Omicron

Say WHAT??!!??

Are you saying that ALL dyno tests done to the car up to this point, by other vendors and such, were with the car in LIMP MODE?

Let's be clear here - this is critical. If you are correct, and can prove it, then all the HP numbers that have been bandied about up till now, such as 188 WHP being the "real" number the car puts out, are WRONG, and the actual is much higher. The lower numbers were generated because the car was on a dyno, and went into limp mode as a result. Is that what you're saying?

Also, are you saying that even if the car is not on the dyno, one good burnout will cause the car to go into limp mode and drastically cut power??? If that's the case, then every test that's been done up to now will be giving wrong performance numbers. Car & Driver, Road & Track, Autoweek, etc, etc - all WOULD have gotten better 0-60 times if the burnouts they did to maximize those times hadn't put the ECU into limp mode. In other words, without this limp mode thing, the car would have gotten better 0-60 times... right?

See what I mean? Heavy stuff. Thought this deserved it's own thread. Please clarify. Thanks.

MazdaManiac

iTrader: (3)

Most, if not all, of the dyno tests I've seen have used the "hack" method that Maurice points out in his post.
Generally, this means attaching the front ABS sensors to the rear with jumpers of some sort.
What he is wanting to do is add a switch of some sort that will allow the user to get the front sensors tied to the rear without having pigtails stretched all over the engine bay or down the sides of the car.
If you dyno without fooling the ABS, you get HP figures closer to 130 to 150. Kinda sad, actually.

Genom

Uhm... I dyno'd the car without knowing what I was doing and it threw out 189.7. I thought that was higher than average, and I didnt do anything before tossing it on the dyno.

I can tell ya that the CEL light stayed lit for about 20 minutes after I got off the dyno, but they also topped the car out to see what it would do, and I am pretty sure thats the time it threw the CEL. However it reset itself after a little while and I havenot had it again.

Omicron

Yes, I recall reading about this in the loooong thread that touched this all off (https://www.rx8club.com/showthread.p...threadid=11614), and understand the concept of tying the sensors together. However, I disagree that most dynos have been done with this hack... in fact, the only one I've heard of doing this is Canzoomer. I wonder how we can find out how many people who've dyno'd the '8 have done this. Hafta post something in the main forums asking...

I also wonder just how much better the RX-8's 0-60 and 1/4 mile times will be with the limp mode "defeated."

Very very interesting. As I've said, I saw this before and did not realize the potential impact of the finding till today.

RX-8 friend

It may be that only some cars are sensitive to this. In Canzoomer's it takes 3-4 seconds to trip, so one dyno run could be completed before triggering it. It seems to trip when the engine is up around 6000 RPM, but that may just be because it takes 3-4 sec. to get it up there. It's winter, and icy. Try spinning the rear wheels on your car for 3-4 sec. to see if it will trigger a light (turn off the DSC first). It's easy to reset it. Shut it off, restart, and drive it around the block.

Read others comments carefully. Did they mention only one run gave the correct power output? Did they mention they had trouble getting a good reading? I seem to recall some mention of that on some tests. As far as 1/4 mile times, if you don't spin the tires for more than 3 sec. you will not have a problem.

Our solution involves isolating the front/rear connection so they can't damage the ABS system. Wiring them together (the hack I mentioned) hasn't harmed anything, but you never know without a schematic. Trust me, you DON'T want to damage the ABS. It's an expensive part!

This may end up being another "racing part" for sale. We'll see.

BigAsian

I don't think that it really affects the dyno on most cars. I belive it was the december issue of SCC, they dynoed with the sensors connected and without and it actually made how hp with them connected....Also, what in the world was mazda thinking when they programmed the ECU to go into limphome mode in this type of situation anyway?!

Hymee

iTrader: (1)

Guys - I have been thinking (maybe that is a bad idea ) They (Mazda) say the only way to measure engine power is on an engine dyno. That part is correct. But on an engine dyno you still need your ECU hooked up, as it is what is controlling what is going on... So what about the wheels turning etc... when they are not even there? Then it hit me like a light bulb. Are you thinking what I'm thinking???

Instead of fooling the computer to think the front wheels are going the same speed at the rear wheels - why not disconnect the rear wheel sensors, and make the computer think the rear wheels are going the same speed as the front - 0 RPM???

Worthy of injecting into the discussion I think!

Cheers,
Hymee.

zerohour

Hmm kinda weird........When I dynoed at SP engineering without disabling the ABS I Dynoed 181 SAE corrected to 179. So Ill have to be skeptical about the validity of your statement.

MazdaManiac

iTrader: (3)

First of all, that statement is two months old.
There have been many, many more dyno attempts made by different people under different circumstances since then; some of them have had some measure of success.

Second, you can be skeptical if you wish, but your best bet would be to use the search feature at the top of the page and see for yourself. You will find the same information that led me to this conclusion.

Tying back the rear ABS sensors is the current favorite mode for an accurate, repeatable test.
How many runs did you make?