History of MembersPage/GergelyLezsak/MoriniMeasurement
Older Newer
2022-12-09 21:32:45 . . . . t-vpn.t-systems.hu [Morini logger project initial and questions]


Changes by last author:

Added:
Moto Morini data logging project

This page is about datalogging with VEMS V3 on a Moto Morini Corsaro 1200 from 2006.

Bike is working OK and running normally, but there are weak spots in fuel mapping. I'm going to make a wideband lambda datalog with VEMS ECU SDCard option. RPM and TPS signals also need to be logged.

Engine details:

* 1200cc four stroke V2 engine, 4 valve per cylinder. No VVT

* Ignition: one coil per cylinder

* Fuel: one injector per cylinder

* Trigger: 48-2 teeth trigger wheel on CAM (front cyl.)

* Weber-Marelli IAW-5AM ECU, very common on Ducati, Guzzi, Fiat Punto, etc.

* Stepper motor idle regulation

* Narrowband O2 sensor, standard TPS, CLT, IAT, atmospheric pressure sensors

* Alpha-N logic, normally closed loop fueling operation (probably at light loads)

Clone install details

Original Marelli ECU operation need to be untouched during the logging.

* Lambda - I have a race map on the bike which has open-loop fueling, so I had no problem with removing the original narrowband Oxygen sensor and install VEMS Wideband sensor instead

* TPS - piece of cake: signal wire shared between the two ECU. 5V supply and GND comes from Marelli ECU

* CLT - this signal is not so important, but now it's working shared. I use a VEMS ECU without pullup resistor to not interfere with Marelli ECU measurements

* MAP - no such signal on Marelli ECU, so idle air pipe split to VEMS MAP sensor

* RPM - Trigger - I had no luck sharing the original VR signal. The problem is that original ECU uses a VR sensor which has +- polarity wired in (two wires to Marelli ECU), and not grounded anywhere. VEMS has one wire VR signal design relative to GND and has some pullup resistor(?). Connecting VEMS to the signal line always ruins the original ECU signal. (alternatively: when original ECU triggers, VEMS cannot get a strong-enough signal). Tried to share with series capacitor isolation, resistors to limit current, -many arrangement, no luck. Working ideas welcome!

* RPM - failover method: this works from original injector output signal, connected to VEMS HALL trigger input through a diode. It is OK, but with limitations: one signal per every CAM rotation - so RPM is halved, I couldn't set up VEMS config otherwise yet. No signal during deceleration fuelcut, inaccurate RPM reading during sudden changes, especially when injection open times and start timing also change.

Pictures

https://drive.google.com/file/d/1OCFP46jUkKWVIGnSi1jUtYqEAZKQRd_l/view?usp=sharing

https://drive.google.com/file/d/1OD-efesTKJgXxtX1Gy7CjZLNt437hWP0/view?usp=sharing

Config

I'm still looking for the right VEMS trigger configuration to see proper RPM.

Currently I have one trigger event per CAM rotation. Current settings:

*No. of cylinders=2

*Trigger=Coil type

*Special trigger=Normal

*TDC after the trigger=60

*#of teeth of full engine cycle=2

*first trigger tooth=0

*sec. trigger=disabled

*Next trigger tooth=7

This ended up half of the real RPM consistently. "#of teeth of full engine cycle" should be 1, but that results an error.

And ideas?