Huge RPM Limiter Problems

[LacknL]

Hello Guy's!

In the jrX the, newest Formula Student car from joanneum racing graz works a turbocharged rotax single cylinder with a Motec M800 ECU. Since a few days we have huge problems with our RPM Limiter.

We desided to set the RPM Limiter to 8000 rpm because otherwise we could overspeed our turbo.

But the RPM Limiter doesn't really work pretty well. When we reach approximatelly 5500 rpms Motec starts with RPM cutting.

Could come this affect from a poor signal from the ref sensor. Or is something with our M800 wrong?

I also attached you our latest Log File where you can see our sensor values and also the strange behavior of the RPM Cut.

If you need more informatins about our engine package or about our motec file don't hessitate and ask! We are very happy about any idea for finding that bug.

Many Many Many thanks in advance for any idea or solution!

Kind Regards
Luke

[NathanS]

Hello Luke,

The file that you have uploaded is the .ldx file, this only contains information about any changes that you have made to the log file since opening it in i2, it does not contain any of the actual logged data. Could you post the .ld file please.

[LacknL]

Oh I'm sorry!

Thanks for trying to help us Nathan!

Kind Regards
Luke

[Mazi]

[Sean]

:shock:

What they said

I am really interested to see what Nathan comes back with.

[NathanS]

Hi Luke,

I've tested your ECU file on the bench and the limiter behaves as it should, what I did notice from the logging though is that the RPM trace is very noisey and the cuts are occurring when there are sudden increases in RPM. For the next time that you run the engine can you add in some of the diagnostic channels to your logging so that we can check that everything is ok from the crank and cam sensor side. If you add in Diagnostic error groups 6 and 14, these will highlight any ref sync errors that occur.
Do you have a Ref/Sync capture for the setup that you are running on the engine?

As an a side I spotted a few things in your ECU file and the logging that you might want to check:

1) The lambda trace for this engine appears to be on the lean side especially as it is a turbo'd engine.
2) The Fuel MAP compensation jumps up at 300kPa.
3) The Primary Injector compensation is set to 0 at 8V and below which means that the injectors won't open for the required amount of time should the battery voltage drop this low.
4) There is no filtering being applied to the MAP sensor or Air Temperature values and in the ECU file that I retrieved from the logging and quite a few of the sensor default values had been set to 999.

[LacknL]

Hello Nathan!

First of all, sorry for the delay of my answer! We were in the last few days on the testtrack with our car, so I had no time to check the forum.

I think the next testing day is already tomorrow, so I will ad this two channels to the data logging. I hope I'll understand you right, do you want to have a ref/sync capture of our engine or do you mean if we have a ref/sync capture from a engine which runs well and so we can compare both captures?

In the mean time, I'll found a tricky but simple way how we can compensate the rpm cuts. I loaded a old file into motec were I knew that it works well and then the problem with the rpm limiter didn't exists any more. The only difference between both files was the value of the limiter. With 9250rpms the file works well, with 8000rpms we se the rpm cuts.

The reason for the value of 8000rpm is that our VTG doesn't work pretty well, so we wanted to reduce the engine rpms, so we can save our turbo for overspeeding. A better solution for the VTG is a external wastegate which is already in production

Now back to your points:

1) When we went to Silverstone to the Formula Student UK event we suddently recognised that the engine runs to lean. So we increased the Fuel Overall Trim. At our testbench we had at full load a Lambda value of 0.9 which was our target to reach. For part load and idle rpm we want a lambda value of 1,05 which we also got at our testbench.

2) We will never be able to reach that amount of boost, because the turbo can't do this. Our maximum boost with this configuaration: Throttle Valve, Restrictor, Turbo, Airbox is under 200kPa. Do you think we should increase the value or what would you advise us to do. I'm very happy for all tips how we can improve our motec file.

3) Normaly when we reach a voltage drop down to 8V then our car will not work any more. Our normal battery voltage level during the endurance is approx between 12,2V and 11,5V. But for safety reasyns I can add a compensation at this point.

4) For finding the error with the rpm limiter i turned off the sensor filtering. I thought that there could be maby a coherence between a sensor and the cut. Which sensor filtering values would you prefer for the common sensors? For the default values I took the values from our previous team. Is the default value for the normal operating point of the sensor?

Thanks in advance for your help Nathan!

Kind regards from Graz
Luke

[Sander]

Is your oil pressure channel assigned to something other than oil pressure? Your oil pressure values are strange looking to say the least. At 1:00.332, you have 7000 rpm and 23.5 psi?

If you're running in to limter problems, could it be that you are interpreting a limter issue to be just a lean running condition? Those lambda channels look very very lean relative to manifold pressure and rpm.

Just curious.

Sander

[NathanS]

If you have a ref sync capture from when the engine was setup and one from now that would be good because you can then check that nothing has been changed such as the wiring for any sensors and there has been no deterioration in signal quality.

What RPM is the engine being run too with the 9250rpm limit?

1) I've seen quite a few times on Formula Student cars that the engine runs differently on the dyno to when it is installed in the car. Are there any differences in engine configuration between running the engine on the dyno to running it in the car? Do you take into account the variation in air flow over the intake on the dyno?

2) At 300kPa the fuel compensation should be 200% not 300%. Something that you need to be aware of is the fact that the Fuel MAP comp table extrapolates the data off of the end of the the table which can cause problems if you haven't set the sensor defaults correctly and a sensor goes into error.

3) I would recommend filling in all of the cells in the battery comp table, it is better to have something in there rather than setting it to 0, there maybe an instance where the battery does drop that low say during starting in which case the injector time will suddenly be reduced.

4) The filtering that you need to use for the different sensors will depend upon how good your sensor signal is, a value of 1 turns the filter off, 10 gives a medium level of filtering and 50 gives a high level of filtering.
Typical values for the filtering to use could be:
MAP - anything up to 4
ET - anything up to 10
The amount of filtering is a trade off between smoothness of the signal verus its response time.

As for the default value, this is the value that is used if a sensor goes into fault, so at the moment if your TP, MAP, ET or AT sensors goes into fault their readings will go to 999.0, which in the case of the MAP sensor will add in a lot of fuel. If you set these to normal operating values it may mean the difference between finishing the endurance event and not.

[Sean]

Luke, is there a specific reason you are cutting the fuel as well as the ignition on the shifts? I thought that in nearly all cases cutting the ignition only is best. Looks like in your case the ignition is back on with the fuel still cut, which gives that nice lean spike as the fuel (and load) comes on.