I've just done a series of tests using scanalyser to see if I could detect any performance differences using different octane fuels . I know people here have done dyno tuning that shows better results on low octane & yet others showing good results on high octane - I wanted to test on the road & see if the difference was detectable.
I have a piggyback EMS so have played with timing & fuel with each octane to try & get the best setup for that particular octane.
I decided I would test the rpm range in which I really wanted to get gains . I know from reading this forum that there is very little that can be done to get any gains below 5500rpm anyway . So my test was to log acceleration from 50km/hr through to 161km/hr (100mph) . This involves 1 gearshift (2nd to 3rd) .
In reviewing the results I remove the gearshift as a variable & only look at acceleration in gear.

I found In 2nd gear the diferences in acceleration are so small that my measuring tool (scanalyser) is not good enough to see a difference . I can only really see a definite difference if it is more than 0.15s . However in 3rd gear I can definately see any improvement - if it is there.
Tested the following fuels :

91 Octane (87 in US approx.)
95 Octane (91 in US approx.)
98 Octane (93 in US approx.)

Findings :

No detectable improvement from any of the fuels .

I got some interesting differences by pulling timing (5deg) - finding 91 worked better with the factory timing whereas 95 & 98 worked better with the timing pulled .

Altering the AFRs has less effect than I had been led to believe. It certainly helps but anything in the 12.5 - 13.0 range gives me identical results .

Following these tests I decided to switch back to 91 as it is cheaper & seemed just as good. However , In certain conditions (hot engine - higher intake temp) I discovered that when I tried hard acceleration I got such bad detonation that the knock sensor absolutely killed the engine . This was with stock timing curve & AFRs around 12.5 - 13

Will stick with 95 now - what Mazda recommends funnily enough

Oh - one last thing . The difference between my best tune & stock ECU is 0.3 s
from 110km/hr - 160km/hr . This would translate to a little more than that on 1/4 mile run.

Edit : See charts Below for full results .

 

I have ran 93 in my car since I had it with the exception of after Katrina during our "gas shortage". 89 was all that was available. I kept close track of my milage and I lost almost 2 miles per gallon. As soon as it was available again, I went back to the 93. I did the math and I was not saving any money with the 89. I also figured that if I was losing milage, it can't be as good for my engine.

 

Good work, brettus...but take care not to create too many variables when trying to illustrate the impact of one factor.

 

Looks like a very well designed experiment.

Can you follow up with the actual times you recorded?

Nicely done, BGuy.

 

Actually never intended to compare octanes but in search of getting a few extra tenths I gave it a shot and realised that it may be of interest.
I'm not saying that I have definately resolved this issue , just that "on the road " the difference is so small as to be hardly noticeable .

 

brettus,
nice job.. where did you get that idea???

how long did it take you figure out how to set up the scanasyer??

btw, what color is your car.. i think we might be twins.

beers

 

Thanks for that .

Absolute best time so far 6.8s 110km/hr to 160 km/hr in 3rd gear
Achieved this (or very close to) with all three fuels.

Worst time 7.2s

All tests done with between 1/4 and 1/2 tank of gas with intake temps between 16 & 20 deg c (at 100km/hr) & coolant temp between 88-92 degC

 

Scanalyser was pretty easy - just took me a while to come up with a repeatable test that scanalyser could measure with any accuracy. It really needs double the resolution to be really good for this sort of testing.

Done over 30 runs now - what fun

Mine is a VRed - & you ?

 

gray, i am relieved now.

keep up the good work. i think mine goes to the dyno monday..

beers

 

A word on the dyno - I'm finding that the tune we did on the dyno does not give me the best result on the road - perhaps the cold air duct comes into play at speed .........

hope to do some tests with premix next .

 

dyno is for testing power before and after the new motor...

i am most interested in your premix tests..

beers

 

These charts may cause a bit of a stir .

One shows tests I did on the three octane fuels available in NZ . Note : the test on 91 Oct was actually with the fuel map in place but no timimng changes. Did not do a "NO tune" test on 91 at the time unfortunately . An error on my part.

The other shows affect of a tune on each fuel . IE AFRs of 12.5 - 13.0 & timing pulled above 6500 rpm . Note this was the timing we found made most HP on the dyno with 95 octane.

proceedure was :
Same stretch of road - same direction.
1/4 to 1/2 tank of gas
Ensure ambient temp within range 15deg C - 19 deg C
Ensure coolant temp not over 93 deg C
Ensure intake air temp stabilised (ie car not left sitting so high initial intake temps)
Ensure not much wind

Log made of run from 4000 rpm in third gear at 100% throttle till rev limiter kicks in.
Each verticle line on chart represents 0.14 seconds

Findings :
1 /very little between octanes with standard tune as per previous post.

2/with a tune I found better results with higher octane . I Know , I know - this varies from what you guys are finding in the states but this is what I found .

3/ If you let a car sit after a high speed run then try and do another run - you will get a poor result . Intake air temp goes very high initially & there is possibly a lot of heat within the engine that the coolant may not have time to get rid of.

4/ 91 octane worked better with the std timing settings - was actually slower than std with timing pulled - while the other two octane fuels were faster.

 

Interesting reading Brettus! I have bought a Scananalyser to and shall be performing some tests just like you do. I will use Shell V-Power 99 unleadad and also a mix of 30% E-85. Shall be quite interesting. Next week i will bolt on my prototype header which is coated inside and outside and then evualate it on my RotoTest.

/Lasse

 

the octane doesn't matter to my car nowhere near as much as the brand/quality of the gas. a quality 87 is way better than noname 93. in mileage and even in feel. it's a diffrent car when i have good gas vs poor gas.

 

If your car is tuned for A octane, putting B octane is not gonna yield more power when B>A. If A>B then the engine will get marbles.