I have been long-time plagued with issues with the AEM EMS.
I currently am running the Series 1 box with an AEM Steel 3.5Bar map sensor.
There have been a couple professional tuners that have told me that there is no way to get around the throttle response issue that AEM is plagued with, and that the AEM EMS ecu's have problems that cannot be corrected unless a complete overhaul is done. (Series 2 included).
I decided to put some of my feelings into this so some people can understand my frustrations with this AEM EMS, this is what I feel, Then I will get into the concerns that I have with my current AEM EMS
Here's a condition that I'd like assistance with that can happen in the wet or dry... and to anyone for that matter during an emergency situation... and the way this AEM EMS is responding, it can get you killed or into an accident. It already has almost gotten me rear ended twice.
Say you take a turn, you fish tail, you are going at a high rate of speed, there is NO time, ZERO time for throttle delay, at this moment you need to be able to put the RPM's where you want them as fast as possible to rev match in the gear you need to deliver the power you need to correct your spin-out by a possible power slide to carve into the turn and away from possible danger. Anyone that has driven aggressively understands what I'm saying here.
I've never been in a car that has had such a delayed response of the throttle. Turbo/SuperCharged/N-A... none.
If I were rally racing... and I approached an e-brake hair pin, I'd clutch down, pull ebrake to initiate slide after appropriate steering input, then nail throttle to raise rpm to rev-match to the next gear and then lift clutch up in a fashion that lays the power to the floor in the way that I demand depending on the turn/camber/angle/terrain... but with my AEM EMS, this is not possible because I'd have to nail the clutch AND the gas at the same time in that extreme time-sensitive situation in order for the gas to interpret my gas pedal movement to say "hey, guess what, he just hit the gas... should we put some fuel in now? or later? what? was that now? oh, ok..." well I want it when I hit the gas pedal, and not later, as I think everyone would.
I'm can be an aggressive driver and when I hit that gas pedal, I want the computer to do something like it used to when I had the OEM ECU, not to chuckle, choke, or make my motor go lean on fuel, or attempt to lower rpm/stall. I know its a turbo car and I know turbo has lag, and I can differentiate the two, but throttle input should not have lag. I heal-toe often, and my inputs to the car are done in a way where they need to be recognized when I do them, not predicted, or delayed.
No stock-ecu Evo I have driven (and I have driven many) has ever had this type of problem. If you feel that the AEM EMS can perform like the OEM ECU in this manner, in these circumstances, then I'd like to be driven in one with such a setup. Or if you are so inclined, I'd like to drive it myself. But I will also say, I'd feel horrible if the truth is that it won't, because then I'd be wasting every one's time. And that, I do not want to do because I know how valuable time is.
Some of the main concerns that I have with "my" current AEM EMS are, but not limited to:
• Throttle response (instant lean AFR reading on throttle input, then slow calculation to regain fuel... which kills response and causes car to "buck" before acceleration or even holding an RPM for cruising)
• Weak idle. (No matter how high I set my idle map to compensate, the vehicle will stall on its own when depressing the clutch, or after a long drive and then finally coming to a slow down. It's as if the idle motor is confused as to when to kick in, and when not to.
• TPS reading not stable. When this reading doesn't read between my 0.3 to 0.9 preferred value, all hell breaks loose. Not only does my idle run richer, but it stalls more frequently, and the response turns to crap.
I have used Browningderek's guide to eliminate AEM's ability to have to "kick on" the idle motor. This way, there is always a static method AEM will attempt to idle the vehicle. It works great for the most part (unless my TPS goes haywire again)
The link can be found here
I have gotten the AEM EMS to run like a stock computer on my car (yet again... as long as the TPS doesn't go haywire) by using the Idle guide above... and the following methods:
I have had my tuner un-do boost compensation. This way, my ECU does not have to over-calculate anything. What's in the fuel map, is in the fuel map, and that's that. The fuel map looks like a pretty dangerous slope to ski down, but it works, and fine tuning each boost level can be done easily. I also completely flattened out the Throttle Injector Correction Table. The only section I'll even consider lowering in the event my idle be too rich... is the 0% throttle value... every thing else remains at 0% fuel added/subtracted. This way, again, less for the ECU to calculate. So therefore, the ECU only deals with the fuel map.
There was no futzing around with the Throttle Fuel map (if you don't know what this is, you're better off) as it is glitched in the AEM EMS... it tends to strap itself to the fuel map for whatever reason.
This is my gripe: When driving the car in a "daily fashion", (4,000RPM shifts between gears from red light to redlight driving... trying to keep up with rush hour traffic) When I clutch down, then change gear, then tip throttle in... and begin to clutch up... In that last step of "tipping throttle in while engaging clutch" is where I have lost many hairs and initiated project gray hair and skin wrinkle.
What is happening is, as soon as I tip in throttle, my AFR is no longer reading anything between 10.0 to 16.0 (just to give you an idea) in fact, it's showing a split second of 21.0 and above. When there is no fuel, there is no response... so what ends up happening is as I engage the clutch... the car is decelerating. Then the clutch engages, and I get a harsh "buck" through my drivetrain as my motor tries to spin the drive train the wrong way because my throttle input/rev-matching/momentum is conflicting.
Anyone with AEM EMS knows what I'm talking about here. The price we are paying for this horse$hit is honestly not worth all the stress. I miss my stock ECU, and couldn't care less for an additional 25-50 ponies.
This is what I did to correct the problem. Maybe correct is a strong word... band-aid? Pacify? The problem is still there.
First off... I checked my AEM Map sensor connection since I found my car would stall if I shook the harness... I found that the 3 pin plug harness was completely corroded inside. I dremeled out the plastic to expose the 3 male pins of the map sensor... shaved them down to clean off all of the green grit. Fluxed them, Soldered solder onto them, and directly soldered wires into the factory evo harness. This corrected a lot of issues to begin with.
Then: I have taken every negative boost load cell above 1500rpm, to about 5500-6000rpm and increased the fuel cells to make my AFR read 9.5-10.0 in those sections. This alone, will solve the problem, however, you will be emitting black clouds of fuel behind you. So what I did to counter that was enabling o2 feedback for the range of 1500-5500/6000rpm, and turn it off above the load I have the fuel set to correctly.
What happens is that the initial fuel map throws that extra fuel that you need for response, and then o2 feedback slowly pulls the fuel away as you drive in order to reach whatever target AFR you have in your o2 feedback map.
Issue with this: o2 feedback on AEM EMS Series 1 is so slow (apparently Series 2 has a faster processor... but I'm not sure it corrected this problem either)... they might as well not even included it as part of the package in my box in my opinion. The correction rate is hideous. If you tamper with the o2 feedback ms rate... what will happen is the car will end up running richer, then leaner than the target. It overdoes itself.
I haven't spent enough time tweaking the feedback rates on o2 feedback, but from what I've seen so far... it won't work as I want it to.
I fully understand that you are supposed to tune your fuel map with o2 feedback completely off first, and then enable it to maybe give you that extra bit of fine tuning or that extra bit of gas mileage for long distance trips... but if tuned that way... I can guarantee you that your throttle response will be horrible.
If anyone has insight as to how to get around this well known issue with any AEM EMS box for the Evo's, (8's and 9's that I'm aware of so far) Feel free to put your two cents in this thread.
If a thread like this exists already, I apologize, and moderators, please feel free to either copy/paste/move or direct me to the right place.
I have heard from other tuners that the TPS's on our cars go bad, and you must get a new OEM one in hopes that it holds the cailbration.
Apparently it's best for it to read 98% throttle when you are floored on the pedal, and 0.5% (or as close as you can get to that) when your throttle isn't touched. Be sure your throttle cable is setup correctly with ZERO pull. A little slack is fine, but don't set it too tight.
I thank everyone in advance for any information and/or any light you can shed on this painful issue that many drivers that are OCD such as myself with tunes are experiencing.
I will be constantly updating this thread with any other insight I may find as time passes since I'm sure many Evo users are using AEM EMS in their vehicles.