Been on this one for a while now and think I have exhausted pretty much all leads.
The Car: 95 E320 with 104.992 Gas 3.2L California
Symptom: Idles rough when warmed up and surges at times. Does not do this when AC is on.
Without AC and warm engine. As you approach a red light and put your foot slightly onto the brake at 1-2 Miles per Hour, the engine starts to surge perhaps up to 800-900 rpm's and repeats until the car comes to a halt. Now it idles at 650 and misses ones in a while well I guess every other revolution. With the AC on I can't make out these symptom's.

Vehicle has the California 16 port / pin diagnostics connector.
Error Codes:
DTC:
21 Fuel purge control valve, open or short circuit (more on this later)
26 Transmission Upshift delay switchover valve (Y3/3) without function (more on this later)

Port 8:
13 Throttle valve potentiometer actual value too low.
37 Injection valve cylinder 3, short to positive

1. The Vehicle has the Updated engine wiring harness.
2. The throttle body has been replaced too.
3. Fuel pressure regulator (new)
4. E-Gas Computer (replaced)
5. Purge valve (replaced)
6. Ignition Coils (new)
7. Spark Plugs (new)
8. Throttle Control Cable (new)
9. OVP (new)
10. O2 Sensor (new Bosch)
11. Some Vacuum lines and connector replaced with new lines
12. Mass Air Meter (new Bosch)
13. AIT (new Behr)
14. Air Filter (new)
15. New Oil Dip stick o ring.
16. Cleaned the EGR pipe all the way.

There are probably other things that I replaced, but can't remember each and everything.

The Fuel Purge Valve, I got it from another 104.992 (junk yard) and it worked initially. Thought it got stuck, cleaned it and put back in. No go!
Put it back on the bench and tested it with power supply statically. YES, it does work!
Took a scope to the car and looked at the signal from the ECU. Static at VBatt. Since the ECU seems to switch the return, I checked there. Static VBatt. Seem the ECU doesn't bother to switch it on and off.

Transmission Upshift delay, seems to be similar as with purge valve. Switchover valve works on bench, but not in car. I did an 0815 test on the check valves and I believe that one closes faster than the other one. Vacuum is present and held, I can hear the discharge when I disconnect the lines.

Throttle valve potentiometer actual value too low.
Don't really know what to make out of this. Only a cracked Rheostat could cause a value too low. What is the Computer looking for? If the resistance is Zero, can't get any lower than that. So, it must be some other value it is expecting.
I had CAN BUS communication errors at times, but not recently. Perhaps related to a weak battery?

Injection valve cylinder 3, short to positive, could be my misses, if the valve is really short to positive, would it run rich at 3? or is it lean.
Unfortunately the PO messed up the bolts that hold the fuel rail, I haven't dared to attack this, I think the only way to get them out is with a chisel.
However, I had in the past similar codes on different cylinder? Is the problem travelling?

In an other post, I read that a weak Battery could throw random errors, I find it hard to believe, not knowing the algorithm the ECU goes by, one would suspect several samples before making a decision and throw an error.

I don't want to throw any more parts at the car, although most of what I replaced was more or less necessary. The PO didn't maintain very well and that goes for the front end too.

At this point I see only two more things that could be defective.
Either the battery is weak (don't really believe it) Did another 0815 test on the battery, but have no Idea what the load condition would be. Ignition on, not running. after about 15 minutes the voltage dropped below 12 Volt somewhere to 11.99. Still plenty of energy left to start. Ignition off the battery recovers to 12.4 Volts after a couple of minutes.

Or the ECU has been damaged while the PO was trying to drive with the bad Throttle Body (really bad wiring pretty much all short) I don't want to replace the ECU with another one because of the coding and it could mean trouble during the next smog check.

Someone any more idea's?

__________________
Mostly, I don't know notin, I just know where to look.
I am looking back, to over 30 years in Electronics Design.
Electrons don't care if they move in a car, computer or relay!

95 W124 E320 M104.992 - Because, I love to repair, naaaah!
Over 221,000 Miles
Cheers,
Norbert

Than (what tool) have been read 8 port error?
led or computer-based scanner?

I think this "slow codes" (led blinks), then:
13 035 with TWC
Lambda control is at rich stop, mixture too lean
Air ingress, fuel injection valves
Diaphragm pressure regulator

13 036 with TWC
Lambda control is at lean stop, mixture too rich
Air ingress, fuel injection valves
Diaphragm pressure regulator

37 088 with automatic transmission
Upshift delay switchover valve (Y3/3) - open circuit/short circuit

Source Error Codes - Mercedes WIS document


The fact that your car can not be code 013 (if the scanner reading) since HFM ECU does not control the throttle itself, and therefore can not see a problem with his potentiometers ....

the best solution - connect the scanner and drive the car to the problem manifested itself when the diagnostic session, and then try to find the problem in the data (logs), which will bring together the scanner.

PS: you have Throttle angle = 0 on idle speed (log data from febrary 2014)
try this:
1. engine is warm (coolant>80*)
2. set ignition ON and wait > 2 minutes
3. run engine.
This adapt throttle on your engine.

Didn't want to be rude.
Just ran across your reply.
Thanks for your thoughts,

It turns out that the crank case breather system was causing the problems, sounds weird, but I'll explain a little more.

The crank vent goes through a mounting port attached to the intake manifold, but not to the vacuum yet.
A second port is connected via a short breather hose to the lower intake manifold.
That's where the oil fumes are submitted to the intake and supposedly be burned through the engine.
If that port breaks off or allows un-measured air to enter the manifold, you have pretty much a messy condition and the loop can't possibly be closed.
Although, I replaced all fuel injector, but believe that the broken breather port was not repaired properly before and caused the problems I had all along.
Now, the engine is running as supposed to and I have to deal only with some other minor codes, etc.

Thanks again!
.

__________________
Mostly, I don't know notin, I just know where to look.
I am looking back, to over 30 years in Electronics Design.
Electrons don't care if they move in a car, computer or relay!

95 W124 E320 M104.992 - Because, I love to repair, naaaah!
Over 221,000 Miles
Cheers,
Norbert