MPGuino VSS Pulses/Mile confusion

[sallen]

Hi there,

I've built an MPGunio and hooked it up to my ancient C-class (vss only for now - I have another thread about the injectors). I'm running "v.1.93tav" version.

It seems to be working (0mph when stopped - some MPH when going) but there are two issues:

(1) It is way off at the stock VSS/mile setting. the MPGuino says 120mph at an indicated 20mph.

(2) It seems to bounce around all over the place (showing 120mph at 20 mostly, but sometimes less (60 or 80mph at a "speedo" 40) - I assume this is the delay setting and have changed it to 0.

It's (1) that I'm having issue with the most. The setup chart for a "group" vehicle of the era (I'm guessign they all used the same ABS sensor) says 160000 for an E-class - My own maths say 166770 is the sweet-spot.

The problem is that the arduino won't save those values. Something goes wrong and it reverts to like 4000. Does the pulses for mile value have some kind of limit? It seems fine if I use a smaller (5 figure) value.

Anyone knoww what's going on?

[t vago]

Yah, I set the number of bits for the VSS pulses/mile count to 16. That results in a maximum value of 65535. That is part of the issue you're seeing.

The other part is that the MPGuino code currently cannot handle raw VSS pulse frequencies higher than about 187 Hz. In contrast, using a tire size of 225/45R17 (which I think is the stock tire size of your Mercedes), and a VSS pulse/mile setting of 166770 pulses/mile, I have calculated that your VSS is sending a raw VSS frequency of about 926 Hz to your MPGuino.

I think I have found a way to beat this within the MPGuino code, and allow your speedo to output the correct speed, but I will need your help to verify whether this fix will actually work.

[sallen]

Cheers - I've left a nice hole in the side of my MPGuino build just for this reason - to access the USB port for uploading updates. Happy to help with fixes.

My C-Class is a UK version, with literally no options specced, so it's rolling on super pimp 15" steelies rather than 17's. Tyre is 195/65/15.

[t vago]

Originally Posted by sallen

Cheers - I've left a nice hole in the side of my MPGuino build just for this reason - to access the USB port for uploading updates. Happy to help with fixes.

My C-Class is a UK version, with literally no options specced, so it's rolling on super pimp 15" steelies rather than 17's. Tyre is 195/65/15.

Here, try this code. I've increased the parameter size from 16 bits to 18 bits, for the pulses per distance parameter, so it should easily handle 166770 now. I also re-coded the data transfer function so it will take a lot less critical time to perform. Hopefully, the issue you're seeing will be solved with the combination of these two changes.

[sallen]

Cheers man will upload that tomorrow and give it a shot...

By way of update - I tried maxing the 16bit (with 0ms delay) and it was more consistent but about 30% high, hopefully the new code can easily cope with that.

[sallen]

Gave this new code a shot - there may be a bug.

I entered 98302 as vss per mile (which is what I now calculate should get me close to accurate) and it accepts, but if you leave the screen and return it had changed to 32766 (roughly 1/3 of what I entered).

Curiously, if I change it back to 98302 it says "param unchanged" - so maybe it has accepted the change but the number displayed hasn't caught up.

Won't get a change to drive with it until tomorrow to check accuracy vs regular speedo.

[sallen]

Just got a chance to road-test the new code and it's not working...

I tried vss/mile values of 98302 and 166770 and both reverted back to lower numbers in the 16 bit range - with significant over-reading of speed as a result.

We are quite close - the MPGuino now seems to read speed consistently. With the value maxed at 65535 I get approximately 30% overread on the MPGuno compared with the speedo. E.g:

Speedo > MPGuino

20mph > 31mph
30mph > 44mph
40mph > 58mph

[t vago]

Originally Posted by sallen

We are quite close - the MPGuino now seems to read speed consistently. With the value maxed at 65535 I get approximately 30% overread on the MPGuno compared with the speedo.

Try this code. Found and fixed a bug in the eepromWriteVal routine, and tested the fix by manually setting the pulses/mile setting to 120000.

[sallen]

Thanks again for the quick turnaround.

I've added the VSS and injector params to get it into the ballpark and will calibrate at the end of the tank as per the instructions here:
Mpguino calibration - EcoModder

Thanks again for all your help, the MPGuino now looks like it will be very useful.

A couple of observations on the new version:

- The uSec/Gallon default value was waaaaaaaaaaay low - like 703 or something, so I used a calculator and input the loooong number for some injectors in a similar model;

- I see there is both mL/min and uSec/Gallon and they are independent - does it use both, or one?

- On the old version's instant MPG, I was getting infinite "- - - MPG" down to 1600rpm, now I get a larger number (100 - 200mpg) or a fixed 0.49MPG in place of the - - -, perhaps some of that is the new settings.

I'm not driving much over Xmas so I will calibrate in Jan and hopefully all will be good.

Thanks again,

[t vago]

Originally Posted by sallen

I've added the VSS and injector params to get it into the ballpark and will calibrate at the end of the tank as per the instructions here:
Mpguino calibration - EcoModder

Thanks again for all your help, the MPGuino now looks like it will be very useful.

Glad to be of assistance.

Originally Posted by sallen

- The uSec/Gallon default value was waaaaaaaaaaay low - like 703 or something, so I used a calculator and input the loooong number for some injectors in a similar model;

Strange. The code should have reset all of the EEPROM stored parameters. Apparently, it did not. Let me know if you have any other issues.

Originally Posted by sallen

- I see there is both mL/min and uSec/Gallon and they are independent - does it use both, or one?

The mL/min parameter is the rating of each individual fuel injector. That parameter, along with the # of cylinders and the reference and system fuel pressures, allow the code to calculate the uSec/Gallon number, if desired. That saves you having to calculate it out yourself. That only goes in one direction, though, so if you enter the uSec/Gallon number directly, it does not affect any of the other parameters.

Originally Posted by sallen

- On the old version's instant MPG, I was getting infinite "- - - MPG" down to 1600rpm, now I get a larger number (100 - 200mpg) or a fixed 0.49MPG in place of the - - -, perhaps some of that is the new settings.

Yah, that's weird. You're probably right, it's probably due to the settings being changed due to the changes in the code. Please keep an eye out for anything that you think is out of the ordinary.

Originally Posted by sallen

Thanks again,

You're quite welcome.