Search RomRaider





  • View Page
  • Page History/Versions
  • Print Page

FMIC Rich Dip

Also known as “MAF Lag”

What is it?

Many people with turbo Subarus, draw-through (stock-style) MAF sensors, and front-mounted intercoolers have reported a “rich dip” in their AFR that starts right after stomping the throttle, and ends about a second later. A change of about 1.0 AFR is typical. I have personally observed this after setting tip-in to zero, just to verify that tip-in isn’t the problem.

This will happen at any RPM. If you consistently do pulls from 2500 RPM, you can tune around the problem without realizing it. But you’ll find that stomping the throttle at (for example) 4000 RPM will cause a rich dip. This cannot be tuned out via MAF scaling or even via the primary open loop fueling table. If you did tune it out for 4000 RPM throttle stomps - by leaning out the cells where the rich dip happens - you’d find that stomping the throttle at a lower RPM causes the mixture to go lean when you cross into the same cells where you were previously rich.

A picture is worth a thousand words, but I don’t have one handy at the moment.

What causes the FMIC rich dip?

Not MAF scaling - the same MAF voltage will yield different AFRs depending on how many milliseconds have elapsed since you stomped on the throttle.

Not the fuel table - the same RPM/Load cells will yield different AFRs depending on how many milliseconds have elapsed since you stomped on the throttle.

The leading theory is that it’s caused by the buildup of pressure in the charge pipes and intercooler core. While pressure is building, there’s slightly more air entering the pipes and core than there is exiting. The MAF sensor sees the the larger volume entering, and the corresponding amount of fuel is slightly excessive for the slightly-smaller amount of air that actually enters the cylinders. Once the intake tract reaches equilibrium, the air flow measured at the MAF sensor equals the air flow entering the cylinders, and the AFR returns to normal.

How can you fix the FMIC rich dip?

A few people have reported that switching to a blow-through MAF setup (where the MAF sensor sits between the intercooler core and the throttle body) solves the problem. I’ve got a blow-through charge pipe fabricated but I haven’t yet done the wiring to hook up the sensor on the far side of the engine bay. I’ll get this done when snowboarding season is over, if not sooner.

How can you minimize the effects of the rich dip?

While you can’t tune out the rich dip, you can tune around it. The key thing is how steeply you transition from 14.7 at cruise, to 11ish at max load.

When I first went to a FMIC, I was tapering my AFR just as steeply as the stock tune - I had basically just raised all cells that were richer than 11.2, to 11.2, and left the fuel table the same otherwise. The rich dip would cause me to go about 1.5 richer that the target, so that left with we AFRs below 10.0, and there would be a noticable drop in power right after stabbing the throttle. Power came on much more smoothly if I rolled on the throttle.

Later I set my fuel table to taper much more gradually. I still get AFRs about 1.5 richer than target, but now that only puts me into the high 10s when I stomp the throttle, and the engine is still pretty happy at that AFR.

The biggest shortcoming to this workaround is that it really only works when you go WOT from cruising RPM. When shifting, where you might go to WOT from 4000–4500 RPM, you’re probably immedately going to be in an area where the fuel table calls for an AFR in the low 11s, which results in a high-9s AFR due to the rich dip. You’ll feel a bit of hesitation when that happens.

The other issues is that you have to be careful not to maximize your ignition advance based on pulls where you stomp the throttle. If you do, you’ll be running too much advance if you cross through the same cells by rolling on the throttle, since the AFR won’t be artificially rich in that case.


RomRaider Forums

Join our forums, the best place to find help and answers!


RomRaider is developed and supported by volunteers working on their own time. To support their efforts please consider making a donation.

Page last modified on March 09, 2011, at 01:58 AM
Powered by PmWiki