Torque dip common to NA boxer engines?

[Sirsimon]

Originally Posted by FRSure

I see headers and tune get rid most of the dip. Why didn't Toyota do it?

Since it seems as if it can't be completely tuned out via ECU config, I'm perplexed as well. They obviously saw and "felt" it during their testing, yet let it slide. The only thing that comes to mind is that a lot of the testing was likely done with one-off custom exhaust components before they decided on the final design. I guess another possibility might be that they wanted the car to mellow out during that period in the rev range to either make the car more tame for the average RWD newbie or to increase aftermarket part sales. ;-)

[86-tundra]

because most of the headers that fix the dip are catless. it allows for more aggressive tuning.

[FR-S Matt]

Originally Posted by trevorovert

because most of the headers that fix the dip are catless. it allows for more aggressive tuning.

I can confirm this after having my car tuned over the past couple days. Catless header lets you adjust timing a lot more to take in more air through the intake and exhaust out more. Therefore, more power to be had and more advanced timing in that 3-4k range eliminating the dip. I don't feel it anymore. Nice smooth upward curve.

[s2d4]

Originally Posted by FR-S Matt

I can confirm this after having my car tuned over the past couple days. Catless header lets you adjust timing a lot more to take in more air through the intake and exhaust out more. Therefore, more power to be had and more advanced timing in that 3-4k range eliminating the dip. I don't feel it anymore. Nice smooth upward curve.

Can you explain how ft86speedfactory catted header achieve this without a tune?

[DarkSunrise]

Originally Posted by s2d4

Can you explain how ft86speedfactory catted header achieve this without a tune?

I've been wondering this myself. It's the only header I've seen that effectively eliminates the torque dip without a tune, as shown both on their dyno and a 3rd party dyno. Very interesting.

[s2d4]

Originally Posted by DarkSunrise

as shown both on their dyno and a 3rd party dyno. Very interesting.

I haven't seen the 3rd party one yet, can you point me to it if you don't mind.

[w00t692]

The nameless EL header makes the torque dip a torque peak. Tends to happen with a proper header at that point. I've noticed it's rare that any of the UEL headers make the torque peak there.

[FR-S Matt]

Originally Posted by s2d4

Can you explain how ft86speedfactory catted header achieve this without a tune?

It's one single dyno with no consistent proof? I'm sorry, but that's a crock of shit if I've ever heard one. A catless header has more tuning power than a catted header so that doesn't make much sense. I should be able to put on a catless header and do the same thing, but there are many headers that don't eliminate the dip by themselves. Rev Works even showed that with theirs and 0 tune. There's something going on there and to me it sounds like a marketing gimmick to get more sales.

Let the third party dynos do the talking about the product, then we'll see how magical it is.

[86-tundra]

I think the ft86speed factory cat is high flow, & I think it may have a different cell count from stock. this could have something to do with it. however I am amazed too.

[DarkSunrise]

Originally Posted by s2d4

I haven't seen the 3rd party one yet, can you point me to it if you don't mind.

Here's the one I was thinking of, but I just noticed he lists a Unichip in his signature, so this might have been tuned:

http://www.ft86club.com/forums/showp...postcount=1073

[s2d4]

Originally Posted by DarkSunrise

Here's the one I was thinking of, but I just noticed he lists a Unichip in his signature, so this might have been tuned:

http://www.ft86club.com/forums/showp...postcount=1073

Well, the scale just happened to not show much..

[DarkSunrise]

Originally Posted by s2d4

Well, the scale just happened to not show much..

Yeah not the best picture, but I think you can get enough information out of it. Typical stock torque dip is about 20 lbs-ft drop starting at 3200 rpm, recovering by 4500 rpm.

On this graph, the torque rises by ~5 lbs-ft at 3100 rpm, dips by ~5 lbs-ft at 3800 rpm, and recovers by 4100 rpm.

So not only does the torque rise where it would normally dip, but it's +/- 5 lbs-ft through the midrange.

[s2d4]

Originally Posted by DarkSunrise

On this graph, the torque rises by ~5 lbs-ft at 3100 rpm, dips by ~5 lbs-ft at 3800 rpm, and recovers by 4100 rpm.

So not only does the torque rise where it would normally dip, but it's +/- 5 lbs-ft through the midrange.

I commend you for pulling those numbers out of that graph, I couldn't do it as the error would have been way too large due to the lack of resolution.

It artificially flattens the graph by using double the scale required.
There are 3 dips visible and there was no baseline run to compare it to as reference.
On top of that, he'd been running unichip.
For these reasons, I don't believe that 3rd party dyno provides substantiated data to actually draw values for valid comparisons.

[zdr93523]

Interesting discussion. I'd be curious how the ft86speedfactory header eliminates the dip as well.