CaliBenzDriver

Recently I made progress surveying more solderless modules on my W212-Facelift.

I've had my eyes on ESP for a long while even before sanitizing my ISM shifter and exploring my sealed SGR radar.

This experimental testing conducted under careful supervision. Just like with the oil solenoid MOD: be conservative when your life and property value are concerned. No claim of fitness expressed yadi-yada don't do it unless....


The more solderless modules I fixed, the more I grew curious to find out what each module would improve. It became obvious that modules with poor connection underperformed quietly without any fault... Ouch, that's clearly not why I custom ordered a premium sedan.

The solution is very simply applied by soldering oxidized pressed-pins !

My engine/tranny got an extreme amount of bang with this quick fix. With GDI precise timings, Engine got extremely powerful, reslonsive and smooth.
Tranny and engine now have coordinated shifts with matched RPM, the oposite of sloppy, banging late shifts up/down.


> FYI :
Working on electronic brake module may yield hazardous results. Not working on it will lead to problems as well.

The normal repair procedure is to replace failed modules with OEM Certified replacements. Unfortunately here all factory ESP modules are likely built the exact same features...

(... A while back I rationalized that modules glitchy connections may prevent facing other less appealing failures.
Meaning it's not great but may prevent worse things from happening.)


ESP disassembled showing 12x solenoid actuator coils


nicely dry hydraulic section with contacts
This design really shines. What a pure marvel with great firmware assembled with precision


BOSCH best in class technology



metallic cover removed


what have we here... opportunity for enhancement!


overall solder-side view with heatpads
I wish I could see both PCB sides but it's not an option to lift the board out


soldered pins... (GND are extra cold joints)


nicely coated with silicone conformal protection


all good ready to perform
Anxious then to test drive this quick fix... OMG!


module re-seated before getting covered up
Extra heatsink paste added for good heatspread.


taped for a clean RTV seal

This excellent module really benefits my chassis with all its pin connections normally soldered. This reduces poor contact resistance that quietly force performance down. No thank you!


I noticed improvements in my electric steering, the autonomous braking under Distronic+ and best of all the ECU/TCU are now INTERACTING much faster.
This leads the tranny to shift differently and engine timing to use GDI multi-shots in the compression stroke before multi-sparks ahead of TDC.

The result is an ABSOLUTE MUST for this platform!


The difference is so huge it clearly highlights that ECU performance depends unfortunately on CAN latencies. This should not be but that's what we are dealing with!


ECU timings should absolutely not be exposed to any external network chaos factors. Realtime mapped calculations of injection and firing should never suffer random delays, this setup does. It kills performance with signal jitter.

At this point, best options are few. I choose to solder pressed-pins. Ppl may experiment by having a local or Internet shops resolder, coat and reseal your own module. Perhaps you want to clone a used unit for a fearless testing experience...

Earlier we've shown that dry-lubed cylinders were introducing uneven contributions easily remedied with precise MOD-2.1 oil pressure.

> Engineered pin connections:
Here are the different areas that I reworked for my testing. I soldered 100% of the solderless connections.


input connector vs. solenoids

Module connector is at the bottom flanked by power supply pins on each side.
Fields of solenoids are up towards center.


groups of glitchy solenoids pulses


input pins used to supply power + signals


The way I rationalize GDI timings are able to transform my M276 is that the extreme pressure is used to pump as much gas as the ECU wants in the compression.
I don't know how Bosch deals with the perfect 14:1 ratio during rich mixtures. This puts out unbelievably insane power/torque... these huge accelerations have to be rich!!

Gas pedal is really granular touchy sensitive and nimble with extreme response.

The next guy who enables this is gonna blow his brain with unbelievable results.... transformative in the best way possible!

CaliBenzDriver

The whole ESP module is connected by one quick release connector. Again let's have a look at... PINS but this time from their other end.

It's super hard to get a good picture of these pins, here is what I got...










Now let's zoom in closer...

power input pins


somewhat oxidized


the female pin is very aggressive... digs right into the male spade


some male pins better shape than others
That's not a call for inaction. A little oxide protection should help stabilize unwanted oxidation.

Beware: we note that module connectors have a limited insertion count. Clearly not Zero Insertion Force (ZIF)..

Overall this module water tightness appears good. It is well protected the way it is unlike say the rear BlindSpot radar connectors.

Oxidation naturally creeps in to deteriorate contacts. Poor contacts cause in-line resistance that cause more oxidation, more heat...

​​​​​Whatever you decide.... DO NOT SCRAPE OFF THE METAL PLATING LAYERS!
The pins are made of sophisticated alloys coated with both conductive and protective material such as silver or something similar but extra cheap.


Improvements are easy to come by

+++ FOCUS ON PINS OF ABS ACCUMULATOR PUMP

The traditional ABS/ESP failure involves the pump motor degrading. Treat the motor power feed with extra care.


pump male spades and sensor golden contacts


limited use of gold plating

Whatever you do, your safety is in your hands!
Enjoy premium performance, it's all yours for free.

CaliBenzDriver

The bottom line here is... SOLDERLESS DISABLES LIMITS PERFORMANCE of networked modules, all of them!


It is counter intuitive up front to connect limited chassis performance with solderless electronics.

However the relationship becomes directly evident working with the VIP CAN-C wana be high-speed network.

Say GOODBY TO LAGGY ENGINE AND COMPLEMENTARY BANGING TRANNY SHIFTS once and for all: its 90% caused by poor coms. The tranny itself is really sharp when allowed to perform.

When modules rely on precise timings the networking performance MUST be closely supervised.

I personally can not believe all this chaos is unmanaged by the premier top German flagship brand.


From this hands-on experimental survey, I can say we miss debugging tools to gain insight on Boch design. The OBDII interface is keep quiet without useful data access.

Luckily with a soldering iron we are not longer powerless. Now I am not going to sit here to complain and do nothing.

People who can't solder electronics should instead get their own solderless modules to a repair shop of their choice.


> GDI Timings :
Bosch injection has multiple timing styles: legacy or GDI multi-shots.
The words dont describe the powerful experience!

I am not a gasoline injection engineer, it's just the extreme difference in power is night and day. The sound is also much different... it's hard to miss all together.


To get ECU to switch over to GDI timings, the key is to have a smooth predictable engine with a healthy CAN-C.
That means oil-sealed cylinders at driving RPM with soldered modules to deliver pristine timely data.

I am not totally sure I got all 100% of this chaos canceled because I can still get my engine to cough (fish-bites misfire) while scanning my chassis. Try scanning your engine idling/Parked.

Where I am... is that our Bosch CGW Gateway is not innocent !!! Now I want to go iron out my solderless MFK - It's on CAN-E, not C and likely transacts across CGW network bridge that gets bugged by corrupted packets.

Also I am growing increasingly curious about the big 100A Bosch fan... its internal circuit board and diassembly procedure. Can we find a YT to dismantle our big fan - I can't assume it's not introducing PWM glitches straight into main pre-fuse box.

The delays are incremental by nature. The less delays there are, the better things work. My car uses a stereoscopic windshield camera to activate ESP brakes as needed. THAT'S A LOT OF DATA STREAMING THROUGH CGW!!!

S-Prihadi

HHmmmm, more delicate fingers work required.... from me.

Nice write up. Thanks Cali.

I thought I saw special effects from Avatar movie looking at the gold contact...LOL. So nice looking.

2 Questions below......

S-Prihadi

You speak of CAN BUS data glitch or overload, I think my data logger maybe able to capture it and seen by expert eye using proper software.
I have this https://www.csselectronics.com/produ...terface-cl2000 it was for my headlight command sniffing.
Car manufacturers use this to debug or whatever the software/technical term is,...... their CAN BUS.

S-Prihadi

My ABS-ESP is the cheapo version, my car does not have the ESP Premium like yours.
I only have 235 - Active Park Assist , that simple beep-beep small led display.
I like it this way, no intervention yada yada from electronics towards my driving style.

So my module ID is N30/4 and yours is N30/7 , right ?

So less data bandwith traffic for my module.....I suppose

CaliBenzDriver

We are blessed by this excellent interface between the hydraulic valves and their electronic control. It's very easy to service when gone bad.

The 12x round shapes are basically.... SOLENOIDS! Both male and female sides: control --> actuator

The golden contacts I believe are connections to some types of pressure sensors. I do not have ESP inner schematic and I don't think Bosch intends anyone to do any tempering inside this precious module.



MS questions
These parts are static, they can not move at all.
I can't tell what sort of signal polarity is involved without Bosch proprietary schematic.



> There is a small issue about CENTERING....
- Both module parts do not have precise centering. That's why each of the gold pads have an area to accommodate the possible male location.

- The solenoids coils are all extremely loose (pin solder!) to be able to accomdate loose location.

- This lack of precise positioning allows all assembly parts to be more compatible and perhaps more vibrations immune.


gold contacts
These round dots carved in the contact pads are likely some sort of junctions to the internal sensor.

At first I thought the holes had been carved by the opposite matting side but now that we are looking at it... the shape looks too perfectly round.

Ultimately I did not deal at all with these contacts as I only wanted to solder the amazing pin collection. I did not clean them, I did not lube them, not even brushed them.

The way I considered this was:
- module has zero fault
- very easy to revisit if needed
- gold is free of any oxidation
So I passed to stay focused.

If facing errors with a single bad pressure then these contacts could be visually inspected.

It seems to me this module is built to concentrate issues on its weak connections **

One of the lessons of fixing systems is to carefully not create new problem while fixing existing ones else issues grow exponentially!


** this is like a shredded serpentine belt when one of the accessories has gone bad... Having a tougher belt would keep spinning defective compressor: ouch
Having weaknesses built all over the place helps make predictable systems!

To rework solderless pins I used a hot temp regulated 40W iron plus a second powerful 100W solder iron to heat cold GND pins. That's fairly easy.
The bea_tch is to clean up the old RTV sealant out of the metal cover... hours

CaliBenzDriver

This seems like a PERFECT CAN TOOL to troubleshoot datagrams. I don't see packet-sniffing going very fast as in... extremely time consuming measured in Month.

Have you dealt with "Wire-shark" network traffic tool to troubleshoot performance?
I think this powerful CAN networking tool as similar.

I would totally be up for it if we did not have any clue or solution.
Thing is I am biased for solutions. If we spend months to uncover issues without solutions, we made no progress, right?

Right now I have the practical proof that our chassis performance is quietly trashed by solderless modules including Bosch crown-jewel ESP.

On top of that I am also sure there are mismatched firmware incompatibility bugs but software quality is a lesser priority while the underlying hardware is wonky unstable.

To sum up where i think we are... a few selected modules are built solderless to stress-test the gateway modules with bad traffic. The CGW is overwhelmed by outstanding connections.
Traffic across gets increasingly slower until chassis is rebooted

..../...

S-Prihadi

No I never use wire shark.... you know me, I hate IT stuff

CaliBenzDriver

It's been over 30 days since I reworked my solderless ESP module - To my surprise the unexpected outcome is the best-effort by ECU to fully charge main AGM perfectly.

The good part is no 90Amp killer MAIN + AUX drain while driving. It took me 10Yr to survey the bugs out of this vehicle. 80% hope it'll holds !


> TEST DRIVE THIS:
I was really not prepared and quite surprised when I took those pictures almost by chance... I saw unusually good numbers charge behavior.

Here my cool engine is not fully warmed up after only 1Mile and the battery charge already enters the no-charge 12.6V float cycle.

This is without any manual charge help! I have not hooked up that car for about 2 Mo. Battery charge status was honestly maintained by the car itself. That's how I aquire actionable valid data.

I am ready to declare the YO-YO factory defect FIXED! If ECU control doesn't drop below 12.0VDC for an additional 30 days, it will be a giant step for Bosch distributed voltage control.


In chronological order:

-01- engine still cool already nearly full (5A!)


-02- ECU switches Batt Mgt over to R-SAM Float Mgt


-03- decelerating triggers a charge burst by ECU


-04- Cruising steady switches back to float, no charge


-05- tiny charge


-06- ECO flag turned ON


-07- reduced load during acceleration


-08- same as previous


-09- gentle charge


-0A- gentle charge current

> TEST RESULTS...
Fixing ESP enable ECU to work with R-SAM normally on a different CAN.... so CGW IS INVOLVED in this solderless fix.
This is badly integrated on many levels.

S-Prihadi

Get an OBD2 logger, so that we all can see 1 full 30 minutes drive cycle by per second...
Sometimes the under 12.6V condition is very very short, like 2-3 seconds only.

CaliBenzDriver

We have passed the early stage of studies and are now testing fixes. Fixing solderless modules has always returned unforseen benefits. ESP is not exception!!

Thank in parts to you Master Surya, I have fixed my ESP and realized the advanced charge control control is now working... knock on wood, still!!

My battery quit getting drained while driving and while parked.
Without this factory bugs, my car is becoming a nice to drive.

I see loading busy CAN + CGW with additional traffic as a bad long-term practice. It's only good for short troubleshooting only.
Your chassis is proven to have factory YO-YO and here I link this to marginal COM'S.

Honestly with your need for good A/C, disconnected ALT is a great insurance. You don't need (12.6V to 14.9v) , 14.4V with minimum charge is nearly perfect.
What's still seriously wrong is this chassis has proven poor COM's.

By fixing the Bosch ESP I evidenced it screws up the Bosch CGW that impacts the Bosch ECU injection timings. One would imagine Bosch test its design integration better than marginal from factory.

No problem, now there's an easy fix.

> Hands-on testing ...
I can not debug the firmware source of the CGW nor can I use a logic analyzer to study the signal delays through broken queue management.


> What can we do to prevent being nagged by managed chaos?
Buy a used identical ESP on eBay, clone with your chassis and get it reworked better than new.
CHEAP