[mercmad6.3]

Further to the above, because the WIS is updated constantly a lot of useful info is deleted such as Sheet GF27.80-p-3012G. It says" operation
In order to ensure a safe driving state and to prevent damage to the automatic transmission,the ETC control module (N15/3) switches to limp home mode in the event of critical faults.A fault code assigned to the fault is stored in memory.all solenoid and regulating valves are thus de energized ".

So... that explains the 2602 fault ,ie no power to the solenoids(3.7 volts in my case.). Now... to find the fault .

 

[mrboca]

So if you clear the TCU and ECU faults to remove the limp mode condition, turn the key to IGNITION ON, and then measure the voltage between pins 30 (C1) and 38 (C2), you should read 11 to 14 V. right ?

 

[Witek_M]

So if you clear the TCU and ECU faults to remove the limp mode condition, turn the key to IGNITION ON, and then measure the voltage between pins 30 (C1) and 38 (C2), you should read 11 to 14 V. right ?

This is one of those faults that comes back instantly after key off/on cycle.

 

[Witek_M]

Ok, I rechecked everything, the data link connector is showing power. I have run the ohm meter over the cables from the ECU to the trans and found yet another dead link... I am not happy, this is the second damn contact plate I have fitted. I am going to see if it is just the pilot bush with maybe a damaged pin before dropping the damn pan again. I am heading out for a job this morning but will update when I get back to the shop and look further into this debacle. BTW using my cars VIn i get the correct wiring diagram for my vehicle in the WIS.

What dead link?

Attached is worksheet to check basics. If you are serious about finding the problem you need to be systematic.

 

[mrboca]

This is one of those faults that comes back instantly after key off/on cycle.

Only if the fault persists.. I do not know if the OP has tried to clear the codes after all these cable tests, and fixes (if any) yet.

 

[mercmad6.3]

Ok, After an hour or so of testing cables ,looking for cracks in the ECM's PCB (nothing out of order there) , I did find the link to the y5 solenoid is dead. That didn't show up on the scan, just the p2602 code. So after doing all the usual continuity tests on the trans feed cable ( I pulled it up into the car to make it easier to check ) all of the cables including #16 ,which was the dead one, tested OK with around .01 ohms.

That only left me with one option, test the contact plate in place with a little test tool I have,basically a cable plug with the leads still in it. I just plugged that into the trans and ran the meter over the lead ends. All except #8 lead were OK . That of course meant dropping the pan and having a look at the contact plate and the seemingly dead solenoid .
Once I had the shift plate assembly out and on the bench I saw immediately what was wrong .... it has a pin missing . Where are these damn things being made? (or need i ask?) .
So tomorrow I am going to retest all the solenoids ,then fitting contact plate number 4 and I will see if pin 38 on the trans ECM is delivering battery voltage . I notice there is a voltage regulator chip in the ECM and if I still can't get battery voltage I will try and get a replacement chip. onwards and up wards....

 

[mercmad6.3]

So if you clear the TCU and ECU faults to remove the limp mode condition, turn the key to IGNITION ON, and then measure the voltage between pins 30 (C1) and 38 (C2), you should read 11 to 14 V. right ?

Nope, in fact the scanner said there wasn't any power to the ECM.... arrgh!.

 

[mrboca]

There are 12 wires, (even though the connector is usually refereed to as "13-pin"). No pin in pin#5 position on the conductor plate connector, and no corresponding socket on the vehicle harness connector.

An open on y5 grounding pin (pin 8 on the conductor plate, when you check continuity between pin 6 and pin 8 as you look into the conductor plate connector) should not cause the P2602. There is a specific code for that (internal code 99).

Also checking the cable harness involves in not only checking the continuity between the two ends, but also checking for discontinuities between the suspected wire and the other wires. This would make sure there is no shorting between the solenoid supply wire and other wires (just check from one end while the other end is disconnected).

I think the TCM power supply has a built in protection against shorts. Few days ago some other poster had the same P2602 with 3.7V at pin 38, and he ended up removing the conductor plate, removing and checking the solenoids, and installing a new plug adapter, and clearing the faults at the TCM and the ECM. His problem was solved.

 

[MAVA]

Glad you found the problem. It comes down to insertion of that connector plug. Yes, the pins are fragile, but jamming-in the plug is not the solution.

Practice the proper insertion and removal a few times. Lube lightly the seals. Here is the process..

http://www.benzworld.org/forums/w210-e-class/1570724-my-attempt-replacing-transmission-plug.html

All the best,

Martin

 

[Witek_M]

Once I had the shift plate assembly out and on the bench I saw immediately what was wrong .... it has a pin missing

Missing pin, where? Pictures?

 

[mercmad6.3]

One of the pins #8 in the contact plate wasn't there .So today I fitted another new plate ,i double checked everything after fitting it to the shift body and again when the shift was bolted back in.. Cables all measured within specs to indicate all the solenoids are working /connected as they should be... annnd .
I hooked everything up,filled the fluid and double checked the level at cold . Fitted the scanner, turned on the ignition ,let the scanner do it's thing and read off the code!!! yes i have a new code .

P2107 The internal electrical check of Component Y3/6y1(Modulating pressure control solenoid valve) has failed.




The ECM is still showing only 3.76 volts and the fault code will not clear . Do I have a ECM fault? this NOT a throttle switch fault or faulty solenoid as they all tested well within spec. (GM cars use the same number to indicate a fault with fly by wire throttles) .


A note on these plates. i have a couple of old original plates here with mercedes markings. The new ones came in boxes marked Febi but there are no part numbers or other ID . The two halves are held together with little white plastic pins. The originals are riveted and glued by the look of them.

 

[tabijan]

I have to ask again since I reread the thread but didn't see a clear answer - how is the solenoid in question acquitted of fault? After all this time I'd have thrown in a set of known good solenoids for the heck of it.

Sixto
83 300SD
97 E320 wagon

 

[mrboca]

When are you measuring the voltage on pins 38 and 30? With the ignition ON or with the engine running ? You should measure it with ignition ON the key position (RUN), when the engine is not running. The MPC (modulated pressure control) solenoid works with pulse width modulation (pulsing the pin 36 at the TCM). This will adjust the exhaust from the solenoid based on differing line pressures (just like a fuel injector operation). So just a resistance check will not 100 percent check the operation of the MPC. There may be something wrong with the MPC (stuck) or not properly installed (no o-rings so the mating surfaces must be absolutely clean).

A solenoid check should also involve in the actual valve operation action, which can be done with a coupling hose, and pressured metered air. This will ensure that the solenoid is not stuck and electrically operational.

 

[mercmad6.3]

There are 12 wires, (even though the connector is usually refereed to as "13-pin"). No pin in pin#5 position on the conductor plate connector, and no corresponding socket on the vehicle harness connector.

An open on y5 grounding pin (pin 8 on the conductor plate, when you check continuity between pin 6 and pin 8 as you look into the conductor plate connector) should not cause the P2602. There is a specific code for that (internal code 99).

Also checking the cable harness involves in not only checking the continuity between the two ends, but also checking for discontinuities between the suspected wire and the other wires. This would make sure there is no shorting between the solenoid supply wire and other wires (just check from one end while the other end is disconnected).

I think the TCM power supply has a built in protection against shorts. Few days ago some other poster had the same P2602 with 3.7V at pin 38, and he ended up removing the conductor plate, removing and checking the solenoids, and installing a new plug adapter, and clearing the faults at the TCM and the ECM. His problem was solved.

Having carried out this exercise more times than i want to count I am getting to the stage where i might just try a new pilot bush and order a genuine plate from MB . Everything is new already which really has me scratcvhing my head .
Today i opened up one the ECMs i have plugged it in and checked the voltages . As before, i am still only betting 3.9 volts at #38 . So it's some sort of shut down mode. I checked the AAM for faults and there's' nothing there .
I have read off the voltages at the voltage regulator chip, 11.97, then 10volts,then 5 followed by a 0 then 3.9 .3.9 and zero. i have no idea if these are correct and the WIS suggests that the fault is cables or the ECM ... i need more info on this ECM before i commit to spending more cash.

 

[Witek_M]

Having carried out this exercise more times than i want to count I am getting to the stage where i might just try a new pilot bush and order a genuine plate from MB . Everything is new already which really has me scratcvhing my head .
Today i opened up one the ECMs i have plugged it in and checked the voltages . As before, i am still only betting 3.9 volts at #38 . So it's some sort of shut down mode. I checked the AAM for faults and there's' nothing there .
I have read off the voltages at the voltage regulator chip, 11.97, then 10volts,then 5 followed by a 0 then 3.9 .3.9 and zero. i have no idea if these are correct and the WIS suggests that the fault is cables or the ECM ... i need more info on this ECM before i commit to spending more cash.

Get assembled valve body back on the bench. Using your test connector apply 12V to voltage supply. Then in turn complete the circuit for each solenoid using amp meter connected in series. Note current consumption for each solenoid.

I am thinking TCM during boot process checks for short or open circuit and if things are wrong goes into protection mode.

 

[Al_Savage]

Get assembled valve body back on the bench. Using your test connector apply 12V to voltage supply. Then in turn complete the circuit for each solenoid using amp meter connected in series.

No!

http://www.atraonline.com/gears/2006/2006-04/2006-04_04.pdf (page 12):

One more thing; all six solenoids
are pulse-width-modulated and have a
low resistance so don’t get cute and fire
up the shift solenoids for any length of
time with straight battery voltage. If you
want to shift this transmission electrically
use a 5-volt power supply.

The solenoids are not designed for 100% current duty cycle at 12v. Use a 5v supply if testing for current consumption.

 

[mrboca]

"One more thing; all six solenoids
are pulse-width-modulated and have a
low resistance so don’t get cute and fire
up the shift solenoids for any length of
time with straight battery voltage. If you
want to shift this transmission electrically
use a 5-volt power supply."


Respectfully, I do not agree with this statement. Only MOD PC, SHIFT PC and TCC work with PWM (strictly meaning, to regulate the flow). The other 3 are ON and OFF.

Second, Witek did not say you connect a car battery directly. The supply voltage to the solenoids is 11 to 14V (see Startek) and the solenoids are PWM'med or turned on/ off via application of ground. I am sure within the TCM the power to the solenoid is "derived" and regulated. so there is current limiting there, plus a rudimentary current sensing circuit, so the solenoid operation is checked. The important factor is the current really, not the voltage to certain extent. The electrical continuity and the resistance of the solenoids can be done with a good ohm-meter, but this does not guarantee an actual operation of the solenoid controlled valve.

Note also that while the car is in Park / Neutral the MOD PC and SHIFT PC are pulsed with 40 and 33 percent duty cycles. So if the voltage measurement are done across these while IDLING, you will have much less voltage across them due to averaging nature of the multimeter. It is best to use a scope for these voltage measurements if available.

 

[Al_Savage]

Only MOD PC, SHIFT PC and TCC work with PWM (strictly meaning, to regulate the flow). The other 3 are ON and OFF.

What you say seems reasonable. I was quoting the Automatic Transmission Rebuilders Association article, and assumed they know whereof they speak, but mistakes are made; maybe they got it wrong.

Second, Witek did not say you connect a car battery directly.

Pretty much, yes, that's what he said:

Get assembled valve body back on the bench. Using your test connector apply 12V to voltage supply. Then in turn complete the circuit for each solenoid using amp meter connected in series. Note current consumption for each solenoid.

If it's on the bench with a test connector, we aren't feeding current to the valve body's solenoids via the EGS, so any theoretical current limiting that may or may not exist is moot.

"Don't use a non-current-limited 12v source to test the solenoids" is my interpretation of ATRA's advice, and I think that's good advice. Just because the shift solenoids are not PWM does not imply that they will handle being fed 12v for more than seconds. As an example, it's quite common for diesel glow plugs to be rated for 4, 6, or 8v and be installed in 12v circuits. Exceed the duty cycle and they're shot. Could be a similar situation for the 722.6's shift solenoids. We don't know: use 5v supply for testing.

I agree with all else you said, esp. use of a scope if you want to watch the duty cycle, but I suspect the OP's problems is not going to require data that fine.

From what I understand, if P2602 is being set (disabling all EGS outputs, rendering the AT to hydraulic-only operation: limp mode), waveforms won't matter: the circuit is open or shorted.

 

[mrboca]

The 11 to 14 V that you get on pin 38 (with ref to pin 30) is the OPEN circuit voltage. During normal operation, you have different solenoids being operated some with PWM and some ON/ OFF. Those which are operated on/off are only on when shifting occurs, so at any point, the current supplied to the solenoids change. Therefore I consider the solenoid supply on the TCM as a variable current supply that is designed to meet the requirements of varying current demands.

If there are no shorts anywhere, when the ignition key is turned ON (engine is NOT running), all solenoids return paths are opened at the TCM, and the voltage should be 11-14V. So if you are in limp mode, The TCM does not need to drop the supply voltage, all it has to do is to open the solenoid returns at the TCM side, and then you would get the full supply volts, not 3.7V when the ignition is ON. When you start the engine, the MOD PC solenoid is supplied with PWM, so the 3.7 V he is seeing between pins 38 and 30 is the average voltage across the MOD PC solenoid PLUS the wiring. I believe this is normal. I would be worried to see full 12V when the engine is RUNNING.

The OP claims after his latest checks, he does not have P2602, but now he has P2107 MOD PC solenoid electrical problem. This tells me that either the solenoid has an intermittent open /internal short, or the solenoid is not making good contacts to the conductor plate, or some problem with wiring / connectivity issue in between the TCM and the soleonid. That is probably why he is having jumping voltages all over the place.

Easiest way to find it is to measure the resistance between pins 38 and 36, and make sure it is within specs (plus some small (0.5 ohm for example) addition of wiring resistance). If the resistance value changes intermittently, then there is b some bad connection somewhere.

 

[MAVA]

Yea, but measuring pulsating-DC is a crap shot with any DVM as all of them have a bad roll-off reading frequency on the DC scale. This is basics-101 in using a DVM.

The best is scope the signal to see if the signal is present. I know one of the regulation solenoids goes from a frequency of around 50hz to 2kHz(and a 30% to 50% duty cycle). I had a chart showing the frequency, but not sure. The lowest frequency is in reverse(I think). I was looking in my bag of documents, and I cannot find that chart.

For all practical purposes the "drive transistor" is shorted ON for that volt reading

I assume you live in the UK as here in Amerika we donot have any kind of ML270-Da AmeriCANs want all the Horse Power you can give them, so the ML 320's were the smallest found here... The ML270 is powered by a m112? or M104? or something else?

What is the outcome now? You are in full limp mode of reverse and second gear, or do you not get that at all?

I would even swap the two black solenoids around...

Any junk yards in your area? I would even get a harness and some solenoids to play with as measuring the signal on DC volt scale is silly. Maybe your DVM has a frequency scale?

Martin