I've been having some small and not that bothersome issues related to car connectivity.

• First, the doors don't always unlock when I touch the handle and even more rarely, they sometimes don't lock.
• Sometimes the key isn't found when I'm in the car until I hit unlock on the keyfob.
• The car sometimes thinks I'm about to crash and gives me the "HIT THE BREAKS" warning.
• The rain sense wipers can't stay on all the time, or they will be wiping sporadically.
• This morning after starting, I got a warning that my blind spot warning system was disabled (but it worked during the drive).

Finally, the second start of the day is sometimes weak and very rarely needs a second attempt. Last night was around 15F and the first start of the day was strong, but that second start was so weak that I couldn't start the car. No amount of restarts would get it going. Luckily, it's parked really close to my work so I'm just going to leave it there until later today. I'm pretty sure it's going to start fine when it cools off.

I suppose it's time to get out the soldering iron. Cali, what order do you suggest one go through the modules to take care of the solderless pins?

 

As expected, I let the car cool off and it started immediately. I then drove the car a short distance such that the engine was not fully warmed up, and it started again immediately.

@CaliBenzDriver has described problems related to the start/stop system and the auxiliary battery which are the result of poor data connections which can result in strange behaviors with the auxiliary battery. I'm going to wait for him to chime in.

Yes, there is a chance that the auxiliary battery is just not holding a charge, and I might just need to replace it, but I also want to fix the crappy solderless connections which are causing, or at least not helping, with the other issues on my list.

 

CT! you're now dealing with the genuine bugs skillfully planted by the design team.

From your description right now the top priority is poor voltage. That helps upset the solderless chaos.... self-feeding Gremlins!

The theme is "one thing leads to another bigger one". The goal is to seek the hidden root cause to cut down the chaos wasting good ownership.

> Getting a leg up:
1- Very FIRST is... clean funky MAIN GND STRAP!

2- Next is battery CHARGED up

3- REBOOT then scan chassis for all faults

4- Sanitize CAN-B: saves battery
EIS, SCM, Keyless + DCU's + AAC +

5- Santize CAN-C: gets performance
ESP, ISM, ...

6-- ALT Self-managed Voltage:
Unplug LIN control to prevent voltage yoyo wasting your time right now. (Courtesy of R-SAM + CAN-B + CGW).

The castle or keystone is CGW that's purposely assaulted by junk signals, mostly CAN-B.

CGW tanks bandwidth and soft-crashes network... until Restarted or Rebooted.


> CAN OF WORMS...
Last week I found out a special twist about CAN-B: part of it works off SAM, part of it work down long daisy-chained 60R bus-bars extensions.
The furthest away has the worst connection...
- I am going to poke around that issue with shielded pairs to see.


We'll help you get your W212 to perform reliably without chaos.

+++ CAN bus network:
Differential connections rely on stable voltage environment.

Having noisy spiky unstable voltage in rocky poor connections is like "melting ice in the sun"... ouch!

Your best bet is to cancel all forms of disturbance implemented by MB:

1. Low voltage
2. Poor connections
3. glitchy voltage
4. undersized wiring for power /signal
5. marginal noise filtration
6. oxidized/oiled connectors
7. ...

These weaknesses lead to strength when turned around.

 

If it is engine temperature related a couple things to check first before going to the electronic gremlins' world:

-The crank shaft position sensor can be heat sensitive not allowing the car to start but after running it can let it run, until you turn engine off and try star again when hot.

-The starter can be shorting internally when hot. If you have the Auto shut-off/start function, I would really look into this.

When MB introduced the auto shut-off feature they claimed the starter was made stronger for taking the repeated starts. I think this is one of the biggest BS statements they ever made. Check your starter if your car has this feature and you have been using it.

 

The only one of those that I can see being true is the starter shorting when hot, but the starter does not always do this. It only does it when the engine is hot enough and conditions are right that the engine treats it more like a restart than a fresh start, if that makes sense. That's why I suspect it has to do with the start/stop, aux battery, or a related system.

Even though I drove the car until it was at temp, the temperature outside was extremely cold today. I don't believe the starter was unduly hot today, but the aux battery was unduly cold...

 

@CaliBenzDriver has spent a lot of energy for all those discovery's and maybe he already thought to use a CAN Sniffer to monitor the CAN signals, at a byte level.

Mostly he is right, luckily all those can't be applied on automotive engineering, as besides laws and regulation, some of the best engineers do their job at the highest level possible, they always add "error correction" ...and a lot of HW/SW standards.

Of course, there are many bugs and maybe wrongdoings, but after all...when you steer to the right the car still goes to the right and steering angle sensor declare the correct direction and angle value to the ESP and longitudinal sensor as well...and all data is calculated in real time so you can continue your safe drive...

The best is the enemy of the good!

Solderless, missing CAN signals, voltage fluctuations, extreme temperatures, vibrations, humidity, radio interferences, etc, all that is calculated with a great tolerance with safety as priority, ....so maybe at some point a soldered solderless connection could be at some point worst than any calculated risk.

I would start learning CAN Sniffing, even DTS Monaco has live CAN messages displayed very clearly with Stop, Pause, Rec...even sending.manually UDS Commands ( https://en.wikipedia.org/wiki/Unifie...ostic_Services).

The Software always monitor latency, network errors, voltages, so ECU's will always switch to any higher safe level if something going wrong. As by CAN standard, if something goes wrong at the network level the CAN communication will be switched automatically to "one wire communication", so the car will still steer or brake...and a red message will force the driver to take care.

 

In post 4 related to this thread I provide pointers in sorted order to eliminate the root cause issue.

This can be overwhelming when first dealing with so many seemingly unrelated issues.

I have studied W212 CAN schematic and can speak to an interesting self feeding-gremlin:
the Noisy voltage chaos...
The W212 ALT power source is controlled in multiple stages to accommodate load demand. At some point the ECU decides to hand over dynamic voltage control to R-SAM once the main battery is deamed charged over 80%.
At that point the R-SAM manages the chassis voltage between 12.6V and 13.3V based on load demand. R-SAM signals its demands to ECU over CAN-B through the Central CAN Gateway Module.
ECU remotely manages directly the ALT over a LIN, on its oen during battery charge or based on R-SAM during float stage.

That's when things can go terribly wrong and chassis voltage can plunge below 11.xV while driving.

Many reasons for that:
ECU has no preset minimum fail safe voltage
CGW corrupts the battery charge level data
CGW has no watchdog to reboot it

W212 CGW module is now co-hosted inside F-SAM enclosure. It still built around its own hardware with multicore CPU. These two modules are the core of this Bosch chassis!! Very busy and built with reliable hardware.

CGW is in charge of tracking messages across networks such that network can function with dome degree of separation.

The flip side of all this wonderful architecture is poor solderless connections implemented on a few hand-picked peripheral modules.

By soldering modules and cleaning GND, I have shown that every module with "pressed pins" had poor performance and impacted their peers by way of wasted network bandwidth.
The various module normally tolerate gracefully disrupted networking conditions expect the Bosch ECU in charge of ALT remote control.

What happening ?
CAN-Bus long undersized stranded twisted pairs connected to solderless pins routed across multiple distribution bus-bar are used to disrupt the chassis with gradually increasing chaos.
That's where reboot comes in handy. CGW/fSAM are always powered live 24x7 to initiate actions and track responses. They work remarquably faster after being power-cycled. This requires a soldered network to notice the difference else solderless nodes are always always slow to the extent of being delayed.

Example of slow networking:
-- alarm can hunk when opening the door after unlocking!
-- Keyless functions are random hit or miss
-- Tranny shifts are delayed and bang when executed.
-- Main AGM is drained in days instead of weeks while parked by delayed response to security pooling.

....///...

 

It's very interesting to know that Vediamo provides CAN troubleshooting. Right now the top issue is with the hardware not the protocol level.
Meaning I don't want to read the data the software can display, I want to cancel the marginal connections.
THIS HAS GRANTED MY CHASSIS EXCELLENT IMPROVEMENTS so I continue to 100% coverage.
meaning... :
My Engine does multishot GDI injection
My engine runs vibration free despite colapsed mounts
My tranny shifts seemlessly perfect
My battery does not drain while parked
My engine does "Start/No crank"
My ALT does 12.6V float voltage
My Keyless coin-cell last 2 or 3 years
My steerings hugs 0° center on highway

​​​​​​I started with the amazin' "painted GND" then naturally discovered the loose-pins on circuit boards in the trunk...

I used the experimental method to realize most of chassis troubles were related to solderless nodes.

I decided to survey and eliminate all of it. Best thing that happened to this car - - very simple to do, without much electronic knowledge required.

My survey led me to realize only selected modules feature solderless - The essential modules are built to perform normally:

1. ECU
2. TCU
3. COMAND
4. F-SAM
5. CGW
6. DISTRONIC
7. IC
8. DISPLAY
9. OCP
10. SRS
11. ALT
12. BlwrSpeed
13. Satyelite radio
14. mBrace teleAid
15. 2xSeatsCtl/2xBeltsTen.
16. 2xSound Amplifiers

The above modules performance are impacted by proxy using CGW.

Top grade trouble providers are... (VIPS!!):

1. ESP brakes!!
2. EPS SteeringRack!!
3. SCM SteeringCol!!
4. EIS Ignit!!
5. ISM PRND!!
6. R-SAM Pwr Mgt!!
7. 4xDCU's Doors
8. AAC AirCon
9. KYLS-G KeyGo
10. LED Headlights
11. Front radars both near + long distance
12. Rear blind spot radars
13. LCP LowrPanel
14. Rotary dial/E-S-Buton Console

Its not just poor pin connections it's also hiw these modules are networked.
VIP's are networked well with direct links on CGW/SAM,
troubles providers are wired with artificially loosy connections.
These days I am mapping how R-SAM and RFK network with F-SAM/ECU to beef-up my twisted pair with thicker shielded cable

 

If you don't have a W212 you don't need to worry about soldering it.

The chaos I've described is increasingly less gentle until it "limp modes "ECU/TCU" on the highway. This is the gradual effect the REBOOT procedure is effective at cancelling.

At the top I point out the amazin' single"Main GND Strap" as it affects everything powered by ALT on the chassis.


> Legacy source... W211:
All this architecture did not get started yesterday - Some of this was already found in W211... I can guess solderless EIS, SCM, RSAM, XenonCTL, AAC, DCU,... already implemented to soft-crash CAN-C VIP's by way of CGW.

I don't think dynamic ALT Mgt was in place 12.6 to 14.9v thank God.

The single GND strap all the way undercarriage is too special. It was handed to me by Master Surya himself!

 

Not speaking for Cali here but the big takeaway is to make sure the batteries are both in a like new state so that that variable can be removed. Then make sure all electrical connections are as solid as can be, because there may be a net benefit.

My experience with wiring diagrams and soldering comes from the pinball hobby. Did you know that those machines historically had more than six miles of wire, equivalent to a small car? What I learned through a dozen years of making repairs to those machines is that vibration and heat will diminish performance and poorly designed wiring circuits required correction down the road. Cold Solder joints, corrosion, and overheated connectors were a constant theme. The first thing to be done when buying a new used machine would be to reflow all of the solder on all of the boards, lights, switches and solenoids. A proper solid connection would fix most problems and some of those boards are now nearly fifty years old. Nowadays, the machines are as advanced as modern electronics and require replacement of more parts vs repair. I had to learn SMD soldering and everything is small and more easily damaged. In one particular forum, I was the guy repairing boards (that were NLA) on a machine I had that nobody else had ever touched. Ive done a lot of electrical work and a good connection is everything. If you ever watch any of the electronics repair YouTube videos, they mostly are just cleaning and making sure there is a good flow of signal throughout the boards.

So when I liken this to what Cali and others may start to explore with soldering boards, it makes perfect sense that a better permanent connection will be reliable in every situation. These boards are designed for manufacturing savings and may work as intended, but they are susceptible to becoming loose and having more corrosion due to more exposed surface area on the pins and connectors. This solid connection will reduce heat (as long as wiring is sufficient) and guarantee a clean signal. When this is done, electronic performance will remain more stable over a longer period of time and allow better diagnosis when problems do pop up.

Hope the story helps.

 

Exactly! Half of this chassis is built like a pinball machine - What sadden me is the intentional purpose.
Ppl who fix systems must understand how they work to derive how they fail. Here we are dealing with engineered failure modes that got perfected to go down as planned. That in itself could be fair if systems worked well for a limited time.

The selection of solderless modules is really specific. Some must be robust and others less so.
There's a huge contrast between "crimped + soldered" top grade connectors compared to "painted GND"

There's a huge gap between ESP and the Mitsubishi headunit.

There's a huge difference in harness routing to split noisy power and signal paths (3-Ph Fuel pump harness!)

There is huge contrast between best-practice and bad-practice... big opportunity is in enhancing CAN-B and CAN-C nodes:
harness grade
solderless PCB's
routing path
That's what I am going to experiment on next.

What I clearly found is the more chaos is cancelled, the closer to perfectly normal ECU/TCU perform.
Soldering is required.

 

.

 

If you check post #3, you will see I have already closed the loop for now. Checking and addressing electrical connections has been on my list since I got the car, and obvious electrical issues elevated their relative importance on the list.

I've only owned MBs (models going back to the mid-70s, gas and diesel) and I'm my mechanic, so some of the perceptions you appear to have make sense to me, but they really only apply to older generations of Mercedes. My W211 was still basically a fairly computerized car. This W212 is a server room on wheels.

 

Well said. WELL SAID!!!
In my technology days in 80's and 90's when we would have constant problems with client/server apps and Microsoft Windows we would say how scary and unacceptable it would be if cars ran on MS Windows . LOL
Well . . . .

 

... it almost works right despite marginal connections.
Commuters just want to drive from A to B and some of us mind normal systems performance.

Bosch fault tolerance can handle less than perfect conditions to still work in reduced mode!

It's easy to cancel the 722.9 factory setup that cause delayed banging shifts... not adaptations! 722.9 excels at self-adaptation given normal conditions.

You can solder SCM + EIS to provide yourself with motivations to complete ESP + ISM: vibration free engine + seemless shifts.

For everyone else you will observe the ECU/TCU moodiness, hanged gears, low shifts and heavy chassis. Bosch performance scales both ways.

You can drive this chassis as is or get the engine nimble and tranny seemless without replacing anything. Free performance is an awesome value proposition.

 

I just ordered the EIS opening/spin tool. It did not state W212 compatibility, but I think it will fit.

 

I feel sorry for you Plutoe.
Here we are enjoying tinkering with OUR car and NOT YOUR car, and you seems to be always forever in a bad mood and very offensive.
It is our car, even if we want to burn it to the ground, it is none of your business.
If you don't like this thread, don't read it.

if you were a kid, for being this rude and naughty, I would pull your ear and make it like Mr Spock in Star Trek....

 

MS! I concur, there's a lot of truth to your statement.

Here to help... else skip drama

 

Not sure why the OP believes faulty crimp terminals are the problem, but trying to solder them is not the answer. If there is corrosion on the mating surfaces of connector pins, electronic contact cleaner might provide a temporary remedy. Automotive connectors used in areas exposed to contamination have polymer sealing glands around each wire and around the connector shell. If the seals are shot, the connectors need to be replaced, or the entire cable assembly should be replaced. You cannot solder corroded wires and contact pins, and attempting this will damage the wire insulation and wire seals.

 

The SOLDERLESS meant by OP is this one : The one at the PCB




You speak of the common connector with wire and crimped terminal, which will be below :




.

 

Thanks for the clarification. I've seen this press-fit assembly technique before -- it's supposed to form a gas-tight contact between the connector pin and the PCB's plated through hole, and remain integral. Has there been moisture intrusion into this assembly module? I'm sure there's been testing on this type of assembly technique, but I would be reluctant to use it in applications subject to shock and vibration, such as automotive. I once experienced an intermittent relay in an automobile that was soldered into a single sided PCB, where the solder was expected to form both the electrical and mechanical bond -- bad idea.

 

The press fit is not gas tight, it is just faster than soldering and less solder use overall.
With those lead-free solder which is more brittle, I guess its a welcome technique by modules/electronic manufacturer.

Since we shall solder using lead type solder, and better with a solder having a 2.5% silver, the soldering will assist the press-fit to be better long term and less chance of oxidation.

Water intrusion NO, but surely overtime all press-fit contact do get oxidized, depending on where the car is being used.

==========

On my boat, I crimp and then solder each and every additional connector I use.
Its very salty and humid air on the boat and it has survived well.



.

.






I only use this solder since the 90s, from Goot, the one with 2.5% Silver : https://www.amazon.sg/SE-0AG08-Solde.../dp/B001PR1L46
It is very reliable long term.



Below : If Bare copper bar, and brass shunt. This is only 12 months on the boat.

All big cables are original Suzuki engine main power wires, except 1 with white mini cable tie.



Nickle plated copper bar for battery, when new in 2005 vs in 2020
.



Nickel for coating and silver for soldering is great. I custom order these from high purity copper and the battery terminal is also custom made, I copy high end audio battery terminal design,
but uses 99.XX% high purity copper. My partner for this boat has all the tooling like wire cut, CNC machine and etc etc, because he has a factory making home appliances

.

 

... in addition to the single "main GND strap" located under carriage to deliver glitches under load. MB used no protection to prevent oxidation.

The combination of noisy voltage reference GND + oxidized solderless pins implemanted in specific modules reduces the performance of functions including ECU/TCU/ESP.

This translates in various common conditions such as battery drains, delayed shifts, basic GDI timings and the extreme "X-Mas Tree" limp mode... Happy Holiday's!

Hardware conditions limit the performance of software modules by way of network bottleneck through CGW.

> Available Fixes:
For normal ECU/TCU timings: fix CAN-C
For battery drain: fix CAN-B.


> Quick rewarding fun:
-- Does your chassis has "Blind spot" LED in side mirrors ...
-- solder one door module in 20mn
-- then observe how bright dim LED becomes!

 

I've worked with lead-free solder, and it's miserable for both electronic assembly and plumbing. I thought certain industries -- automotive, aerospace and medical -- were exempt from RoHS lead-free requirements due to the solder's brittleness and unproven long term reliability. Pressfit connector manufacturers claim it's better than solder.

 

in theory press-fit is better than solder for mechanical vibration because solder joints remain somewhat soft and sensitive to stress.

I don't think we're dealing with an honest mistake for a collection of reasons...

--1- Only peripherals modules use solderless pins.
- Core power train modules are soldered !
- Does that choice tell you anything about what works well??

--2- Then at some point you realize more details are lined up 180° backwards against best-practice.

​​​​​​Implemented offenders are:

1. noisy harness bundled jointly!
2. undersized conductors!
3. underated caps filtering!
4. artificially extra long runs!
5. painted GND's!
6. single noisy main-strap!

Every single item above is a big no-no!!
Instead of being made more resilient, details are used to create weaknesses.


--3- Another layer is the incremental effect of degradation.
You'll realize that rebooting shapes things up temporarily.
That helps you realize software is involved. (hardware faults can not be power-cycled)


All the above is skillfully setup from factory without any aging or mileage necessary to behin acting up.

​​​​
+++ Solderless Gas Pedal... Recalled
Some of us remember the accelerator pedal was recalled early on because of a vague "pin connection resistance issue" causing ECU limp mode CEL


+++ Confusing For Most:
What can be misleading is how voltage issues are related to ECU/TCU timing chaos by way of CAN bandwidth.


> NO CANT DO... fix:
This chaos combination is a master piece by system integrators, well hidden.

Solderless modules tank performance without warning or DTC early on.

The amazingly weak battery has long been a Mercedes staple.
Same for banging tranny poor adaptations.

Fortunately for us this can be made better than new with zero part.