MembersPage/FSTeamDelft

Hello,

Welcome to the Delft University of Technology Racing Team memberspage.

1185186947_endurance2.jpg

We are a Formula Student Recing team. We compete in the UK and Germany events and we are also present at the Baltic Open event. We are known to build the lightest car in the competition. We were weighed @ 125,5 kg in the UK in 2006.

We are investigating in the VEMS system to use it in the future instead of the MoTeC? system.

We are working on a Traction and Launch control system with four wheel speed sensors on the VEMS or with an external system communicating with the VEMS.

Chassis:

Full carbon monoocoque

Engine:

Yamaha wr450 one cylinder.

5 valve cyl head

Power output currently 55 hp with 20 mm restrictor

Current manamgement system: MoTeC? M4

36-1 trigger wheel with v/r sensor

1 injector aimed @ the intake valves

Coil on plug ignition coil

5 bar fuel pressure

have a look @ www.dutracing.nl

Best result so far: 2rd overall in the Formula Student UK event. (www.formulastudent.co.uk)


Trigger problem

We have a problem with the trigger wheel. It is a 36-1 trigger wheel that worked fine with MoTeC?. The RPM reading is too high (700 instead of 550). We cannot check the timing at the moment.

I do not have the configuration file, because the dyno-pc does not have an internet connection.

I have the following settings in megatune:

Rising edge,

Multitooth,

Filtering disabled,

Advanced filtering disabled,

Special N-1 ENABLED,

Number of teeth on wheel: 35

Trigger tooth: 1

Next trigger tooth: 0 0 is BAD ! Use FF if you don't need a "next trigger" event

Crank min period: 768 us

The timings percentages are as in the help, 75-150 and 150-250 (but these only matter with Advanced filtering enabled)

The constant for RPM calc is 12000

If I set Next trigger tooth to FF, I have a trigger, but I have to enter a RPM constant of 6000 instead of 12000 (we have one cylinder). Is this correct?

I keep having problems when I configure the VEMS as one cylinder.

The ignition dwell time and ignition time changes unacceptable much from cycle to cycle.

At the moment, I have the VEMS configured as a four cylinder with injector divider of 4 and the 'wrong' ignition evens output to an unused ignition output. This works o.k., but it needs to be noticed that the injector divider does nothing when cranking, so the cranking pw should be 1/4 of the 'real' pw.


Ignition and Fuel Pump cross-talk

When I enable power to the vems, it makes the ignition driver high (no pull-down) for the same time as the fuel-pump powerup. If I change the fuel-pump powerup time, the ignition driver time changes the same way.

This is a problem since we have a coil-on-plug that accepts only a few ms dwell time, or it will die.

It has a high input impedance, but it needs a true 12v/0v signal, so I use a driver and pull-up resistor. It charges on a high signal and ignites on a negative edge.

I tried different ignition outputs, it is currently connected to an injector driver.

I tried different fuel-pump outputs, but it doesn't help. Even when I disable the fuel-pump output, the problem still exists.

Since I tried so much inputs and outputs, and the problem only exists on outputs that are configured as ignition output, I am quite sure it is a software problem.

I also tried different firmwares, 1.1.18 and 1.0.73

I will upload my config soon.

I wonder if this is a known problem?


Crash Problem

I have a very strange problem with my VEMS.

After upgrading to 16x14 firmware and setting all the tables, my VEMS crashes after about 5 minutes running. When the engine is not running, the VEMS does not crash

A crash means: No communication with Megatune, all outputs high.

I downgraded back to 12x12 firmware and loaded my old config file, but the VEMS still crashes.

I don't see any strange values in the Megatune log.

Is there a known problem that makes the VEMS crash like this?

There is no such known problem.

Some windows-USB-serial adapter driver combinations can sometimes crash MT (so that windows reboot needed) if ochblocksize=62 ('D' command) but they usually work with ochblocksize=56 ('A' command). Controller does not crash though.

Maybe this behaviour is somehow special to your configuration.

I am thinking about turning on the Watchdog timer, because when the vems is crashed, it leaves the Injector open. I wonder why the watchdog timer is not implemented in the firmware?