txgriz

Audi won't (apparently) pay for the ECM update unless you have the DTC codes set:

DTC P008800 (fuel, rail/system pressure too high)
DTC P01C900 (fuel, rail/system pressure bank 2 too high)

I have the TSB for this update. I have a theory on what the real problem is and it may explain why some cars behave differently that others (on the same firmware). My theory is that the car has a power supply regulation circuit that reacts too slowly to an instantaneous increase in the load current requested. The TSB has text that says "If there is a low system voltage, the on/off ratio may not be enough to keep the fuel metering valve constant in all areas."

So the theory goes, you are slowing down and the current draw from the power supply is relatively light. You jam on the gas pedal and the car immediately requests a lot more current from the power supply than before. This current draw drains whatever capacitance is holding up the system voltage and this causes a momentary drop in the voltage rails across the car. As the power supply control loop reacts to the load to restore the correct voltage, the other systems see a momentary lower voltage which stalls out various systems in the car that are sensitive to that voltage rail being in spec.

The variability can come from the various tolerances of the inductors and capacitors in the system may allow for some car's power supplies to naturally react faster than others. /engineer talk/ The power supply control loop is probably a digital PID controller that the ECM in this update reprograms to react faster to load transients, thus solving the problems for all the cars getting the update/engineer talk off/

My car has the lag issue and my car also has the problem that if i'm at a dead stop and I mash the pedal, my radar detector will spuriously reset. I believe the reset is due to a voltage drop on the 12V supply to the DC adapter. To really be sure, I need to monitor this voltage over time as I drive around town like an idiot.

DeerHunter

The above being the case, why would a tune solve the problem (as reported by several owners)?

nobbyv

Just to clarify: when you say you "have the TSB for this update", you mean you have the copy of the TSB I posted, right? When I first read your post, I thought you meant you had had the update done and STILL had the lag.

Couple things: first, you should only need ONE of the DTC codes to guarantee being covered by this update. I only had one. Second, I understand that some dealers may claim that AoA won't pay for this update without the DTCs. Personally, I wouldn't let it drop at that: I'd call AoA Customer Care and explain that you want this update. If you delicately throw in a few words like "dangerous, liability" etc. when speaking to them (which I think we all agree are applicable in this case), I think you'll find that AoA is willing to write off this update even without the DTCs. It's going to take some persistence, but it can be done.

Next, your theory is interesting, but there are a few holes I can think of:
1) Load transients are typically measured in magnitudes of nano or maybe micro seconds, NOT whole integer seconds, which is the duration of this "lag" effect that we see (or in my case, saw). Perhaps you're right, and the state machine for this control loop causes some pre-defined time period to sit in this "undervoltage" state, but multiple seconds seems like a LONG time.
2) I think you're overestimating how much extra power (of course, I mean current*voltage, not HP) would be needed even when flooring the accelerator: the duty cycle of the injectors will change, and the fuel pump will start working harder, but other than that not much else on a car changes that quickly when stomping on the gas. A modern car's power supply should be buffered with more than enough capacitance to withstand this increase.
3) Just throwing this out there: any chance your radar detector behavior is physical, not electrical? Like maybe a power cord under stress that gets pulled/moved slightly from the quick momentum change of the car accelerating rapidly?

You may be 100% right, I'm just playing devil's advocate here. Control Signals and Systems is a distant memory for me; I play in the firmware world these days.

txgriz

My SA sent me the TSB 2035235/2. It reads nearly the same as the one you posted. The latest revision date is 1-23-2014, but it was initially published 10/9/2013.

Working on power supplied for 10-20W systems, I have seen poorly designed systems take 50-100ms to respond to "brown out". That could be enough time to ripple through the system. Whatever effect the short voltage drop caused may take significantly longer to correct. Pure conjecture on my part. As for the radar detector, it actually resets before the car gets going but as the engines revs very high.

Since the tune rewrites the ECM, I can guess that time ordering of events changes the load transient profile.

[edit]. Also, the TSB's text itself indicates low system voltage is the problem causing the DTCs. How is the system voltage low unless it's due to a transient condition?

nobbyv

No, you were likely experiencing a different condition, which is simply electronic nannies interfering and trying to prevent sensed wheel slip.

The lag issue occurs even on dry, smooth pavement at part-throttle applications, and even with ESP turned off.

nobbyv

When I first read the bulletin I actually interpreted it the OPPOSITE way: a fuel pressure rail gets too high under this condition (request for moderate-to-high throttle just after braking), and some governing control loop cuts power to try and bring it back in line. But rereading it with your theory in mind, I agree that it makes more sense that the reduced voltage is causing the high pressure scenario, and not the other way around.

I wonder if we'll ever know for sure?

MalibuCANE

I said all the right things, the dealer tried to download it to my car and said they couldn't do it. I'm not sure how hard they tried or if that is just Audi trying to keep them from downloading but this sucks. I almost got clipped today pulling out onto a busy road.

Richard Solomon

Look in your warranty booklet (or trunk IIRC) for the equipment build sticker and it will give you the month and year of your build. If you got it in November '13 I'd imagine it's somewhere in the 08/13 to 10/13 range.


If folks would post up their ECU coding info that might help also - my 04/14 build is:

MalibuCANE

I think figuring out how the trip those fault codes would be a better use of time. Maybe I should just trade up to that unclaimed Daytona Gray Matte RS7 they had sitting in the showroom.

Richard Solomon

Wow, just wow.