Trying to understand the DME

[Robert986]

Was about to ask in the 987-MAF-thread, but to avoid destroying that thread I make a separate one. I´m trying to get my head around how the Motronic 7.8 actually works. And, I really feel kind of stupid, I´m sure I got some major parts wrong.. But here we go.

1) ADAPTION
I take it that the DME adapts to a certain amount of change in airflow after a improvment of intake or exhaust. This "adaption" is that reffering to the adjustment of fuel-input that is made by the short term fuel trim and after a while moves over to long term fuel trim? Or is it the actual fuelmaps that changes by some kind of magic?

2) MAF CALCULATIONS
As discussed in other threads, when using the 987 MAF-tube the DME doesnt know of the larger area of the 987-tube and calculates to little airflow leading to a lean state. If I have understood this correctly a pro tuner will be able to adjust this in the DME. How is this done? Is it a simple constant in the DME-calculation that is adjusted to represent the new area or are the whole fuelmaps adjusted?

3) MAPS
A very wild guess is that there are three states of operation:
a) Startup/Idling: I would guess that this is managed by a fixed map in a open loop condition.
b) "Middle ground/half throttle" I guess there is ONE singel fuelmap covering this area operating in "Closed loop" thus adjusting STFT/LTFT based upon feedback from the o2-sensors.
c) WOT (Driving like a man) I guess this is covered by a separate fuelmap operating in open loop.
d) Finally I guess that also the ignition is managed by separate maps just like the fuel.

4) CODE/PASSWORD
Is there a password to get into the Motronic that we need to get from the dealer? Then how do the pro tuners do? I have got an offer to tune my car and I don´t think they need a code from the dealer.

Just trying to learn, this is so interesting! Please feel free to tell me where I´m wrong above, or just guess together..

[thstone]

Here are some general responses. We could create an entire new forum dedicated to this subject!

Originally Posted by Robert986

1) ADAPTION
I take it that the DME adapts to a certain amount of change in airflow after a improvment of intake or exhaust. This "adaption" is that reffering to the adjustment of fuel-input that is made by the short term fuel trim and after a while moves over to long term fuel trim? Or is it the actual fuelmaps that changes by some kind of magic?

The OEM map is static, meaning that it doesn't change. Its hard programmed into the DME.

The DME monitors the O2 sensors and makes small adjustments (up and down) in air/fuel mix to find the optimal point. This is short term fuel trim.

The long term fuel trims are averages of the short term trims.

Trims are stored in non-permanent DME memory and will be lost if the DME is reset (and then adaptation has to start all over).

Originally Posted by Robert986

2) MAF CALCULATIONS
As discussed in other threads, when using the 987 MAF-tube the DME doesnt know of the larger area of the 987-tube and calculates to little airflow leading to a lean state. If I have understood this correctly a pro tuner will be able to adjust this in the DME. How is this done? Is it a simple constant in the DME-calculation that is adjusted to represent the new area or are the whole fuelmaps adjusted?

The pro tuner will create a new map by monitoring the air/fuel mixture using an exhaust probe while the car is on a dyno to find the optimal performance at varying engine rpm, loads, etc. This new map will be written into the memory of the DME, creating a new map.

Typically, the entire map will be replaced but some points in the new map may be close to the original map. It really depends on how aggressive the tune is and what changes were done in the engine (differing from stock).

Originally Posted by Robert986

3) MAPS
A very wild guess is that there are three states of operation:
a) Startup/Idling: I would guess that this is managed by a fixed map in a open loop condition.
b) "Middle ground/half throttle" I guess there is ONE singel fuelmap covering this area operating in "Closed loop" thus adjusting STFT/LTFT based upon feedback from the o2-sensors.
c) WOT (Driving like a man) I guess this is covered by a separate fuelmap operating in open loop.
d) Finally I guess that also the ignition is managed by separate maps just like the fuel.

Its more complex than this. There are at least two start routines (cold start and hot start) that have different parameters within each routine depending on the monitored variables.

Start is open loop, warm operation is closed loop, and WOT is open loop.

Yes, ignition timing is map-based just like fuel.

Originally Posted by Robert986

4) CODE/PASSWORD
Is there a password to get into the Motronic that we need to get from the dealer? Then how do the pro tuners do? I have got an offer to tune my car and I don´t think they need a code from the dealer.

Some tuners hack the DME and re-progam over the existing data (its really not that hard). Others will piggy back an external controller onto the stock DME to intercept the signals from the stock DME and modify them before they are sent to the engine. Or some tuners will replace the stock DME with an easy to program/modify aftermarket unit.

Go and buy this book:

https://www.amazon.com/Bosch-Fuel-Injection-Engine-Management/dp/0837603005

[Robert986]

Brilliant answers thstone, thanks!! Will look into that book!

Some more reflections:

"The OEM map is static" -So, does this mean that there is only ONE fuel map used in all states of operation? I was expexting at least a separate map at WOT, giving a bit fatter mix for cooling.

The main reason for my interest is if I would be able to use the 987 MAF-tube (as a part of switch to the 987-airbox and 997 throttle/Plenum) without a re-map. Let´s say I manage to convert the MAF-signal so the voltages going into the DME represents the actual airflow.. Would that be enough? I mean then the DME would get the correct airflow into it´s calculations, the difference should be that the new intake system would allow for a better airflow, giving more air. Then I take it that the STFT+LTFT would compensate so it would all be working.

However.. Since the imroved airflow + extra fuel will give more power, would I then need a fatter mix at WOT for more cooling? Or will everything be ok?

[thstone]

Originally Posted by Robert986

"The OEM map is static" -So, does this mean that there is only ONE fuel map used in all states of operation? I was expexting at least a separate map at WOT, giving a bit fatter mix for cooling.

You are correct, there is more than one map. What I meant by "static" is that the maps don't change. They are pre-programmed at the factory and stored in permanent memory.

Originally Posted by Robert986

The main reason for my interest is if I would be able to use the 987 MAF-tube (as a part of switch to the 987-airbox and 997 throttle/Plenum) without a re-map. Let´s say I manage to convert the MAF-signal so the voltages going into the DME represents the actual airflow.. Would that be enough? I mean then the DME would get the correct airflow into it´s calculations, the difference should be that the new intake system would allow for a better airflow, giving more air. Then I take it that the STFT+LTFT would compensate so it would all be working.

This makes sense. The question is whether stock DME has enough range in STFT/LTFT to accommodate the additional air from the 987 intake? For example, if the 987 intake flows 20% more air, the STFT/LTFT will have to be able to adjust the trims by +20%. That seems like a lot for a system that is meant for fine tuning engine operation. However, if its only +2%, that might be possible. I don't know the answers to those questions. I'd suggest searching to see if you can find someone who made that mod and ask them.

Originally Posted by Robert986

However.. Since the imroved airflow + extra fuel will give more power, would I then need a fatter mix at WOT for more cooling? Or will everything be ok?

I'd have to look at the book that I referenced to remember how the ECU handles the switchover from closed to open loop for WOT. My gut says that it takes the closed loop trims and then fattens up the mix even more (that is how I'd write the algorithm!). I'll take a look today, with the virus I'm just hanging around the house and this is way more interesting than cleaning my workshop.

[Robert986]

Thanks thstone!

I do plan to take it to the tuner for a re-map, however if one can convert the MAF-signal eg via an arduino to make it all fly, that would be neat for everyone wanting to switch to the 987 airbox/MAF-tube without having to re-tune.

[jsceash]

Originally Posted by Robert986

Thanks thstone!

I do plan to take it to the tunder for a re-map, however if one can konvert the MAF-signal eg via an arduino to make it all fly, that would be neat for everyone wanting to switch to the 987 airbox/MAF-tube without having to re-tune.

There is a signal range returned from the MAF to the ECU. This value is compared to a table for air flow so to speak, each millivolt or milliamp has a value. So the return signal triggers an exchange in the mix calculation. So in the end the signal has a range in the ECU and a maximum value in the table. The table has to be changed.