Oskar Nilssons Audi S4
Technical data
Audi S4 2,2t -92, AAN-engine without modifications so far, VEMS v3.3 with Motronic connector, box no. 1045.
Status
081019
After a couple of months in the garage the car is now updated with a new turbo, injectors and exhaust manifold. I've changed the injectors from the original 349cc/min injectors to Siemens Deka 630cc@3bar. Just changed the req.fuel-value from 13ms to 7.2ms. The engine started with no problem and ran smoothly. After 10km of driving without problem the car started to run on only 4 cylinders.
After some fault-finding I discovered that the injector on cylinder number one was open all the time and filled the cylinder with gasoline quickly. Tried with the old stock injectors but the same thing happened on cylinder number one again. Something has to be wrong in the VEMS-box.
What can have caused this to happen? I'm using the original cable harness and haven't modified anything on the engine besides a new turbo and larger injectors. Since the install of VEMS I've been driving approx. 300km, no problem at all with the old setup with completely stock-engine. Would be grateful if anyone could help me out with this annoying problem.
- You're not the only one. I've experienced similar problem with a car running over 5000km without problems, and nothing has been changed on it, one day it started to run very rich... This unit was sent back to the HQ to repair. The relevant FET driver IC and injfet were both changed and problem still persists. Probably the CPU pin is damaged (for unknown reason). I've got a call these days from another customer with very similar problem but didn't get that ECU to my hands yet. Annoying.
Measurement of cable harness to injectors
Did some measurement of the cable harness and the ECU to recoqnize possible sources of the fault.
Injectors and idle-air output measurements
Measure injector and idle air pins against AANpin37 (that is the flyback pin, same function as EC36pin23 in standard VEMS docs)
- inj cyl1, AANpin17, EC36-pin20, INJ-D -> 12.0 Ohm
- inj cyl2, AANpin16, EC36-pin8, INJ-C -> 12.1 Ohm
- inj cyl3, AANpin36, EC36-pin9, INJ-E -> 12.0 Ohm
- inj cyl4, AANpin35, EC36-pin19, INJ-B -> 11.9 Ohm
- inj cyl5, AANpin34, EC36-pin7, INJ-A -> 11.9 Ohm
By measuring directly on the injectors I got values between 11.9-12.0 Ohm.
Grounds
Measure that grounds are strongly (max 0.1 Ohm) connected inside the VEMS ECU.
Harness:
- AANpin10 to AANpin19 -> 0.5 Ohm
- AANpin19 to AANpin24 -> 0.6 Ohm
- AANpin24 to AANpin10 -> 0.5 Ohm
- AANpin30 sensor ground -> -
- AANpin48 trigger ground -> -
ECU:
- AANpin10 to AANpin19 -> 0.3 Ohm
- AANpin19 to AANpin24 -> 0.3 Ohm
- AANpin24 to AANpin10 -> 0.3 Ohm
- AANpin30 to AANpin48 -> 0.3 Ohm
Measure ECU pins: verify that all the above grounds are connected inside the ECU.
Measure that first 3 are connected in the harness. AANpin30 (sensor ground) and AANpin48 trigger ground are likely not connected inside the harness.
I used a cheap DVM with a measurement range of 0-200 Ohm, is 0.3 Ohm a to high value on the grounds inside the ECU?
Internal flyback testing
Found the information about this test here http://megasquirtavr.sourceforge.net/manual/Detailed.Output.Flyback.html
The first test must be performed to ensure that the internal flyback connection is in place and working.
- Make sure the Genboard module is powered off.
- Disconnect the first injector.
- Connect DVM negative/common to injector connector common pin (the wire that runs to EC36-pin23).
- Connect DVM positive to the other injector connector pin.
- Measure voltage drop with DVM (in diode mode).
The measured voltage drop should be approximatly 600-1500mV. Repeat the test for the other injectors.
- injector1 -> 1661 mV
- injector2 -> 1698 mV
- injector3 -> 1698 mV
- injector4 -> 1698 mV
- injector5 -> 1698 mV
When I swapped and measured with DVM positive to the injector common pin and DVM negative to the other injector pin I got these values.
- injector1 -> 1193 mV
- injector2 -> 0 mV
- injector3 -> 0 mV
- injector4 -> 0 mV
- injector5 -> 0 mV
I assume this value should be 0 mV for all injectors. Is the flyback for injector1 inside the ECU broken?
Is 1698 mV a to high value on the other flybacks or is it in the range of what you can expect to get from a new flyback?
The engine has been started and given a base VE tune up to 0.6 bar boost, this process took less than 15 minutes after engine was warmed up using the wideband.
The installation was very smooth and it started up right away thanks to great configuration tips and files from MembersPage/MiskaPeippo and MembersPage/TeppoKalske.
More work is needed on the boost controller settings and idle control. Both worked out of the box but two things needs work with the settings:
- Only 0.6 bar boost (solved, read below)
- 1700 rpm after hot start, went down to normal idle after 10+ secs.
- Idle control doesn't catch the engine fast enough
The boost-problem is solved. It was caused by a huge leak in the rubber hose between the turbo and the intercooler. Tried with open loop to control the boost solenoid manually and with the solenoid completely closed to get maximum boost I still only got 0.6 bar boost. So Miskas configuration for the boost control seems to work. I'll test it when I've fixed the leaking hose.
08.04.03. Have fixed the leaking hose and now the boost is fine when I use open loop for boost-control. With closed loop the boost was very instable. The boost pwm table for closed loop was messed up, just need to fix the table with the values that worked in open loop. The quick response in boost impresses me, full boost comes much quicker now compared to my original 280 hp chip-tuned ECU.
Software configuration
I used 1.1.27, it is the latest stable version to use (as of 2008-03-12). The tables are 12x12, which seems more stable and gives good control over this very easy to tune engine (stock cams).
With some documentation I hope this can be pre-loaded in the pnp Audi boxes delivered.
Latest config 2008-05-06
- Stock engine tuned to get most power
- Changed ignition map to get higher advance and more power
- Cold start needs tuning
- Runs well for daily driving
- Open loop boost control works ok, closed loop not so ok
Config 2008-04-22
- For a stock engine
- Cold start needs tuning
- Overall runs well for daily driving, but not tuned for max power
- Open loop boost control works ok, closed loop not so ok
- Idle air control valve works adequately
I made up a start configuration from various sources :
- I have config/tables from Jörgen:
- An MSQ from an RS2 from one of the Finnish guys:
- The lambda table uses AFR (In my opinion wrong) and the spark map seems to use (too) high advance on load.
- Miska Peippo sent an MSQ for AAN using firmware 1.1.27:
- Very nice config with boost control and idle air valve control
- One must use a special vemsv3.ini to get control of the special boost control settings that Miska has set up.
Deviations
In the majority of installations it's very probable that the stock IAT sensor which is of PTC type will be replaced by the web shop IAT that uses standard NTC type. This requires changing the on-board pull-up on the analog input for the IAT back to the standard 2.7K.
This makes the installation less pnp but everyone seems to agree that the stock PTC can be replaced with favourable gains in response and accuracy.
The stock PTC sensor is used for now, but will be changed when the engine is upgraded (intercooler, turbo, exhaust manifold).
Wiki links to Audi information
PlugAndPlayMotronic
VemsAudiAan
MembersPage/MiskaPeippo/AudiSSix
MembersPage/JorgenKarlsson/AudiStwo
MembersPage/TeppoKalske
MembersPage/MarcellGal/PowerAndTraction/PnpHarness
MembersPage/MarcellGal/PowerAndTraction/IdleAir
MembersPage/MarcellGal/PowerAndTraction/BoostControl