MPGuino VSS Pulses/Mile confusion

[t vago]

Originally Posted by sallen

Thanks again - Settings work like a charm - have input the VSS and mL/min and hope it will be good to go for my next drive.

One final Q the main MPGuino thread says your version "supports both metric and imperial units" - I might be thinking my instant is off because it's giveing me (higher than I expect) US MPG, when I'm used to UK mpg.

I looked through the code (I'm no arduino expert) and didn't see anything to define or chose between US and UK mpg. Is there a figure I can change to the size of a UK gallon, to get UK MPG?

Yah, that was somewhat inaccurately stated in the MPGuino wiki page, although I just added support for Imperial gallons instead of US gallons.

For your purposes, go to around line 2460 in the code, and change this:

	,3785411784ul	// numerator to convert gallons to liters
	,1000000000ul	// denominator to convert gallons to liters

to this:

	,454609ul	// numerator to convert Imperial gallons to liters
	,100000ul	// denominator to convert Imperial gallons to liters

[sallen]

As always, thanks very much.

I'm driving Saturday, so fingers crossed.

[t vago]

You're quite welcome. I'm glad you like the code.

Just a heads-up - this new code has an alarming tendency to lock up on occasion. It appears to lock up any time that the code reads really, really small amounts of fuel, and I think it has something to do with the speed improvements I introduced to the 64-bit division code.

I should have either the fix, or a reversion to the earlier 64-bit division code, available tomorrow.

[t vago]

The 64-bit division bug has been fixed.

Update: The current version of my MPGuino code may be found here.

[sallen]

Thanks - I was just logging on to say it was locking up.

I'll give the new one a try.

In terms of the imperial mod, all that did was give crazy high gal/hr figures (and correspondingly small MPG figures), and broke the autocalculate of the mL/min to (which was coming out very small again).

I changed it to keep the same number of digits (using a several decimal pace version of the UK gal / litres conversion figure.

,4546090281ul  // numerator to convert Imperial gallons to liters  ,1000000000ul // denominator to convert Imperial gallons to liters

...and it still came out crazy high on the gal/hr - I think this was because the microsec/gal figure was autocalculating very small (e.g. 5 or 6 digits).

I self calculated microsec/gal and it was wrong (crazy low this time, bu a factor of about 100 or so). So I guessed that the microsec/gal param was in US gallons still and used the value from lb/hr to microsec/gal spreadsheet.

I'm now using American microsec/gal and the imperial mod in teh code tags above and it's reading sensible numbers. Hopefully enough to make meaningful calibration adjustments at the end of each tank!

[sallen]

Just noticed in the new code that there is a define for imperial - will give that a go.

**edit**

Got it all set up using: a self-input microsec/USgal figure; and the useImperialGallon true defined.

Nice work on the code - No more crashes and it's giving the kind of figures I'd expect from a fuel guzzling german saloon.

I'll run a few tanks through it and calibrate accordingly, then add my vehicle details to the Ecomodder table.

[sallen]

Gave the new code a try and while instant MPG is now calculating correctly - tank and current are way low. Tank even reset itself randomly to 0.

I usually get 20's-30's mpg (UK) out of a tank. Today I was getting this:

https://www.youtube.com/watch?v=LVHm-oVh2bs

[t vago]

Originally Posted by sallen

Just noticed in the new code that there is a define for imperial - will give that a go.

**edit**

Got it all set up using: a self-input microsec/USgal figure; and the useImperialGallon true defined.

Nice work on the code - No more crashes and it's giving the kind of figures I'd expect from a fuel guzzling german saloon.

I'll run a few tanks through it and calibrate accordingly, then add my vehicle details to the Ecomodder table.

If you could, would you provide me the figures you used to arrive at your usec/gallon value? And could you also go to the current data and tank data screens and take pictures of those values? This will help me in troubleshooting.

[sallen]

Here you go,

I used the flow rate spreadsheet from the MPGuino wiki to calculate my ms/gal. My injectors are 22lb/hr (which as an aside - google says is 166ml/min, but the spreadsheet says is 231.3 cc/min and I thought cc and ml were interchangeble 1:1).

The attached image shows the uSec/gal figure I'm using, which is returning a believable instant MPG of around 20-50mpg at cruising speed depending on throttle position. - As another aside, before the 64 bit bug was fixed, and before I tried to "fix" it by using a UK MPG conversion figure the same length and the US one - the instant figure was similarly crazy low like the Current and Tank are now. Maybe they are expecting a number with more decimal places to work with?

The image also shows Current and Tank (which jumped from around 3mpg straight to 0 about 5-10miles ago and has stayed there).

Hope this helps.

[t vago]

Got to look at the issues, and will come back with a fix in a day or two.

231 cc/min is the correct conversion from 22 lb/hr. Not sure how Google arrived at 166 ml/min.

The current data and the tank data screens were switched around, so what is labeled "current data" is actually tank data, and vice versa. You can tell what is actually "current" by the presence of a slowly blinking "c", as with a slowly blinking "T" for "tank". The "current" trip is designed to reset after the MPGuino has gone to low-power mode, and then either waken up by a button press or if it receives a valid injector signal or VSS signal. This is what you're seeing.

As far as the usec/gallon figure, that's a major bug that I will have to tackle. This is why your current and tank fuel economy figures are off, and why it'll take a day or two to fix.