AvramD

Well, I'm not concerned with CPU- or motor- delays, at least I don't think I am, b/c I was happy w/ the B5 S4's system.

Unless one of these things is physically sticking. The motor itself could conceivably be sticking, I hadn't considered that (as obvious as it seems now). But I don't think I need to be the one to get into that detail.

If this is the case, it sounds to me like measuring pedal position vs throttle position will show an approximation of the delay, although it will be off by the CPU delay. That's ok; as long as I can demonstrate that it is a) inconsistant, and b) sometimes unaccceptable, I would think that would be enough for Audi to agree they have to figure it out. Don't you think?

So, if I understand correctly, the diagnostic tool will report throttle positions to me via a sensor that looks at the angle the butterfly is actually at, not the angle the motor is being told to put it at. Thus it shouldn't matter where in the system the delay is, I'll catch it.

Am I missing anything here?

mtu_audi

Well it seems to me you are concerned about the cpu delay if you are worried about throttle response... If you are seeing a lag between when you push on the pedal and the throttle plate moving to the desired position then it could be the cpu delay and or the maximum velocity the motor that moves the plate. Is it that sometimes you have a delay and sometimes not? Or is it just there is a consistent lag between pedal and throttle? I guess I'm not sure exactly the problem you are seeing.

As far as a diagnostic tool I don't know if it will give you access to those parameters. Usually a diagnostic tool will let you view diagnostic codes that are set by the ECU when a fault is detected. I don't know that they give you access to the sensors. Even if it did, it would have to be something other than the raw data otherwise it wouldn't tell you anything meaningful.

Because you and more than likely your service shop do not have the proper tools to test with, it may be better to see if you have the latest version of ECU software in case a newer version was released.

mtu_audi

Oh, by the way it is possible the throttle plate is sticking. We had this problem with dual bore throttle bodies. The electric motor didn't have enough wraps to overcome the coefficient of friction at low voltage. You would see low voltage to the motor at cold temperatures so if the motor is not properly "sized", it won't break the throttle free at lower voltages...

Also another thing to consider, we tested our physical throttle body against the competition including Bosch, which I believe are on Audi's, and the maximum velocity that their throttle plate could move was significantly slower than ours...

AvramD

And we're not talking about the minor sorts of magnitudes that a CPU takes to make a decision. We're talking in some cases about a full second or more.

As for diagnostic tools, the vag-com tool does everything that Audi's in-shop tool does, which is to say it can poll the entire ECU's memory in blocks, up to 3 "data blocks" at a time. The key is to know which blocks, and which data within the blocks, and as you say, how to interpret the numbers.

From both reading the vag-com forum and talking to the US distributors of the products, both pedal position and throttle position are routine things to monitor, and their data positions are known. The US distributor has also said that the B6 S4 (and presumably A4) are capable of polling 10 times per second; I think that will be precise enough for my needs.

As for the meaning of the data, again according to the forum, the data field actually does represent the position of the potentiometer. It seems to me they said it's actually expressed as a value between zero and one, with the car set up to use an effective range of actually only the middle 96% of that, b/c pots aren't reliable enough at their extremes. So if I cared about actual position, I'd have to scale it.

But I don't, only need to show that delay is significant and inconsistant.

mtu_audi

Sounds like you should be able to detect significant delay then. Can you record data then for both pedal position and throttle position and be able to look at it afterwards on a computer? Anyways, just for reference, our spec states that we need to acheive a position within 150 milliseconds of issuing a command. If you are accurate within 100 milliseconds as you stated, you should easily be able to tell if there is significant delay... As for the values you get from pedal being from 0-1 it is probably either a percentage of an analog reading of voltage or a percentage of a PWM (pulse width modulation) signal. Throttle position sensors is probably a percentage of voltage as well...