** Hi all, please bear with me as this is a long and fairly complex issue that has evolved over a long period of time, and I’m trying to provide as much technical detail and diagnostic information as possible to avoid random guessing or parts swapping and so the issue can hopefully be narrowed down properly. **

I’m dealing with a major no-crank/no-start issue on my W221 S-Class, chassis 221.071 with the M273 engine and BOSCH ME9.7 ECU. The car has been parked for a while now (~1 month), and I’m trying to get it running again. The issue has developed into a situation where the car powers up, shows multiple drivetrain/ESP/battery faults on the dash, displays a red battery warning icon, and will not crank or start. The starter itself is not the main suspect because it has previously been proven capable of operating when jumped manually. The problem appears to be something disabling crank authorization or preventing the engine ECU/drivetrain network from waking and communicating properly.

The dash currently shows multiple warnings including ABS/ESP inoperative, run-flat indicator inoperative, visit workshop without shifting gears, and the red battery warning icon. On the gear selector display, there are no boxes/highlights around P R N D at all. The gear display appears “dead” or inactive. Through my ODB2 scanner, I am unable to communicate with the ME/engine ECU, transmission module/EGS, or the ISM module while installed in the vehicle. However, most other vehicle systems/modules are still accessible through Xentry and scanner.

The ISM module itself has already been replaced with another unit previously, but I was unable to proceed with proper teach-in/adaptation because communication with the ISM was not possible. Since the ECU, EGS/transmission, and ISM are all simultaneously offline/inaccessible in the vehicle, I suspected a drivetrain CAN, wakeup, Terminal 87, or front SAM-related issue rather than three independent module failures. One interesting detail is that the ISM does make spinning/motorized sounds when the key is removed from the EIS and also when locking the vehicle via key fob for a few seconds, which suggests it is at least partially alive and reacting to vehicle state changes. However, despite those sounds, the gear selector display remains inactive and the module is still inaccessible through diagnostics while installed in the car.

The original suspicion was the ME9.7 ECU because of very strange behavior observed both in the car and on the bench. With the original ECU installed in the vehicle, communication with drivetrain modules was missing. The ECU appeared to affect the CAN bus when plugged in. Previously, with the ECU disconnected, CAN readings looked more stable, around the normal 60R termination (±0.5R) and normal CAN voltages. With the ECU connected, drivetrain CAN voltages became skewed, and multiple modules such as the engine ECU, transmission control unit, and ISM appeared offline or inaccessible. However, after obtaining a donor ECU, it behaved properly on the bench but still disappeared when installed in the vehicle, I am now confused as to what module/power feed/wakeup path/CAN branch in the car may be collapsing the network or preventing proper ECU initialization in-car. More on this later on.

To investigate the CAN and power distribution side more deeply, I stripped a large portion of the interior and accessed drivetrain CAN splice areas and related harnessing. I followed wiring manually using WIS diagrams, tracing lines between the ECU, front SAM, drivetrain splice points, X26 connectors, and related modules. I performed extensive continuity testing between multiple harness sections and connectors, especially around drivetrain CAN and Terminal 87/wakeup-related wiring. I verified continuity through several front SAM paths and checked multiple fuse and relay related circuits. Resistance across drivetrain CAN was generally around the expected 60R, but voltages on drivetrain and chassis CAN-L and CAN-H changed depending on what modules were connected. Another important detail is that while measuring drivetrain CAN directly on the ECU harness connectors with the ECU connected in the vehicle, CAN-H and CAN-L did not appear to sit at normal idle behavior around ~2.5V each with proper activity. Instead, the lines appeared skewed/statically biased, often with little or no visible CAN activity. Interestingly, CAN behavior appeared more stable once the ECU was disconnected, which was one of the original reasons I suspected the ECU may have been disturbing the drivetrain CAN network.

I also performed voltage injection tests using a power probe into specific lines and connector sleeves related to Terminal 87 and ECU wakeup feeds. In several cases, injecting voltage caused temporary behavioral changes in the car and network. Some lines that should have been active (in theory) showed 0V until externally energized, which even then didn’t allow the car to crank/start. Certain wakeup-related lines around X26 and Terminal 87 M1i behaved inconsistently, which increased suspicion toward a power distribution or wakeup/control issue rather than a simple dead ECU. I traced continuity between front SAM sections and engine harness paths, and verified several large pre-fuse and busbar connections. The issue became especially confusing because some circuits appeared electrically correct during continuity testing but the ECU and drivetrain network still failed to wake or communicate reliably in the car. The engine is also not locked up as I was able to manually turn it over using the crank pulley.

On the bench, the original ME9.7 ECU could communicate with Xentry, it was identified correctly, but it showed an internal battery voltage/ADC fault. Xentry reported battery voltage inside the ECU at around 1.6–1.7 V even though the bench power supply was feeding the ECU with 12–13 V. There was also a current/stored fault related to engine electronics voltage supply/battery voltage too low for ADC, P0607. Because of that, I suspected an internal ECU power/reference rail issue. I also heard a high-pitched squealing or electronic noise when the original ECU was installed in the car and with key at position 2. That sound was not present on the bench, which made me think the ECU may have been failing only under vehicle load.

I then opened and inspected the original ECU and found a slightly leaking blue electrolytic capacitor (in its beginning stages) near the orange connector/busbar area, marked 470µF 35V, but not catastrophic failure-level. Cleaning the area with IPA seemed to make more fluid appear, which made the capacitor suspicious. No board damage was observed, at least from the top and bottom of the ECU, neither was there blackened sections near the capacitor like you’d see on some ME9.7 ECUs with severe damage from the electrolytic capacitor fluid which is corrosive and destructive. The power IC would overheat within 15 seconds of powering on the ECU on the bench to the point of burning your finger, all of which resoldering the IC's legs fixed, probably due to poor contact between the legs and pads or had cracked solder joints from heat cycles and age. I also found that certain internal rails seemed stuck around 1.7 V. Pin 5/wake or relay-control-related circuitry also measured around 1.7 V repeatedly. Touching a fresh discharged capacitor to the removed blue capacitor pad momentarily rose that rail to around 3 V before it decayed again. Some ICs and driver circuits showed related voltages around that same abnormal range. This made me strongly suspect the original ECU had an internal rail/reference fault or power management fault.

Because of that, like I previously mentioned, I bought a used donor ECU with exact matching numbers. I inspected the donor externally before installing it. It looks relatively clean for a used module, and most importantly, there are no obvious signs of prying, lid deformation, tampering, localized heat spots, corrosion, burn marks, or bad smells. Before installing the donor ECU in the car, I tested it on the bench, the donor ECU communicated properly on the bench, Xentry could enter the ME9.7 control unit, read control unit version, live values, fault codes, event memory, and drive authorization values. Current draw looked stable and logical. With the ECU ON and idling on the bench, current draw was about 0.510A on the PSU. With only Xentry connected, current rose to about 0.560–0.570 A. And with only the scanner connected, current draw was around 0.980–0.990 A. There were no abnormal smells, no obvious overheating, and no aggressive noises on the bench.

The donor ECU showed typical bench faults, which are expected because the ECU was powered alone without the rest of the car connected. Faults included open circuit faults for oxygen sensor heater circuits, accelerator pedal sensor fault, coolant temperature sensor open/short to positive, pressure sensor open/short to positive, and event memory faults for no CAN message from transmission and traction systems. These made sense on the bench because those components and modules were not connected. Drive authorization data showed initialized, transport protection detached, personalized, and activated as YES, and start enable showed YES in the ECU’s own data. The donor also showed the same 1.6V battery voltage reading in Xentry even though the PSU was set to 12V. Since this was the second ECU to show that reading while otherwise communicating normally on the bench, I now think that specific 1.6V value may not be a reliable raw supply voltage reading in a bench setup, or may depend on missing vehicle conditions.

After the donor ECU passed bench testing, I installed it into the car. But unfortunately (and with a heavy heart), the car still showed the same dash faults, including the red battery warning icon, and still had no crank/no start. The battery also seemed to lose charge quickly even with a 25A charger connected (which FYI I later connected a jump starter just to see if I get any signs of life but to no avail). I then tried to communicate with the engine ECU through Xentry while installed in the car, but I could not enter the ME module in the vehicle.

This is the big issue now: The donor ECU communicates normally on the bench but becomes inaccessible when installed in the car. That makes me think the problem may not be the ECU itself anymore, but something in the car preventing the ECU from waking, powering properly, or communicating on CAN.

The current working theory is that something in the vehicle-side infrastructure is disabling crank or killing ECU communication, or another module dragging the drivetrain CAN bus down. Earlier in diagnosis, Terminal 87-related behavior was already suspicious. I had found weird behavior around terminal 87 M1i / X26 wiring, including circuits that had 0V unless power was injected. Injecting power in some areas caused temporary changes, suggesting a wakeup or supply path problem. I also previously checked pre-fuse and front SAM-related wiring and found odd symptoms around power distribution.

At this point, both ECUs communicate on the bench, but neither is accessible in the vehicle. The original ECU may still have internal issues beyond surface level, perhaps deeper in the multi-layer board, but the fact that the matching donor behaves the same way in-car strongly suggests there is also a vehicle-side issue. The car still has no crank, no start, dash faults, red battery icon, fast battery drain, and ME inaccessible through Xentry or OBD2 scanner when installed. The front-end is dependent on whatever feed or wakeup path is failing there.

I have started checking/replacing fuses in the front SAM area to establish a clean baseline but obviously throwing parts at the car is not the solution here especially on a car like the W221 which is essentially a rolling computer network on wheels.

Also, small backstory, the issue first began when I drove the car and parked it for ~1 hour, then came back and the car would start briefly and shut off on its own, accelerator pedal input was ignored for some reason, and the starts where really soft starts, not the known aggressive powerful jump at startup. Then after a few times it would not start, only crank. Got the car towed, and while diagnosing, came to turn ignition ON and the dash lit up like a Christmas tree and lost all comms to drivetrain CAN. The scanner will not automatically identify or enter the vehicle as it normally would. Instead, I have to manually select the chassis/model and manually navigate into the systems menu in an attempt to access modules individually.

What I urgently need help with now is narrowing down the exact part of the W221’s start authorization/power distribution/drivetrain CAN chain or honestly ANYTHING that can realistically cause this combination of symptoms because at this point, I’m genuinely starting to lose my mind with this car and I urgently need it back on the road. I’ve gone far beyond basic diagnostics and random parts swapping. I’ve stripped interior sections, traced wiring through WIS diagrams, accessed drivetrain CAN splice points, performed continuity testing across harness sections, monitored CAN voltages under different module configurations, injected voltage into suspected wakeup/Terminal 87 circuits, replaced the ISM, bench-tested both original and donor ME9.7 ECUs, monitored current draw behavior, inspected the original ECU internally down to component level, and verified multiple fuse/power distribution paths. Despite all of this, the core issue remains: the drivetrain side of the car appears offline, the ME/EGS/ISM remain inaccessible in the vehicle, the PRND display is dead, the car throws multiple drivetrain/battery/ESP faults, and there is still absolutely no crank/no start. From my earlier EIS/EZS readings, EZS/EIS turn enable issued = YES, A80 ISM start enable = NO, ME (engine ECU) start enable = NO. And there were also drivetrain CAN timeout indications and EIS reporting ISM not communicating.

And for the record, the ECU, EGS, and ISM are all part of the drivetrain CAN/start authorization chain and are heavily interdependent during wakeup and operation. In practice, if one critical part of that chain goes down, whether it’s CAN communication, wakeup power, gateway communication, or one of the modules themselves, all three can appear offline simultaneously. And the Central Gateway Unit (ZGW) is the bridge that lets diagnostics and the rest of the car “see” them.

Any detailed guidance, measurements, test procedures, or known failure points would be hugely appreciated because I’m trying to avoid throwing more modules at the car and instead finally isolate the root cause properly.

 

so much going on this is not going to be fun

I would hazards a guess too many modules messed with - all the immobilisation / theft protection elements are so muddled nothing is happy to communicate

I don't think you normally see 12v in an engine ECU 1 to 5 volts is what I'd expect - put as much back to standard and one by one get each replacement part correctly coded / cloned / communicating correctly

 

Hello,
Thanks for your reply.

I understand what you mean about internal ECU voltages, but some ECU pins/railswill be 5V reference, sensor supply, logic supply, etc... However, the ME9.7 main power supply at the connector is still 12V battery supply and ignition-switched 12V. The ECU then regulates that internally down to 5V/3.3V/lower logic rails.

So my concern was not that every ECU pin should show 12V, my concern was with the ECU’s main supply/wakeup behavior and the fact that the ECU was reporting an abnormal internal battery/ADC voltage despite being supplied correctly on the bench.

Also, the original ECU and original ISM already had the same no-crank/drivetrain communication issue before any replacement modules or adaptation attempts were introduced. So while mismatched or unadapted modules may add another layer of complication, they do not explain the original failure where the drivetrain CAN side, PRND display, and ME/EGS/ISM communication progressively going offline on their own. Neither the cascade of events from starting and dying on its own, then only crank, and now no crank no start.

 

Your original post is way too long and detailed with all voltages etc. for me to read, but one thing I KNOW is that all those modules DO NOT fail at the same time!!!

The problem is something that the modules have in common. You seem to have checked a lot around the wiring already not finding the issue. What about the CGW (ZGW)? I don't know what kind of CGW your car has but mine is (like 2013 model has) a plastic 1" x 4" x 4" box mounted in the end of the dash on driver's side together with the AUX battery.

I had multiple module failures, and my problem was the unsoldered pins on the CGW board, the very same pins that the wire harness connects to.

CGW is the one connection point that connects with all other modules and problems with it can/will cause issues with multiple of the other modules.

You have Xentry so if you are member of ninja group go get you a new CGW (about $150 if I remember correctly as I bought one) and have ninja code it for you unless you know how to do it yourself.
.

 

Hello,
Thanks for your reply.

I absolutely agree with your main point that modules generally DO NOT all fail simultaneously on their own. That’s actually one of the reasons I started suspecting a shared/common link issue fairly early on rather than assuming ME, EGS, and ISM all independently failed together.

The CGW/ZGW was considered during diagnosis for the exact reason you mentioned, it is effectively the bridge between the drivetrain CAN side and the rest of the vehicle/diagnostic systems. I also noticed that despite ME/EGS/ISM being inaccessible, many other modules in the vehicle STILL communicate normally, which does point toward either a gateway/drivetrain CAN issue or a shared wakeup/power issue affecting that side of the network specifically.

I also opened and visually inspected the CGW board earlier on looking for obvious issues such as burnt areas, corrosion, cracked solder joints, lifted pins, etc., but nothing immediately stood out visually. That said, I completely agree that an intermittent/internal CGW issue is still possible and will investigate further, especially considering the amount of other diagnostics already performed. I appreciate your input on this matter.

 

OP EIS failures are common. That you are not getting a command to crank implies as much. TRP... 'Theft Related Part' says it must come from MB and they must program it to the car.

When you hear hoof beats... Think Horse... Not Zebra. Or look for pattern failures first. EIS and ISM failures are right at the top of the list.

I assume you are using a burly battery maintainer as you are doing testing. Low voltage will introduce its own extensive list of disasters.

 

Yeah I'm not reading that.

Replace the battery.

 

yes I also have attention deficit disorder

 

Hello,
Thanks for your input. I agree that EIS and ISM are common failure points on these cars, and I’m definitely not ruling them out blindly. However, one of the reasons I’m not immediately blaming the EIS is because of my earlier EIS/EZS actual values. From what I previously recorded on my scanner, EZS/EIS turn enable issued = YES, while A80 ISM start enable = NO and ME start enable = NO. There were also drivetrain CAN timeout indications and the EIS was reporting that the ISM was not communicating.

That is what makes me question whether the EIS is truly the root cause here. If the EIS is accepting the key and issuing turn enable, but then reports that the ISM is not communicating and the downstream start enables remain NO, wouldn’t that point more toward the drivetrain CAN/wakeup/power side of the chain rather than the EIS itself? I am using external charging support during testing, while trying to keep it controlled as much as possible such that I don't get undervoltage chaos as you mentioned.

 

Hello,
I understand the post is long, but it is long because a large amount of diagnostic work has already been performed and documented to avoid random guessing and unnecessary parts replacement.

As for the battery suggestion, I’m curious about the reasoning behind it in this specific case. The car has been under external charging support during diagnostics, so I’m trying to understand how the battery alone would explain the sudden drivetrain CAN communication loss, ME/EGS/ISM simultaneously becoming inaccessible, dead PRND display, and the progression from starting and dying on its own, to crank-only, and eventually complete no-crank/no-start.

 

Because the batteries are dead.

 

W221 is so very sensitive to voltage that if your battery or batteries is stale, you will get all kinds of crazy codes. Do what Senecat says. Start with a new battery and use a charger when you are working on the car.

 

OP if you are using Xentry be sure to have it test the batteries. It will let you know if it approves. The little one in the dash makes all manner of hell when it is iffy.

 

Simply...

You need the key, EIS, TCM, ISM to also match the engine ECU. They all need to be from the same car. It won't start as long as you have a hole in your a** if they don't match.

The alternative is to buy a new, virgin engine ECU from Mercedes and have it coded to your car. This is not possible with a used unit.

 

Keeping in mind we’ve seen enough MU’s pack it in that I’d not suggest going the used route.

You have a membership with Peter? He is very helpful.

 

Senecat, vettebk, and JohnLane, the battery was temporarily replaced with a new one and the exact same symptoms remained.

 

I’m aware that the car won't start as long as you have a hole in your a** if they don't match, the problem is I’m diagnosing a CAN/communication issue first not asking why a random unmatched ECU won’t start the car.

 

Hello,
I don't have a membership with him, where can I get one?

 

utter genius

https://benzninja.com/

 

Don't you have two batteries? (technically 3, but we're not worried about the dash battery)

 

Yes, technically 3 batteries as you mentioned, G1 on-board electrical system battery at the rear, G1/4 starter battery in the engine bay, and the small AUX battery.

The starter battery was temporarily replaced and the rear battery was externally supported using another battery, yet nothing changed.

Front SAM is the main suspect now considering its role in wakeup/Terminal 87 control and power distribution to the drivetrain-side systems, despite having checked its circuit board, if corrupted messages were being sent by the MCU, you wouldn't be able to tell unless you have the proper tools.

 

The front SAM can be swapped out for a known good used one. It will need coding eventually to match the config of your car, but it's not VIN coded from a security perspective.

It just needs to be correct to the MY.

 

That’s what I was thinking as well. At this point I might as well try a known good used front SAM just to remove it from the equation and see whether drivetrain communication/start authorization behavior changes at all.

 

I also think there is a problem with your CGW.

 

Whilst you're swapping SAMs investigate both the box and the area the box resides in for evidence of water ingress.

This will invariably end up at the CanBus connection blocks under the front seats.

CGW is as swappable as the SAMs.