Modifying Motronic 2.3.2 ECU hardware and software.

[Tille]

I'm in love. Motronik with perfekt Knockcontrol + MAFless + Tunability VEMS like. I would take one, maybe two. Even if it will be more expensive than VEMS i would chose this solution.

[Thuppu]

Just WOW!!!

[-ice-]

Hello

I'm currently rebuilding my engine, so i made a break with disassembling.

But I restarted with the boost/knock chip this weekend. There is also an "adress table" for the maps (0x2000) and with looking at the map read procedure i'm able to tell which map is read. Interesting of course is the boost control procedure where also the target boost map and n75 map is read. There are many other maps in this procedure, i have to find out really not easy...
I've looked at the "target boost" map with WinOLS. The format is 16x10 which is easy to see...So whats the best way to find out the meaning of the axes? Since we know that one axis is rpm i devided 7000rpm by 16 and the TPS axis per 10 to get a "unit". To convert the 8Bit value into a pressure i simply took the maximum pressure of ADU engine and the highest point in the map. Does anybody know other conversion factors?

Regards

[vwnut8392]

I know the only thing i'll probably hear is crickets on this but does anyone know how to disable the lambda sensor heater circuit weather it be in the ECU code or via a resistor? im going to run an innovative LC-1 wideband now and use one of the analog outputs to feed the ECU from that since the analog outputs are fully configurable in the LC-1.

[Acki]

Is this needed for the I5? At V8 motronic the ecu only switch a relays.

[Billman]

Cat is out of the bag huh.

I'll make an announcement when I have time.

The hardware schematics are finalized, the board needs to be designed and I am manufacturing a prototype soon, then work can start on the software.

Basemaps - yes.

What was omitted from the posted feature list, was stuff like built-in direct ethanol sensor support, wideband / egt control, 4 bank map switching, possibilities to add logging and safeguards for oil temperature/oil pressure/fuel pressure and so on.

The price point is about right 750-800$. The components to make a board that fits into the ECU and can emulate both the fuel and the boost chip at the same time in full realtime are very expensive. And I would not settle for anything less. Flashing? Give me a break.

Let's put it that way, I don't have any expectations to get "rich" off of doing this, and I've been working on the concept for quite a few of my weekends now... but I just will never have time to tune every single 2.2T out there, and if I can put all my knowledge into an easy accessible package then I have succeeded in my goals.

As for standalone ECU's. There are a few concrete limitations in this ECU and I will be upfront and honest about that right away. But it's not really anything people under 1200hp have to worry about. It still should be the most cost effective and solid way to tune anything up to 1000hp.

Very cool! Keep it up!

[vwnut8392]

Is this needed for the I5? At V8 motronic the ecu only switch a relays.

there is a fault code in the list for it. so i figure if the heater is taken away this is the fault i will get. my car currently has the cylinder head off of it so i cant do any testing till i get it back together.

# 2342 (Blink Fault Code), 00525 VAG1551/2 Fault Code Number
Oxygen Sensor (G39)
Sensor Wiring or Sensor defective Heat resistor in sensor defective, Signal wiring shorted to ground, Fuel tank empty, Ignition system component failure, air leak at mass air flow sensor, air leak in exhaust pipe before catalyst, poor ECU ground wire connection at intake manifold.
Symptom: Emissions not within specifications, increased fuel consumption, rich exhaust, spark plug fouling,
Carbon Monoxide (CO) percentage above 3%,

[prj]

No fault for heater but with simulated narrowband you will have severe issues on start if you do not build in a delay.

[vwnut8392]

No fault for heater but with simulated narrowband you will have severe issues on start if you do not build in a delay.

thank you dimitri. would something like a timed relay work to stop the signal from getting to the ECU for a short period?

[prj]

The delay must be built into M2.3.2 SW.

[vwnut8392]

well that does me no good than. i'll set it up and mess with it for awhile to see if i can come up with a solution that doesnt involve writing custom code in the ECU.

[chris..]

Any updates on the PRJ super chip system?

[domas]

I know the only thing i'll probably hear is crickets on this but does anyone know how to disable the lambda sensor heater circuit weather it be in the ECU code or via a resistor? im going to run an innovative LC-1 wideband now and use one of the analog outputs to feed the ECU from that since the analog outputs are fully configurable in the LC-1.

In my 200 20v, there is no such code for lambda heating circuit failure as there is in V8. Mine is heated from ignition directly, from FP relay iirc.

What I have done (even though it should not matter in my case) I am feeding the main voltage supply to WB controller from original lambda heating circuit. My reason for that is that I wanted everything as plug and play as possible - i.e. there are 3 wires going to WB controller now from original lambda connectors (heating circuit for powering WB, and emulated NB return wire from WB controller)

As WB has to be heated as well, there is load on the circuit. Would that possibly work for you? Or is the heating output not constant 12v in your case..?

As far as I understand the only concern is lambda fault code that you want to get rid of, right?

[prj]

There will be no lambda codes. The ECU has no idea about the heater either.

The issue is there is a plausibility check done by the ECU before it considers the signal valid and O2 feedback control is enabled.
This plausibility check passes once the o2 heats up.

The issue with a simulated signal (besides it being imprecise usually) is the fact that the plausibility check passes instantly, and the ECU starts using it right away, even while the wideband is still heating up.
Because while it is heating the simulated signal is usually stuck on one value, the ECU rapidly either fully leans out or fully enriches the mixture and the car runs like **** until the wideband heats up.

To solve this issue, I put in a timer into the ECU software, which introduces a fixed delay after engine start before the plausibility check is allowed to pass.

[domas]

I see. That may be a problem I have myself then.

IMHO it should be considered as wb controllers that are not smart enough. It is strange that this function is not available out of the box.