View Single Post
Old 11-21-2012, 10:29 AM   #62 (permalink)
AndrzejM
Master EcoModder
 
AndrzejM's Avatar
 
Join Date: Oct 2011
Location: Poland
Posts: 840

Berta - '97 BMW 318 tds Compact
90 day: 62.03 mpg (US)

Charlie - '07 Citroen C4 Grand Picasso Exclusive
90 day: 37.58 mpg (US)

Corsa - '05 Opel Corsa C
90 day: 53.22 mpg (US)

Mruczek - '03 Audi A2
90 day: 60.61 mpg (US)
Thanks: 185
Thanked 167 Times in 117 Posts
Quote:
Originally Posted by meelis11 View Post
It already has VSS input logic similar like we want here - it counts VSS pulses.
So we need to modify injector logic to measure pulses instead of pulse length. MPGuino uses two input pins (and two interrupts) for injector measure - one for pulse start and other for pulse end detecting. Currenty arduino pin 2 and 3 are connected together - here we need to separate them in schematic again.
I thought the same, but counting pulses can be tricky.
For example - engine is idling and using 0.5 litre/hour when actual fuel flow trough the whole system is around 2.5 litre/min. Lets say that fuel flow meter counts 1000 pulses/ litre so one pulse is 1ccm

Now some maths:

First fuel flow will count 2500 pulses per minute = 150000 pulses/hour
Fuel used will count as -500 pulses / hour so second flow meter will count 149500 pulses per hour

If we want to update LCD screen once per second we'll get:

150000/3600 = 41.66666666666667 pulses/second from FM#1
149500/3600 = 41.52777777777778 pulses/second from FM#2
so the difference between those two numbers will be: 0.1388888888888889

So we need 7.2 seconds to see 1 pulse difference in flow!

As you can see that's not so easy to have instant reading from flow meters.

I'm thinking that best option will be to count time between pulses on both FMs
So for FM#1 we'll get 1 pulse per 24000us and 24082us for FM#2 accordingly

The difference is 82us (pretty small IMHO), anyway I think that would be the most accurate way of measuring fuel efficiency in a real time.

Time between pulses will allow us to get to those numbers above with pretty good accuracy and in a real time.

Then math is pretty simple

One liter for FM#1 will pass trough in 24 seconds and 24.082 seconds for FM#2
So fuel flow in liter/hour is accordingly:

3600/24 = 150 for FM#1
3600/24.082 = 149.5017 for FM#2

150-149.5017 = 0.4983 l/h

Combining that with VSS signal will give us l/100km or mpg

Error will be ~0.34% so it's not so bad for an instant FE gauge.

Of course tank average or trip average will be way more acurate, after some calibration.
__________________


Quote:
Gerhard Plattner: "The best attitude is to consider fuel saving a kind of sport. Everybody who has enough money for a strong car, can drive fast and hit the pedal. But saving fuel requires concentration, self-control and cleverness. It's a challenge with the nice effect of saving you money that you can use for other more important things."

Last edited by AndrzejM; 11-21-2012 at 11:01 AM.. Reason: Additional info
  Reply With Quote