Mshift confuses MPH and Km/h?

[Philip Lochner]

Today I went for a short drive and immediately it was clear to me that MShift was wayyyyy over optimistic with vechicle speed. The ratio seemed familiar and sure enough when I got to 50km/h on the GPS, Mshift was reading 80Km/h. Stopped, restarted and voila, all normal. Hope this will not be a frequent thing.

Just thought I should mention it FWIW.

[Bernard Fife]

Philip,

As is sometimes the case, the apparently obvious problem isn't the the actual problem.

As of 4.123 code, the MShift code doesn't know anything about converting metric/imperial units. *All* internal calculations are done in imperial units, and Tuner Studio handles all of the conversions based on the INI. See: viewtopic.php?f=4&t=240&start=60#p3381

It is possible that TunerStudio is mishandling the units conversion (especially if you had just changed the units settings and not re-started TS). It is also possible that the INI has become mangled (though it would be unlikely to fix itself if that was the case).

However, I think a more likely cause of a low speedo reading is input masking not set up to match your sensor and wiring, and as a result a significant portion of true pulses are being rejected.

Lance.

[Philip Lochner]

It is possible that TunerStudio is mishandling the units conversion (especially if you had just changed the units settings and not re-started TS). It is also possible that the INI has become mangled (though it would be unlikely to fix itself if that was the case).

However, I think a more likely cause of a low speedo reading is input masking not set up to match your sensor and wiring, and as a result a significant portion of true pulses are being rejected.

Lance.

Thank you Lance, really appreciate your willingness to help.

Drove the car more today and found that speed readings seem to vary from pull-away to pull-away.... Without restarting TS, which suggests that something is amiss in MShift. It have not yet noticed that speed readings vary while driving but I'm certain it varies from pull-away to pull-away - and not every time either.... BTW, I'm using the old algorithm. Just seemed more confidence inspiring than the "BETA" nature of the new one. I figured I need reliability now, not uncertainty.

I'll have another read of VSS part of the manual, then go play some more.

[Philip Lochner]

OK it certainly varies from pull-away to pull-away, and I also see it change in speed calibration when it shifts into 2nd (which is still rather hard - have not yet been able to soften it.)

Its not as much that the speed is wrong, its the inconsistency that baffles me.

Also, tried the beta ISR but that seems to get stuck at very low speeds ( < 5km/h) , if I set masking higher, the needle starts jumping around. Quickly gave up on the beta ISR and went back to old ISR.

I'm going to reflash the code and try again....

[Bernard Fife]

Philip,

If you post your most recent msq here, I will try it on my bench and see if I can reproduce your issue.

Lance.

[Philip Lochner]

Thanks Lance

MSQ attached.

I tried reflashing the code but then CAN would no longer work and I could not figure out why. I was low on patience at that moment and then flashed the Extra Transcontrol code. That code was better in the sense that it was accurate and stable BUT NOT AT extremely slow speeds (<4Km/h) where its erratic.

So I'm still without a solution...

[Philip Lochner]

FWIW, here is a scope trace of the VSS input to GPIO with rear wheels raised, transmission in N, engine running and the wheels turning ever so slowly on their own.

[Bernard Fife]

Philip,

I loaded your msq and rebooted MShift, and with a VSS input of 1606 Hz I got a speed of 100 kph which seems about right. The speedo is smooth and responsive to changes in the input frequency over a wide range. I have been unable to provoke any other displayed speeds at that input frequency, but I will keep trying.

One thing that will screw up the speedo is an intermittent signal (i.e. a single that drops out completely) as this interferes with the masking and can lead to multiples of singles being dropped once the signal is restored. However, I can only accomplish this on my bench by changing the signal by 10x instantaneously. I suppose this could happen in a vehicle if the wiring connections (or the VSS circuit build) were not perfect. Since you had trouble with the VSS when using the stim, you might want to check that your VSS circuit is correctly configured.

Lance.

[tobmag]

Phil.

Long time no seen

I have been in the same place as you are right now and many otheres with me.

What you should be looking at is actualy the signal AFTER the VR conditioner going into the processor pin.
Its easily accesible on the header.

I assume you have alot of logs so look at them and look at the VSS signal there, most likly it jumping all over the place.

I also learned that a clean VSS signal from the sensor wont solve all problems, I had to use another VR conditioner to sort it out!
Drawing is avalible here on the forum just search on "tobmag"

My XJS is extremly noicy in the whole electric system so I needed to add some filters on the positive feed lines as well as it also affected the VSS!

BR//Tobmag

[Philip Lochner]

Right, Ive got speed sensing sorted. I've got a TP500 module doing the conversion from the sinusoidal VSSensor output to 12V square wave which is then fed to GPIO as well as the VDO speedo gauge:


When the speedo still went haywire above 50km/h, I realised that something was wrong with the TP500 setup and that is when a friend of mine raised the possibility of the TP500 saturating.

I think I had 2 main problems:
1) my TP500 module saturated above 50km/h. The input signal voltage became too large for it and it stopped became erratic. The 220Ohm resistor between the two wires loads the sensor output and reduces the output level of the sensor keeping it to a level that the TP500 can cope with.
2) I might have had a dry joint on R45 - hard to tell though with all the component replacing that I was doing.

I modified the VR1 input circuit to be more suitable for a square wave, but now that I found the TP500 issue, I'm not sure if the mod was needed at all.

My car is now, at last driveable and I'm beginning to find my way around the transmission tuning.