I've been struggling with this minor issue for a while, tried a few things, but I haven't cured it yet. Figured I'd get some opinions.

When I am at low RPM, let's say 2000, and I smack the throttle open, the AEs are perfect and the car flies forward. If I smack the throttle at 5000rpm in 2nd, though, the boost shoots up REALLY quickly and the AEs can't keep up. If I increase the AE, then I get ideal fuel delivery under high-RPM throttle openings, but then my low-end gets these huge bogs when I open the throttle.

So what do I do? The only way I can see to avoid this is to go back to MAP-based enrichments which means a loss in throttle response. :( I know MSII can do a mixture of the two but I don't have that :(

 

Try using the RPM based acceleration enrichment. This lets you specify four RPM bands and put in different amounts of enrichment at different RPM.

 

Holy crap, I didn't even know that existed. Matt, does it extrapolate between the bands/values? Also, from what I see, activating the RPM-based AEs turns off the TPS-based AEs. Any way to keep both and have the RPMs just augment the TPS-based stuff?

 

From what I see in MegaTune, with MS1-extra. It triggers with TPS or MAP but the actual AE is only RPM based.

 

Spookyfish is right: It can be triggered off TPS or MAP, but the amount added is dependent on RPM instead of the rate of TPS change. Here's the coverage in the manual:

http://www.msextra.com/manuals/MS_Ex...nual.htm#rpmae

Although the manual claims a lot of installs don't need this option, a stock Miata is already pretty close to a "high strung small displacement motor."

 

have you tried setting the accel timer to "cycles" instead of "sec"? That changes the "accel time(s)" in the wizard to ignition pulses instead of time, thereby making the AE RPM dependent. I tuned out the same problem you mention by changing this one parameter and re-tuning the TPS-based numbers.

 

From the Manual:
"The Accel Timer can be changed from Seconds to Engine Cycle counts. Note that a value of 0.3 could mean 0.3 seconds or 3 ignition pulses. The Cycles option is useful because the enrichment will reduce at higher rpms"

Can you explain further?

 

what part?

 

I think he wants more enrich at higher rpm not less? That was what his last post said before editting

 

Sav: i think rpm based, time based is you best shot. Did you experiment yet?

 

You shouldnt be having any problems with your enrichments savington, if im not having any at any rpm, then you shouldnt either.

Shoot me a IM, and ill send you my AE table.

 

If you aren't running a Knocksense you'd never know you had this problem. The car runs and acts totally, fine, but on high-load high RPM acceleration (5000rpm+ in 4th, 5th, and 6th) I get a single detonation event.

I'm going to experiment a bit this weekend. I'm worried about a loss of throttle response if I move away from TPS-based accels, though. Pipefather, switching over to ignition-based would give me less AE at high RPM, not more like I need.

 

^ You didnt mention any of this in your OP. All I got was that you were losing throttle response, not that you have knock as well.

Are you sure it isnt mechanical noise? Especially for 1 knock event, it maybe be a fluke.

 

Please post your results as soon as you have something useful to report. I'd be very curious to see what you end up with.

 

I was hoping, but I checked the logs and the event always correlates with a lean spike, like a 12.2-12.3:1 AFR and 12-14psi. I run the car pretty close to ragged edge on timing, but my AFRs are 11.4-11.6:1 so when it goes that lean, it obviously detonates.

 

Played around with it a bit, no improvement. Still detonates when I apply partial throttle to 1-2psi, and then snap the throttle open. It happens at 4500 in 5th, 5000 in 5th, and 5500 in 4th.

My internet is on the fritz, so i will post more later. I included a zip file of the datalog I took; I put a marker in after every knock event. I can post a datalog of what it used to do with the TPS-based stuff (basically the same)

 

Back off the timing, its not worth the power vs engine life. Besides its not THAT much you would lose.

 

A small aside: I really do enjoy it when my computer decides to randomly corrupt various MSQs, so when I try to load my "last map saved" after a failed experiment my car runs like I've loaded an MM9093 basemap. One set of plugs and a session later...

 

If you look around timestamp 5940 in your datalog for example, after tip-in, your TPS has pretty much stopped increasing, but MAP is increasing (turbo is spooling).

This is why on a turbo motor one ideally needs BOTH TPS AND MAP based enrichment.
And why I still like my AEM over an MS...

 

Well, my acceleration settings were off forever, and it sucked. Car had LOTS of hesitation everywhere, especially when cold. Then I made one that was TPS based and it helped a little bit, but it still sucked. Anyways, I just spent an hour and a half making this one. Works soooo good compared to before. The car is a lot smoother when shifting. It makes it a lot easier to drive. And when I punch the gas it lurches forward. Never done that before without first falling on its face. MS2 FTW. I mainly tuned the MAP based cells, as it seemed more important. The TPS ones are probably way off. But it's running better than ever so I'm happy for now. If anyone else with MS2 wants to post theres, that would be acceptable...