[NZ00Z3]

A clue from the 112 code. Potentiometer 1 does not have a power supply. It should be 5V. You need to check the voltage at pins 2 to 4 and 1 to 7 at the MDK or there equivalent at the DME end.

Attached is some MDK information I have found on the Interweb over the years. Hope it is useful.

 

[NZ00Z3]

What are the readings for pins?

• 2 - 4
• 1 - 7

Its a resistance divider circuit, so need the above reading as the starting values.

The information was found on the Interweb some time back. It's likely the readings some owner got when testing a MDK. I would not yet use the numbers as a hard and fast pass/fail test until we get some experience with them. Get me the above requested values and I run them through a spreadsheet for you.

 

[NZ00Z3]

I've had a good look at the test results.

The design of the MDK is interesting.

• There are 2 sets of Potentiometers. This allows the DME to compare one with the other to create a fail-safe system.
• Within each set, one potentiometer is biased the opposite to the other. This allows for mathematical checks within the potentiometer set. More fail-safe features.
• All connections within the potentiometer sets are available to the DME. This allows for mathematical checks between the potentiometer set. Yet more fail-safe features.
• The potentiometers within a set are connected in parallel. From the test results we can conclude that the individual potentiometers have the same resistance, that is twice the resistance of the paralleled pair, so that is 2,000 ohms.
• The test results show that the minimum resistance on each potentiometer is 1,000 ohms.
• This makes sense. Engineers like using easy numbers, so a 2,000 ohm Potentiometer with a 1,000 ohm minimum reading, gives a 1,000 ohm range for 100% travel, or 10 ohms per every 1% of travel.
• The potentiometer supply is 5V. This means that the maximum current draw is 5/1,000 = 5mA and the minimum current draw is 5/2,000 = 2.5mA on any leg of the system.
• We test for resistance but the DME measures Voltage.

What we do not know, is what mathematical checks the DME does to ensure that the MDK is safe to use. There are a lot of options with the layers of fail safe in the design.

We do know that the MDK has problems with poor electrical connections. The added resistance upsetting the checks and balances (fail-safe) that the DME does.

If you look at it from a resistance perspective, you get these results. This shows that your MDK is out of spec.

Remember that this is a test sample of 1 and we are trying to do something new and different. We could be doing this all wrong.

 

[NZ00Z3]

Here is the updated Resistance table for the first MDK

Here are the test results from the second MDK

• All of the high resistances are within tolerance.
• All of the low resistances are not in tolerance.
• You have noted 50 ohm resistance jumps/changes between pins 4 and 5 with the MDK closed. This is a sign of a faulty MDK.

The M52TUBXX engine runs on the Idle control valve for lots of the low rev's operation. My testing has shown that for steady speed operation, the Throttle body does not start opening until about 50kph (30 mph). The throttle body does open during acceleration around those speeds, but the results are related to steady speed conditions. At a steady speed of 100 kph (60 mph), the throttle body is only open 12%. This means that the throttle body spends most of its time operating in the 0% to 30% opening range. This is the area of most wear on the carbon resistance strip and probably why you have detected damage on the second MDK at the start/closed position for pins 4 and 5.

I've been doing some more work on testing the MDK. It's not finished. Looks like I'm going to have to pull a known working MDK and do some testing to confirm my thinking.

 

[NZ00Z3]

Had some time to test my cars with the Foxwell NT530 scanner with BMW software. Interesting results. Got all this data from the scanner. Still working my way through it, but thought that you may like to see it in in the raw.

Car was stationary during the tests, key in position 2 with the engine off.

 

[NZ00Z3]

The metal arm that over-rides the cable cam is connected to the throttle valve/butterfly side. You were putting forces on the spring coupling between the butterfly and the cable cam in reverse. There will be some slop in the system, so by turning the system with reverse forces, the potentiometers will be turning slightly differently.

 

[NZ00Z3]

Where the pins that did not change relate to the throttle potentiometers and the ones that did change relate to the pedal potentiometers?

 

[NZ00Z3]

If you are having jumping resistance with the MDK closed try this.

• Fit the MDK to the car. Start it up and let the engine warm up to its normal operating temperature.
• Reach in and adjust the accelerator cable adjuster. Turn it counter clockwise until the RPM just starts to rise. Then turn it back a bit, so the idle is at its normal level (circa 700 rpm).
• Clear the trouble code and go for a drive. See if that fixed the problem. Let me know how it works out.

 

[NZ00Z3]

The idea of adjusting the cable is to try and lift the potentiometer wipes just a little away from the damaged area that causes the resistance jumping.

That's a good idea to practice our soldering skills before attempting the harness. I'm sure there are you-tube clicks that will give you all the soldering tricks and tips.

There should be a rubber grommet on the "F" mount that the cable adjuster goes through. There was one case, where it was claimed that the lack of the grommet was the cause of the problem due to too much slack. Does your cable adjuster press into a rubber grommet on the "F" mount? The part number for the grommet is 35411164739.

 

[NZ00Z3]

Here are a couple of phots of how the cable should look. It looks like your cable has been replaced at some stage with one from a different car.

Have you tested the resistance of the wires in the harness? I suggest that you do, before you go the effort to change the plug.

• Locate the connector X60003 at the DME. It's the big 52 pin connector, number 3 counting from the fire wall to the bumper.
• Remove the outer black plastic cover, 2 tabs and the end.
• Back-probe next to the wires to get a connection. A metal paper click works. Do not front probe the connectors. You are likely to push the probe into the small connector and damage the connector. This just make another hard to find fault to locate.
• Attached is a PDF with a form that I use for another part of MDK testing. It gives the DME and MDK pin outs.
• If your digital multi-meter has a "REL" button. When on ohms, hold the probes together, when the screen say zero or near zero, push the REL button. This will give you a reading relative to zero. It makes the meter more accurate.
• Let me know what the harness resistance are. My thinking is that anything < 0.5 ohm is good.

 

[NZ00Z3]

I am about to go thru this myself. I have a complete spare harness for my '99 E39 528i with M52tu. Rather than cut and solder, I think I am going to replace the entire harness with the used one I bought. I will put "Stabilant 22" or DeOxit on all the connections of the 10-pin connector for the MDK (throttle body).

I'm on the fence regarding the need to change the harness for MDK repairs. You can test the continuity to the harness wires. Test the MDK connector through comparison of the MDK potentiometer resistances both off the car and on the car when tested at the DME. The one area you can't easily test is the DME connections. So just replace the 10 connector pins. Plus, there is already around a 800 ohm resistor in the circuit/series between the potentiometer wipe and the MDK connector. A 2 or even a 10 ohm high resistance joint in series is not going to make very much difference. Rather than highjack this thread, maybe we should discuss this in a separate thread or private conversation?

 

[NZ00Z3]

Yes the wires are dirty on the outside. That does not make them damaged.

Spray a lot of contact cleaner into the back of the MDK harness plug and give it a light scrub. That is where the wires come together and any pollution/salt build up can cause a short between cores. Especially pin groups 6, 7 and 8 plus 3, 4 and 5.

Had another look at your fault codes. Looks like there is a short in the power supply to the Potentiometer 1 group. Suggest you do the following tests.

• Make sure you have a fully charged battery.
• With the MDK removed, Turn the key to position 2, do not start the engine.
• Test for 5V on harness MDK connector pins 1 and 7.
• Test for 5V on harness MDK connector pins 2 and 4.
• Turn the key off.

If you have 5V at both sets of pins, then connect the MDK to the harness and redo the above tests. This time back-probing the MDK harness connector.

Report back with the results.

 

[NZ00Z3]

The pin numbers relate to the numbers next to the individual connectors (pins) on the MDK harness connector. A good light/flashlight will let you see them.

With the MDK removed. You can place the multi-meter probes onto the connector side that goes in to the the MDK. As the probes will have a voltage on them, do not let them touch. They will spark and likely blow a fuse somewhere.

With the MDK installed. You need to back probe the wires side if the MDK harness connector. Back-probing is where you use a small conductive item to extent the connection. I use a bent straight metal paper clip. This give a long wire to put into the connector by the wire and still have a bent bit at the end to easily grip and connect to the meter probes. Once again. As the probes will have a voltage on them, do not let them touch. They will spark and likely blow a fuse somewhere.

 

[NZ00Z3]

The 2 "activation idle adjuster" codes are for the Idle control Valve. Have you remembered to plug it back in?

 

[NZ00Z3]

Did you take the Idle control valve out and give it a clean? It should rattle easily when shaken.

The Idle control valve connector can only go on one way.

When installing the MDK, the accelerator cable should be just slack with the MDK closed.

 

[NZ00Z3]

If checking the ICV plug does not solve your problem, then take the ICV out and give it a clean with carb cleaner or brake cleaner. Lots of black will come out of it.

 

[NZ00Z3]

Sounds like you are closely getting all of this sorted. Well done.

Code 208 is saying that the EWS is not getting the rpm signal from the DME. Clear the code and see if it comes back.

If you are still getting the throttle poti plausibility codes, try a MDK adaptations reset.

Using INPA to delete the adaptations

• Switch off the ignition for 30 seconds.
• Turn on the ignition to position 2. Do not start the engine.
• Delete DME fault code memory and adaptations.
• Press the accelerator pedal to wide open throttle (WOT) four times.
• Remember that this procedure has deleted the engine adaptations, so don’t get surprised if the engine’s idle is a little off when you first start the car. It will correct itself with the first test drive.

 

[NZ00Z3]

My Z3's live with the battery disconnected and a trickle charger on the battery. They do not have any problems when I re-connect them and take them out for a run. One is a M52TUB20 with MS42 DME, similar to your car.

 

[NZ00Z3]

Wow, now that's some file you have posted. Lots of really good information in there. Where did you find it?

I have had a read of B1214_M0EWS. It is saying that you have a communications problem between the engine DME and the EWS. The EWS is not receiving the speed monitoring signal. The communication is via K-Bus, which is a type of CAN-Bus.

Have you had the instrument cluster out recently? If so check that the connections are tight.

This is a good resource on fault finding K-Bus: BMW E46 K-Bus Troubleshooting

 

[NZ00Z3]

I was looking at the pwg angle and dwk angles on INPA and noticed that then opening the throttle body the dwk angles would only be registered when the pwg angle was around 15° then the dwk would creep up... anyone has noticed this behavior?

Under light loads, the first 15% of the accelerator (drivers wish) is provided by the Idle Control Valve. See below

I have also written up the MDK potentiometer cleaning procedure: How to Clean the M52TU MDK Throttle Body Potentiometers