Well, after a couple of days of hearing a strange sound at WOT over about 6500-7k rpm... I finally figured out what it was.

I was broke last week and filled up with 87 octane gas (first time ever) and my car basically took a **** on itself.

This morning I filled up with 93 octane and it runs like a champ...

So I guess my car is one of those nothing below 93 cars.

Anyway, I am happy now. Beers for everyone.

 

I've never so brave as to put in anything less than 93. :D
________
Avandia Death

 

dang dude, sucks you had to learn the hard way, even in my brokest of times I was still using 93, maybe not a full tank, but still 93 lol I've heard some people say the mid-grade is okay, but i'm not so sure how many of those same people drive it everyday etc.

 

My car runs great on 87 and I've been using it since I bought it. :D

 

Perhaps this didn't occur to you, but it could've just been a bad tank of gas... And the octane had nothing to do with it.

 

The RX-8 seems strange when it comes to gas. Some people can run good on 87. Others can only run on 92. Yet others do fine on 89 but not 87. It's weird. It all has to do with tuning. The engine doesn't care. Rotaries have always loved low octane but this damned ecu keeps things expensive.

 

Yeah it all has to do with the tuning. But more importantly in Sovine's case, he switched abruptly from 93 to 87 and the car didn't have time to relearn how to handle it.

In the winter months, I use 89 or sometimes fill up with a mix 93 with 87 (5 gallons 93 and 9-10 gallons 87). In the summer I stick with 93.

-Mr. Wigggles

 

any particular reason? what's with the mix?

 

i drove mine every day, and it ran virtually the same low grade or high grade, WOT runs and all. In the hot FL heat even. 36,000 miles.

 

Out here the best i can get is 92.. and it s*cks. If i get shell gas i'm okay... but any of the other brands is horrible on the car.

-hS

 

Mine runs just fine on regular (87) and it doesn't seem to matter whether I get it from Chevron or Texaco, or Shell. It might run better on premium, but I don't see a need unless I was planning to do some really heavy-foot driving.

 

87 + Turbo = Death

 

Yep, that extra compression would make a difference.

 

me want turboz

 

anyone care to explain why not to run 93? I mean, is it really that big enough a difference in price to not use it? Is it a preference thing?Why doea Mazda recommend 93 if it's not necessary? Do I need to go search happy or? I haven't follwed any thread where they wre talking about the difference in octane etc. (other than I know the higher the content, the more hp can be had right?) so I really don't have a clue, most vehicles I've owned previous to this one had no specific requirement stated or recommended.

 

well okay, so much for 1 theory lol

 

Mazda says 91 octane is required, not 93. Whether they recommend 93, when and where available, may be another thing entirely.

 

The ideal thing is to have gasoline with octane just high enough to prevent detonation. Octane is not the same thing as energy content, it's just detonation resistance.

RG said that some RX-8's can run low octane while others can't. That seems to be true from reading the boards, and I can confirm that myself too. I have a 6-speed that does fine even in summer here in Houston; my brother has an auto 8 but he tried 87 and had problems.

 

cool, thanks for the info!

 

I ran 87 for the first 20K, no problems. Then I would get detonation and went to 91, no problem. Tried a bottle of "Chevron-Techron" in two tanks in a row. Now it is not detonating with 87! What Im wondering is it because the carbon buildup is gone now or its not as hot anymore. If I get some days in the 80s it will tell me.

 

93,000 miles 87 octane, no problem except did get one tank of bad gas, but nothing since then.

 

The ECU has a learn mode that tends to offset timing curves a bit. I know that on Bosch and Delphi (GM) systems there are a handful of factors such as air pressure at startup (& temp if it's somewhere below 100° - so it knows if the motor is "cold"). That sets a type of altitude correction factor. I would suspect the Mazda does the same thing.

Knock sensors are monitored continuously. When detonation counts exceed a certain number the ignition curve is altered under those RPM/Throttle conditions. To an extent fuel mixture can be altered to help control detonation as well. Delphi systems added a second O2 sensor at one time to help manage mixtures so I suspect Mazda is doing the same thing here.

The alteration range is limited so it's not always a cure all for widely varying conditions. The alterations to ignition/fuel is generally stored in deviation tables against a standard curve/fuel tables. It took some pretty fast custom processors to get the fine tuning systems that are found in just about every new car today.

Various conditions will affect whether one's car will run on 87 or 93. Starting & drive cycles, carbon buildup, sensor drift, Factory program release level, even the fuel used.

To get the most out of the motor Mazda is aggressive on the ignition curve. Three things contribute to detonation, in this order; Compression, timing advance, and lean mixture. Considering that it's possible to add a turbo and not destroy the motor after 3000 miles of pump fuel, I would have to say compression is not the problem with detonation on an NA motor. Mazda runs pretty fat in the upper RPM band which is probably a safer way to manage an aggressive curve considering the way learn modes work. Especially for folks that live in the mountains and change altitudes by 3000'+ feet in a single drive.

I've tried 87 octane but it takes a couple days of my 90 mile round trip to work before it seems to behave. Living in North Texas the altitude change is about 200 feet on my daily commute. I just run 93 octane anymore.

As compression is the major factor in detonation, the time to run 93 is in the cold air not the warm air. Higher air density. Those familar with calculating Relative Air Density know why.

To understand the dynamics of detonation takes a bit deeper understanding of what's driving it. The spark fires and a kernal is born - which propogates outwards at the speed of sound - in a cylinder running 10:1 thats about 1800 feet per second. Convective transfer. Detonation occurs when seperate kernals begin to "spontaneously" ignite beyond the front of the initial kernal. Guess what else is emitted from the kernal - Infra-red. At the speed of light. Moreover the slicon carbide crystals in the nikasil plating absorb this IR and then re-emit it at about 1400 nM wavelength. Now guess what wavelength most of the content of gasoline soup love to absorb most?

BTW, if you want to learn more about the dynamics of detonation Porsche published some really excellent white papers back in the early days of the 930 turbo when working with Dow chemical to develop Nikasil plating.

 

I use 91....only thing I can get.