Economy- My Most Boring Thread Ever!

[Sandy 191]

Probably like most of you, I'm finding the fuel to be my biggest expense after the Mortgage. I'm not against fuel duty, I mean taxing fuel as a way of taxing vehicles is pretty fair because it's directly related to the type of vehicle and kind of use.

So I want to try and cut my fuel use as much as possible without getting a boring car. I'm putting together an economy biased 205 soon, but more on that later. The first and most obvious choice to me is to make sure I'm driving the car I have as efficiently as possible.

 

To get a handle on where the fuel goes. I've always used the tankful to tankful method of recording my mpg, but I need to be more specific than that to find out exactly when I'm wasting fuel. I've set about monitoring how much fuel is being used as I drive, this isn't all that hard with an injection engine, because you can quite easily monitor the "duty cycle" of the injectors.

 

Duty cycle: The percentage of the time the injectors are open. The amount of time the injectors are open dictates how much fuel goes into the engine, so if the engine at 3000rpm rotates once every 20 milliseconds and the injectors are opening for 5 milliseconds every rotation, the duty cycle will be (20/5)x100 = 25%. The great thing about measuring duty cycle is that it takes engine speed and injector duration into account together and just tells you literally how much of the time they are open. We can then relate that figure to injector flow, so if you have 4 off 300cc/min injectors running at 25%, you'll be using (4x300)x0.25 = 300cc per minute. You can then compare that figure to your speed to get an instant economy figure (albeit slightly approximate).

 

How to measure duty cycle: Injectors on most petrol engines have battery voltage on the positive side and the negative side is connected to the ECU. The ECU earths the injector for however long it wants to hold it open, which controls the fuel flow. If you connect a volt meter between the ECU side of the injector and earth (don't do this on my say so, I'm just explaining what I'm doing and don't want anyone to cock their ECU up!!), then with the injector closed (0% duty) it'll read battery voltage, if the injector is fully open (100% duty, which should never happen in reality) it'll read 0v. Now whatever the voltage is relative to the battery supply voltage, will indicate the duty cycle, so 13v at the battery, reading 6.5v on the ECU side of the injector, will be (13/6.5)x100 = 50% duty.

To make it easier to read, I connected one lead of my digital meter to the ECU side of the injector and the other lead to the battery voltage, so the result will read as the difference between the two. I tested the supply voltage with the engine running first, which was 14.15 volts.

 

First phase was to do a simple test of seeing what the duty cycle was on a flat (private) road at a steady speed with negligible headwind (perfect day today). I got the car to a steady speed/throttle in 5th, glanced at the meter a couple times to take an average reading and made a mental note, then stopped and wrote it down. The results are as follows:

 

30mph 0.32v 2.3%

35mph 0.41v 2.9%

40mph 0.51v 3.6%

45mph 0.42v 3.0%

50mph 0.71v 5.0%

55mph 0.65v 4.6%

60mph 0.90v 6.4%

65mph 0.93v 6.6%

70mph 1.08v 7.6%

75mph 1.38v 9.8%

80mph 1.55v 11.0%

 

Now if I had to create a formula to turn this into approximate mpg. There's probably an easier way, but a bit of mental exercise won't hurt me. I always got marked down in maths for skipping the intermediate equations, but since hardly anyone's still reading, I'm going to skip them again; it can probably be simplified, but this is what I came up with:

(V/60)/(((Fx4)x(D/100))/4546)) = E

V is speed in MPH

F is injector flow in cc/min

D is duty cycle in percent

E is economy in MPG

 

Here's the result as a graph:

 

It's an interesting pattern and I'm confident the pattern's accurate, but because the injector open times at small openings to tally that accurately with the flow rate, the mpg is probably a touch overrated and it is an instant reading, as soon as you move the throttle down the voltage shoots upwards!

But now I know my car uses less fuel at 45, 55, 65 than 40, 50, 60!

[steve@cornwall 100]

Seeing as you have a handle on wiring - could you utilise a fuel computer from another model? Poss a late 80s ford? The one in my old Orion 1600i seemed pretty accurate.

 

If you want something cheap to run I have a little petrol sipping seat here...........It's not boring, either - you have to be awake to be bored

[mattmk1 0]

I was always told that for any car 56 Mph was the most economical speed, which ties in partly with your graph.

[Ryan 99]

I was always told that for any car 56 Mph was the most economical speed, which ties in partly with your graph.

 

Depends entirely on the engine, gearbox, wheel size, etc. Speed has little to do with it.

[TurboSam 5]

although wind resistance has quite a bit to do with speed..

[KRISKARRERA 2 1 Cars]

Didn't the 56mph thing come about because they discovered that was about the point where the vehicle's weight because less of an issue and the effects of cutting through the wind came into play? (just to describe it in a really crappy way).

[Sandy 191]

Thanks for the offer Steve, I'm already building a 1.1 205 that will appear shortly when I've cleared some work. I expect the fuel computer probably does the same, or maybe it uses manifold pressure like the LED display my Maestro HLE had all those years ago!

 

56mph is a nice round 100km/h and the point at which many think engine efficiency and drag intersect, so to speak, I wasn't expecting it to be more economical at lower speeds, but it clearly is. The 45mph peak on mine is possibly down to the shortish 5th (0.87:1 5th, 4.266 FDR); with a taller fifth it might be more like 50mph.

[vern 0]

Not totaly relevant but, i have a SEAT altea diesel with a fuel computer and it also shows the best fuel economy at around the 40-45 mph mark (about 70 - 80 mpg!).

I think the 56mph is an averge speed the car industry work to calculate the general efficiency of a car, so many cars are setup to give best economy at this speed, so improving the advertised mpg figures. I think this has been largly supersided by the urban and extra urban methods of calculating efficiency.

[Sandy 191]

Cool, sounds like I'm barking up the right tree then

[Rob_the_Sparky 9]

The DVM may not average the fuel pulses that well so there may be some errors in your measurements, you really want to put it through an op-amp and then a simple filter before the DVM but TBH without trying it you won't know how much difference it would make. You could try it with a simple RC filter though, only a couple of components and should be good enough. Just tune it to filter out the pulse frequency and measure the DC voltage.

 

Juding from experience of driving my wife's Golf (with instant mpg) then you get good mpg when driving like a granny and as soon as a hint of right foot is used you are knackered. However, it is harder to work out whether v poor economy for a second or 2 is worse than better economy for a longer period (e.g. better to accelerate hard for short period or gently for a long period). You have to do the integration in your head and that isn't that easy!

 

Rob

[Batfink 201]

my beemer is most fuel efficient I think around 60mph but thats when the car is producing its most torque at 2000rpm so I think this helps pull it along effortlessly.

[M_R_205 5 1 Cars]

30mph 0.32v 2.3%

35mph 0.41v 2.9%

40mph 0.51v 3.6%

45mph 0.42v 3.0%

50mph 0.71v 5.0%

55mph 0.65v 4.6%

60mph 0.90v 6.4%

65mph 0.93v 6.6%

70mph 1.08v 7.6%

75mph 1.38v 9.8%

80mph 1.55v 11.0%

 

What sort of RPM`s are you doing at those speeds? im running a gearbox out of a 306 deisel (i think..) and this is going to sound obsurd but i try to average around 2500rpm in 5th when the juice is low based on the thought trail that the engine is quietest then and therefore possibly at its most economic as less energy is being wasted... at 2500rpm in 5th im doing just shy of 60 (not sure exactly as my needle is rather eratic lol)

[Sandy 191]

The DVM may not average the fuel pulses that well so there may be some errors in your measurements, you really want to put it through an op-amp and then a simple filter before the DVM but TBH without trying it you won't know how much difference it would make. You could try it with a simple RC filter though, only a couple of components and should be good enough. Just tune it to filter out the pulse frequency and measure the DC voltage.

 

Juding from experience of driving my wife's Golf (with instant mpg) then you get good mpg when driving like a granny and as soon as a hint of right foot is used you are knackered. However, it is harder to work out whether v poor economy for a second or 2 is worse than better economy for a longer period (e.g. better to accelerate hard for short period or gently for a long period). You have to do the integration in your head and that isn't that easy!

 

Rob

I did think there might be some error there Rob, but it's updating quickly and very repeatable, so hopefully not far off. On the DTA S40 software I can read the duty cycle off the display (which i'm sure would be accurate), so I might try that and see how it compares to my figures. I was thinking of making a small staged LED circuit that would light one LED up for 1% duty steps upto 10%, but really it needs to relate to speed to be relevant, which is getting beyond my electronic engineering ability at the moment.

On the acceleration angle you mention there, I think taking your time probably is more efficient, Because when you accelerate moderately hard, the duty about triples, but you don't get up to speed in a third of the time; subjectively,but perhaps I could simulate that well enough to time it.

 

It's 18.1mph/1000rpm in top, so:

40mph 2210rpm

45mph 2486rpm

50mph 2762rpm

55mph 3039rpm

60mph 3315rpm

[M_R_205 5 1 Cars]

It's 18.1mph/1000rpm in top, so:

40mph 2210rpm

45mph 2486rpm

50mph 2762rpm

55mph 3039rpm

60mph 3315rpm

 

thats quite interesting then, you most economic speed, 45, is reving about 2500rmp, im guessing your running a 1.6 box?? if my maths isnt to shaby my engine at 1000rpm will be around 24mph? seems alot, but there again im only going off a very bouncy needle!!

[jackherer 543]

im guessing your running a 1.6 box??

 

I assumed Sandy was talking about his 309 with a Honda Accord Type R engine/box.

[M_R_205 5 1 Cars]

I assumed Sandy was talking about his 309 with a Honda Accord Type R engine/box.

oh crap, do i feel like an ape!!

[Sandy 191]

Er, yes, Honda. 4.266:1 FDR and 0.87:1 5th, which is shorter than 1.6 GTi a little.

[Richie-Van-GTi 71 2 Cars]

Im averaging about 150 miles to a tank this last 2 weeks, but then I only pay £20 for 60 litres

[jeremy 70]

Slightly off topic I found that when my wishbone was beginning to fail I drove the most conservatively and managed 320 miles to the tank on my 1.6GTI, since replacing it it will be back down to the 260 miles per tank. My point being even if you find your answer will you change your driving to get the best mpg, because if not then you are going to a lot of effort to achieve very little.

[Sandy 191]

I enjoy driving, generally, trying to get the best economy is more of a challenge than going fast. Mos was saying to me yesterday that people might read this and think "why the **** fit a VTEC engine blah blah and then drive it like a granny?" etc, but that would be missing the point of my car. I like to do both.

 

I'm putting together a 1.1 205 which I'll be using some of the lessons from this car on, to optimise it. Fun at all speeds hopefully.