CaliBenzDriver

I thought we did trace the schematic to realize F25 was only feeding ECU based on VIN options.

Infernally we witnessed the charcoaled leaky capacitor pin that burned the board beyong repair.
This failure is squarely resting on capacitor failure, right?

Do you think an unrelated circuit could toast the F25 path ? In reality many factors get to contribute chaos synergy.

- I am going to strongly recommend CLEANING THE PAINTED GND INVOLVED with F25 circuit: W16/5...
- Call that crossing my T's : we have seen oxidized W16/5, how is yours?
- When W16/5 plays out it's like ECU power going on/off on/off on/off and all caps paying for it.

Today I've tried to spray shoot my nasty W16/5 with contact cleaner... not easy reach to the firewall.

I have familly visting from overseas for 15 days so lots of test driving. When the engine runs with quiet injectors tic-tic, its intake does not stink like giant oil leak - Overheated burnt oil fladhed through PCV is related to ECU workabilities. Hello people with melted piston No1: ECU screwed oil pressure control much like it can screw ALT YoYo !!!
✌️

S-Prihadi

Albeit M connector installed is not contributing to fuse 25 blow out........... but :
Peter have not done inspecting possible short to ground each suspected wires out of F connector where F25 is at.
Example pin F-17 goes to fuel pump controller N118, we do not know if this pin F-17 is powered from Fuse 25 or Fuse 6 or Fuse 27 ?
Accelerator pedal pin F-10, F23 and F-11 is another power consumer wired to F connector which we also do not know which 3 of those fuses are powering it.

By cable size and number of pins and the bigger SLK2.8 class size terminal for Fuse 25 at 15 amps, I doubt it is only for powering the processors,
the must be some other "consumer/s" which is/are at the engine.

Soon I will also amp clamp my fuse 25, I am curious what is the actual amperage it is powering ?
Also why in M276 3.5NA and M276 3.0 , M278 and M157 and now M273 ............fuse 25 is not shown at engine computer wiring diagram, one has to trace it to Front SAM and is always from connector 4i pin 10 of Front SAM.
It seems Bosch and MB has certain wiring topology they keep using, which is good for us as we get more familiar with ease.
The only difference at M273 is, Fuse 25 get only 2 of fat SLK 2.8 terminals at Pin 3 and Pin 5 of F connector.
All other engines I mentioned, get 3 of fat SLK2.8 terminals at Pin 1, 2 and 3 of F connector. Maybe because M273 is not yet a direct injection engine ?

S-Prihadi

My comments in green

T100T

Thank you for the replies, let me add some details to what has happened, reply to some comments and give an update on what's going on.

More details on what happened when tested, the ECU board was in a polythene bag with the F connector attached, a new fuse was installed in f25, Xentry was up and running and I turned on the ignition. I had checked the resistance from the load side of f25 to ground, it was ~3.5ohms piror to reconnecting the battery. I have to say I was confident, too confident - what resulted was additional burning on the ECU in the area of the capacitor, quite a bit given there was f25 in the line. At this point there's a larger burned area where the cap was and the input resistance (pins 3/5 to 2/4/6) is low. I don’t think I can get the board to talk again. My conclusion is that, as stated by others, the blown cap did more damage than appeared.

I'm not frustrated with MB, the ECU was probably designed 15+ years ago; for sure the heat sink on the cap could be improved but there's no way to fully replicate real life use, mine has lasted 14 years and approx. 6,500 hours of operation. I do enjoy driving this car and just prefer not to spend a few thousand dollars if it can be fixed for a few hundred.

A question, CaliBenzDriver in post #38 you mentioned that ECU covers are available, do you have a link? I saw some covers on Alibaba but not for the A273 900 0700.

Lesson learned, If making repairs/testing an ECU where you're unsure of its state then power with a bench supply and limit the current.

I did contact an ECU expert who told me there are three or four chips that physically would need to be moved to clone the ECU, unfortunately one or two of those are BGA (see attached photo) and to make matters worse they use high temp solder. So I sent my ECU to said expert and he's going to try and get power back to the BGA chips then he can download the data, he will not try to remove the BGA devices. Wish us luck, the pins/sockets arrived, I'll take the time to look into the intermittent low oil and various other error messages.

thanks everyone,

Peter

CaliBenzDriver

Peter, I am pleased to see your issue is moving along well towards a refurbished ECU transfer.

Needle in a hay stack... I can't locate the shop that sells replacement covers. I spent couple hours yesterday searching. I kept seeing it when I had no intetest and now it's gone - I'll update when I find source.
It turns out actually here, you don't really want to buy a fresh cover part because there's more to electronic module mechanicals than meet the eyes.

Bosch is pushing the envelope of the greatest electronics architecture by stressing its systems with the worst environmentals. Dedicated mixed engineering teams work together to manage gentle failures. Meaning when you only fix leaky caps expect a less well controllable failure is going to byte you.... Uncle Murphy!

The ECU location is extremely severe for electronics so work inside the ECU enclosure needs to meet that challenge. You don't want the used replacement to fail shortly.

Let the shop dealing with your data transfer also deal with refurbishing and buttoning up your new ECU:

• heatsink compound !!!
• new enclosure cover
• case seal integrity
• capacitors kit replacement
• next designed failure (coil/injector driver chip?)

You need the above steps done 100% properly with no detail left to chance. Whoever does this refurb process will provide you with some type of guarantee. Clearly that is to say: I would not personally refurb my own ECU but find a 3rd party provider.

Hopefully your cloned ECU should recognize your existing keys out of the box because all the specifics are moved over.
​​​​🤞

MBNUT1

Thank you for taking us through this process.

MBNUT1

What is the purpose of the fence?

T100T

I believe it is cooling, having said that the plastic cover on the cap doesn't help. It might also be mechanical support for shock and vibration.

Left Coast Geek

I would suspect that aluminum 'box' is electrical shielding.

CaliBenzDriver

finger shape is a spring loaded holder


Holder cage has 4 opposing fingers - Look at rainbow color oxidation on pin connector...

The fence around power supply filtering capacitor is a mechanical holder to keep the engine vibrations from rocking the tall cap loose.
Normally some type of glued fastening is used but under extreme conditions even BASF advanced glues did not stand (aerospace glue $$$).

The goal of mechanical holder is to preserve the cap solder joints from failing. The lead-free ROHS solder is weaker than the old leaded favorite.

> New cap sizing:
The current cap holder solution forces any cap replacement to be of matching diameter to work.


> PLAN B:
The proven reliable solution to implement power filtering caps is to mechanically fasten capacitors and connect them with thick wires crimped or soldered.
These caps can not be downsized because the microFarad value needs to be the highest and the rated voltage fair to provide low impedance against spikes and riples.


> ECU as noise purveyor :
Clearing most of the noise out from the power is essential before 5V/3.3V voltage regulation for the digital processors.

Here we can expect the ECU ignition/injection/VVT coils to be the source of noise for other modules a distance away. Meaning caps are really exposed to the worst conditions both external and internal.


> Path from Positive to Negative return:
Heavy load circuits are directly powered by a F-SAM's fuse and ECU manages power from the GND side.
Meaning you get all the spikes on GND side that best should be low-resistance path to reduce spikes.
The positive-side spikes are in the SAM away from the ECU undersized cap. Meaning the F-SAM better have it's own local army of filtering caps.


++++ Types of filtering caps....


tantalums are YLW, others are known as ceramics like .1uF
Tantalum caps can go up 100uF are some of the best dry caps available but can not rival with 3300uF/25 low-ESR. Meaning each type of cap is good at doing only one job due to technical limitations.

There are multiple ways to deal with electrical noise.
-- The most effective is filtering nearest to the source. A double-choke wired in X with caps is the best at creating a great stop point - I am surprised we don't see that! (choke-coils can vibrate themselves loose).

-- Next is physical distanciation of unlike signals. Keep power and signals apart. Shield the sensitive lines.
The winning solution is a mix of all counter measures where noise has been confirmed to be an issue (3Ph LPFP) - You should see the army of caps before/after I was there. I through in couple larger caps I had laying around.


> ECU Rainbow connector :
I wonder if the pins contact area have an electro-plated coating or does it look like rainbow oxide too?


+++ MB cap solutions :

LPFP: glued caps and choke coil


Blower motor driver: NICE! Caps and coil are captured - Honest circuit


trouble around corner: backseat 110V module parts not glued 🙂

S-Prihadi

The "rainbow" looking connector male pins could be overheating sign. Maybe heat gun too hot when trying to undo the cover glue ?
https://www.basketofblue.com/how-to-...ors-on-copper/

MBNUT1

Thanks for the answer on the fence.

T100T

Re heat causing discoloration of pins. I only applied heat when removing the board from the lower, diecast, part, heat was not applied to the plastic connector.

My damaged ECU is presently in the mail to an expert, when I hear something I'll post an update.

If you have an aging Mercedes would you spend the money to download data from the ECU as a preventative measure, then should it fail you will be able to buy a used ECU and be up and running without too much hassle?

Peter

CaliBenzDriver

You're right, most us really like the idea of being prepared for known module failures - What can we do...?

1-- Download a backup image of the ECU while it is 100% functional. I think Xentry can help you do that.

2-- To accept a cross-flash, a used ECU coded to a donor VIN need to be virginized. This is done routinely by repair shop with bootlegging toolkit.

3-- Upload your ECU image into virgin donor ECU.
I think Xentry can help you do that.

This procedure is trying to avoid swapping the tiny storage chip from both ECU boards - Some people like this quick guaranteed way best, the hardware guys that is.
The software guys want to fork data files around and not even know what any of the hardware is doing. In between hardware and software there is a no-men's-land... that's where I like to camp.
Best views on either sides...:
Xentry integrated software suite is build to swap modules that are built to spin business. You can bank on Xentry making the process painless with an active On-line SCN access... smooth & well scripted MB procedures. 👍


Here is what I believe...:
- the chip trick is quickest since the ECU BOX already opened for a mandatory "Filtering Cap" rework...

- the software copy is the long route sometimes virtuous. It definitely sounds like a guy's way for a great hacking adventure.

I know a diver who fixes all his gear .

Left Coast Geek

I think what that vendor claimed to be able to do would not only copy any 'coding', but also copy the 'flashware' (what the rest of the world calls firmware)... dealers charge $$$ to update flashware and I don't believe indies can do it at all

juanmor40

Very likely. For example, I do not do backups of my disks, I "image/clone" them. Backup software has logic, manages files, compresses/filters them and is error-prone in my experience. An image is "in theory" an identical copy of its now twin. Something goes wrong with the original, and it is plug and play and only loose the "delta" between the dates, i.e. a very clear known/deterministic quantity.

Similarly, decades ago even with my limited electronic/electrical knowledge I was able to read "flashware" from a given chip of interest with borrowed electronic gear, and I was able to clone a few copies for other uses. As long as the cloned component does not depends on something else also coded in the old hardware, it should work flawlessly on the new hardware.

S-Prihadi

Peter,

Why not ask BenzNinja what he can do. I believe BenzNinja can do A LOT and more, otherwise I would not have paid for his lifetime service.

Today I backed up my car using Autel, but honestly I do not know to the full definition of Autel's BACK-UP for MB cars
By the looks of it it is all relevant "data" where I can swap a new module and place my old data to it.
However, I do not know what if the module in question is a used module ........... which I read some will need to be de-virginize first or make "empty", before I can put in it a new car data which is my car's data.

I read some module is a 1 time use only, so once it is married to a car, no way to use it on other car. VAG cars some modules I read is that way.















After all modules READING SUCCEEDED it will later become Backup Succeeded.


I have 47 modules in total but 10 are LIN and these LINs are sub-module of another module, that means out of 47 less 10 = 37
and now less 29 = 8 modules which some could be Autel un-able to back-up or plug and play module not needing special coding ?

VERY Important modules not backed-up by Autel :

01. A80 Shift Module
02. N80 Steering Column Module
03. N68 Electric Power Steering
I must check with WIS if these 3 components need coding or not when new one installed. I believe A80 and N80 will need to be coded.

On a W205 of year 2020, N68 Electric Power Steering need to be coded and once coded, it can not be used on other W205, the status will be LOCKED.
I hope W212 is not like W205.

Start at 4:20



============

Not so important modules not backed-up by Autel :
04. A2 Radio head unit or A40/8 Audio/COMAND display (ZAN)
05. A40/9 Operating unit of control unit 'Audio or COMAND' ( COUS [ZBE])
06. A40/4 DVD Player ( at my rear seat )
07. A40/5 Display ini left rear passenger compartment ( baby sized TV at left front seat head rest mount )
08. A40/5 Display in right rear passenger compartment ( baby sized TV at right front seat head rest mount )





.

MBNUT1

"On a W205 of year 2020, N68 Electric Power Steering need to be coded and once coded, it can not be used on other W205, the status will be LOCKED.
I hope W212 is not like W205."


Is there a legitimate reason for that?

juanmor40

there are electronic components that are one-time write only, and there are multiple writes type as well. Most modules contain the latter; otherwise, they could not be updated.

Question is if MB modules ( designed by others) have both types internally to store one time data, marrying/pairing. The alternative, for cost reduction, is to disable the access to that region of the module memory in software to look one time only.for us mortals.

Those immortals in software and electronics know how to virginize a component regardless of protection.

MBNUT1

Can a power steering unit from from one W205 be removed and run in another?

S-Prihadi

This video shows ME9.7 being virginized ( data wipe ) and then inserted current VIN yada yada...... = work well.
Using Autel, but some process need bench work , but no need up to removing EEPROM and the like....simply ECM out of the car and direct wired to its High and Low CAN BUS, this way no involvement of GateWay N93 at Front SAM.
This Autel version used is a special model IM608 for this kind of programming work and not focused as a scanner, albeit it can be one too.

NOT CHEAP. US$3K ish
https://www.autel.com/immotool1/3250.jhtml

Some programming tool may not work without current/running subscription, because some car's data is at Autel server HQ and not inside the programming device.
I think 1 year is like US$400 or so

T100T

All,

So the update is that all data has been downloaded from the ECU, the "expert" had to clean up the damaged area to remove the short and then was able to download the data using BDM (background debug mode - this sidesteps the data protection). I have been away from this for July 4th but will get back to checking other sensor errors, check grounds, etc as suggested by others.

Hopefully this will all work and I'll get the car running again, at that point I'll create a new post with a more appropriate title.

thanks again to everyone who contributed,

Peter

Arrie

Just curious, but what would a new ECM cost?

Anybody here know?

juanmor40

https://mbparts.mbusa.com/oem-parts/...Y4LWdhcw%3D%3D

plus coding labor costs.

Edit: I updated the link for the 2010 E550.. $2000+

Personally, until the source of the short is not understood, I would not install a brand-new unit on that vehicle at that price. I rather damage the old ECU one more time before committing a new unit, i.e. use the old ECU as an expensive fuse.

Left Coast Geek

yeah, add about 2 hours shop time for the flashware and coding to the parts cost, and a dealer likely would charge more for the part, which is undoubtedly special order.so yeah, no 'sales price'.