 100% right about that resistor. According to the data sheet this resistor is provided to allow the auto manufacturer to compensate for calibration variations in individual speed meters. That would not work if it created serious non-linear effects. My resistor was marked 75.0ohms (an "odd" value), and measured 76.1ohms so it can't be causing the 10% error.

I now have the (adjustable) speedometer reassemble and on the test bench. It comes up as correct and linear all the way from 40kph through to 200kph, and only shows errors of 1-2 kph below 40 (not a problem).
Next step is to finish the cluster overhaul and get back out there on the road. Then I'll know the real truth.

I have no idea why you had such problems with your 300SD conversion. Is it possible the sensors are different? There ARE some components that effect the way the ITT chip "shapes" the sensor pulse-train before it is processed. Two of the resistors appear to provide a custom-trimmed value, and one of these appears to have been installed at Mercedes "assembly" time, not when VDO made the PCB. Can you remember if the other resisters and capacitors next to the chip were identical between the two instruments?

3D printed gear (12 tooth pinion) is fitted and works fine. Pattern now posted on Thingiverse. If the next gear in line lets go it'll be time to contact Pelican Parts, it would much easier than designing another.

Thanks for odometer tip
 

1988 190E 2.3, identical loose gear problem. Odometer sometimes ok on cold days, but really slips on warm days and numbers stand still. Purchased car recently and suspect actual mileage is much higher than indicated. Saw one suggestion using careful application of super glue, but didn't want to try that in such a closely fitted mechanism. Scoring the odometer shaft looked like a much better idea. Small conflict on chasing the shaft out with a drill bit. Rather than risk further disassembly, I filed a little notch in the odometer frame to allow axial access. See photo. #33 drill bit (0.113 inch) worked fine, just barely long enough at 2.75 inches.

Wasn't sure what sort of 'scoring' would work on the shaft, so I just took the vise-grips to it, figuring no material would be removed, merely rearranged. See photo. Evidently that was overkill, since aggressive tapping was required to force that deformed shaft back into the drive gear. Solid connection now.

Electric drill test looks good. I use a short length of 1/4-inch nylon zip-tie to couple the drill to the odometer cable fitting. Kind of loose and eccentric but not likely to do any damage (wear eye protection). My variable speed drill goes up to 75 mph (turning counter-clockwise). Real test will be an 800 mile road trip this weekend. I may finally get some valid miles-per-gallon data.

85 380SE nonworking speedometer, what to do?
 

You should be able to see the small board fastened to the clear gear cover. I'm guessing that cover needs to be removed to replace those gears.

Shall I buy a soldering station and 'unsolder,' those connections to replace the gears or am I missing something?

Been years since i did both SD's. But i do recall it was a purely mechanical repair only required fidgeting with the tiny axles and removing the melted and crumbling yellowed old gears to replace with the new.

No unsoldering anything unless the SE has a different setup.

What you see is what I have.
 

Like you, I remember doing my 84 SD by removing that clear, plastic cap on the gear end and changing the gears. It seemed quite simple though proved to be tedious to handle such small and delicate parts.

Here's another shot.

<img src="http://www.peachparts.com/shopforum/attachment.php?attachmentid=146452&stc=1&d=1517224392">

I soldered in 2 capacitors that I picked up from the local electronics shop. At first the clock only worked for 5min at a time. I noticed that I did not cut one of the leads short enough on the back side of the board and I'm guessing that capasitor was shorting out. Now the clock will keep perfect time as long as the ignition is on. Once the ignition is off I think the clock continues to run for around 4 hours before it stops. There have been various upgrades to the stereo system in the past. I am not stokum on electronics at all, so I'm wondering, does the power for the clock come directly from the battery, or do the capacitors power the clock when the ignition is off. Would they power the clock for 4 hours with no power at all. I'm thinking maybe someone in the past may have spliced wires behind the dash and the clock power is no longer continuous but should be.

To chime in on the odometer gears, I now have some experience that may help someone. I payed for the $65 set of gears, but suspect I could have gotten the same thing for $17 and not paid the middle man. The second thing I learned was, as far as the replacement gears go, the 2 new bigger gears should be riding on clean spindles. I cleaned with wd40 on a rag. The small gear in the package is the only one that needs to be secure on its spindle, it determines the rate all of the rest of the gears turn right up to the trip and odometer. I learned after gluing all gears and trying with dirty spindles, neither of which worked.