Outlawstar98
Member
- :
- 2003 MazdaSpeed Protege / 2013 BRZ
Mmhmm. I still I a ton of work to do on that sticky. I'm getting there though. Well not a ton but my like is boring so I have the time haha.
AEM F/IC writeup
- Thread starter Ricktalife
- Start date
You have to recalibrate your throttle for every calibration file. I had this problem at first too. You also have to recal the RPMs for every calibration file, which makes it very difficult to compare two calibration files unless you get it dead on.
You don't want the car running for the TPS calibration. Just follow the instructions. Leave the ignition on. Hit the auto button, leave your foot completely off. Hit enter, put it to the floor, then hit enter, and you're done.
You can then immediately test it with the Gauges window and see if it goes to 100%.
Ricktalife, did you buck before you had the AEM in at all, meaning completely stock PCM?
Also, what wideband are you using, and do you have it hooked up to the aux in? AFR has nothing to do with your MAF readings. Yes your car will add fuel based on the MAF reading, but that doesn't mean it's going to be rich or lean. You can have a high MAF value but the car does not add enough fuel and you run lean, or you can have a low MAF value and the car can add too much fuel and run rich.
The reason you see high AFR's at idle or cruising is because the PCM is using the narrowband O2 sensor to keep the AFR around 14.7. Narrowbands have opposite behavior to widebands. Our narrowbands are lean at .0 v and rich at 1 volts. A wideband however is rich at 0v and lean at 5v.
Hopefully that helps you understand the aux in table. The aux in table is setup for a wideband, not for a MAF signal. I was just being "sneaky" and using the aux in to read my stock MAF in values, you aren't supposed to do that.
Also, do you have the acceleration settings on? Make sure your fuel and limit settings are at 0% in your setup window. This is to add fuel quickly when you stab the throttle since the PCM takes a second to catch up. If you dump too much fuel you will misfire.
You don't want the car running for the TPS calibration. Just follow the instructions. Leave the ignition on. Hit the auto button, leave your foot completely off. Hit enter, put it to the floor, then hit enter, and you're done.
You can then immediately test it with the Gauges window and see if it goes to 100%.
Ricktalife, did you buck before you had the AEM in at all, meaning completely stock PCM?
Also, what wideband are you using, and do you have it hooked up to the aux in? AFR has nothing to do with your MAF readings. Yes your car will add fuel based on the MAF reading, but that doesn't mean it's going to be rich or lean. You can have a high MAF value but the car does not add enough fuel and you run lean, or you can have a low MAF value and the car can add too much fuel and run rich.
The reason you see high AFR's at idle or cruising is because the PCM is using the narrowband O2 sensor to keep the AFR around 14.7. Narrowbands have opposite behavior to widebands. Our narrowbands are lean at .0 v and rich at 1 volts. A wideband however is rich at 0v and lean at 5v.
Hopefully that helps you understand the aux in table. The aux in table is setup for a wideband, not for a MAF signal. I was just being "sneaky" and using the aux in to read my stock MAF in values, you aren't supposed to do that.
Also, do you have the acceleration settings on? Make sure your fuel and limit settings are at 0% in your setup window. This is to add fuel quickly when you stab the throttle since the PCM takes a second to catch up. If you dump too much fuel you will misfire.
Ok I called AEM and as expected, they suggested increasing the resistance values between the O2 sensor and the AEM until I see the voltage at the PCM match what the AEM says in the software. Said to give 4.7k a shot but not to go over 10k. Hopefully this works or I'll be putting the FIC up for sale. The reason our cars blow up even with no mods is because we have a 16:1 AFR at full boost. This is very dangerous and the reason we have our hesitation. However, the AEM tech guy said not to try to tune the O2 without a scan tool since you need to know the STFT values.
I ordered an OBDII bluetooth module from some some Chinese gadget website called DealXtreme, but I am sure it is shipping overseas so it'll be a couple weeks before I see it.
I think I'll take my code reader back to Harbor Freight and see if they'll let me trade up to the scan tool, since I don't want to wait for the bluetooth module to come in.
I want the bluetooth module because eventually I plan to put a double DIN carputer in this car and have digital gauges for everything, and I can play with values on the fly. I really wish the AEM software could print the data log in real time. That would be nice. I'd like to avoid using windows too, but making a usb driver for the AEM for linux or android or meego may be pretty tough. May be stuck with windows.
I ordered an OBDII bluetooth module from some some Chinese gadget website called DealXtreme, but I am sure it is shipping overseas so it'll be a couple weeks before I see it.
I think I'll take my code reader back to Harbor Freight and see if they'll let me trade up to the scan tool, since I don't want to wait for the bluetooth module to come in.
I want the bluetooth module because eventually I plan to put a double DIN carputer in this car and have digital gauges for everything, and I can play with values on the fly. I really wish the AEM software could print the data log in real time. That would be nice. I'd like to avoid using windows too, but making a usb driver for the AEM for linux or android or meego may be pretty tough. May be stuck with windows.
Ricktalife
Member
- :
- 2012 Subaru WRX
Thanks for the info jdwk. I will double check the TPS with the ignition on to make sure it goes to 100%. Yes I did have bucking issues with the stock PCM when the AEM was not installed. I have a Prosport wideband, which is not the best I know, but it seems to work pretty well.
For the accelleration settings, I do have fuel and limit set to 0, but what should I have the other fields set to? I think the sensitivity is set to 51% and I did not touch that.
Also, I'm thinking that I might have to reconnect the MAF wire to my ECU. Not because I want to alter the MAF table, but because I think it is necessary for the MAF voltage fuel table. If the AEM doesn't have that wire connected, then it won't be able to fool the MAF into thinking the voltage is normal. Am I right, or completely off lol? How do you have the MAF wire connected on your setup jdwk? It's just a tap connection right?
For the accelleration settings, I do have fuel and limit set to 0, but what should I have the other fields set to? I think the sensitivity is set to 51% and I did not touch that.
Also, I'm thinking that I might have to reconnect the MAF wire to my ECU. Not because I want to alter the MAF table, but because I think it is necessary for the MAF voltage fuel table. If the AEM doesn't have that wire connected, then it won't be able to fool the MAF into thinking the voltage is normal. Am I right, or completely off lol? How do you have the MAF wire connected on your setup jdwk? It's just a tap connection right?
Destineal, do not buy this unit until I can successfully get O2 sensor control working. The FIC is done this way so it can be more of a universal unit and work with various O2 sensors and PCMs. The SSAFC is more car specific and can intercept the O2 siganl, which is why it is PSC-1-020 (as in version 20). This is also why it is more expensive.
The reason the resistor is needed is because when the FIC goes into output mode, you basically have two voltage sources in parallel (the O2 sensor and the FIC). Try taking a 9v batter and a 3v AA and put them in parallel and measure the voltage across them. You need enough resistance that current will not flow from the AEM to the O2 sensor, but not too much that the O2 sensor voltage drops since the FIC still needs to read that voltage for 50ms to alter it.
In this respect the SSAFC is the much better unit. What it doesn't do though is intercept the actual fuel injector signals, it just modifies the MAF signal. So you have much less finer control over fuel. You also can't do things like MAF clamp, throttle based fuel dumps, and of course timing control unless you go to a $700 SSFTC. At that point, you are better off going to a full standalone. And I am fairly close to doing that myself. If I can't get the O2 to work, I'll either go full stand alone or just go back to stock, and wait for a used B8 S4 to pop up or for the ATS-V to come out.
Seems silly to spend that kind of money on a car worth $7k, but I guess it's a hell of a lot cheaper than an S4.
The reason the resistor is needed is because when the FIC goes into output mode, you basically have two voltage sources in parallel (the O2 sensor and the FIC). Try taking a 9v batter and a 3v AA and put them in parallel and measure the voltage across them. You need enough resistance that current will not flow from the AEM to the O2 sensor, but not too much that the O2 sensor voltage drops since the FIC still needs to read that voltage for 50ms to alter it.
In this respect the SSAFC is the much better unit. What it doesn't do though is intercept the actual fuel injector signals, it just modifies the MAF signal. So you have much less finer control over fuel. You also can't do things like MAF clamp, throttle based fuel dumps, and of course timing control unless you go to a $700 SSFTC. At that point, you are better off going to a full standalone. And I am fairly close to doing that myself. If I can't get the O2 to work, I'll either go full stand alone or just go back to stock, and wait for a used B8 S4 to pop up or for the ATS-V to come out.
Seems silly to spend that kind of money on a car worth $7k, but I guess it's a hell of a lot cheaper than an S4.
Don't worry about the acceleration settings just yet. I'll play around with them and let you know if they help at all. Just keep fuel and limit to 0% and it won't do anything. The higher the sensitivity value the faster you'll have to jab the throttle for it to trigger. The fuel will be the % that it inceases the injector pulse, the limit is the max injector duty cycle, and the duration is how long you want the extra fuel to last. This could be fun later, but need to get a lot more sorted out before I start playing with it.
Is your MAF wire not hooked up? You want it to go straight into the PCM just like it was stock. Don't even tap it. Tapping it causes the voltage to drop a bit, and you'll have a hard time idling and can run lean. Like I said, I just have it because of the my super hot summers, that way I can add more fuel when my MAF values are lower at higher rpms.
I'll PM you my google voice number, give me a call, and I'll try to explain it better. Maybe we can figure out why you are bucking with the stock ECU too. Send me a link to a list of your mods too.
Is your MAF wire not hooked up? You want it to go straight into the PCM just like it was stock. Don't even tap it. Tapping it causes the voltage to drop a bit, and you'll have a hard time idling and can run lean. Like I said, I just have it because of the my super hot summers, that way I can add more fuel when my MAF values are lower at higher rpms.
I'll PM you my google voice number, give me a call, and I'll try to explain it better. Maybe we can figure out why you are bucking with the stock ECU too. Send me a link to a list of your mods too.
Ricktalife
Member
- :
- 2012 Subaru WRX
I've got a full mod list in my build thread which is linked in my sig. I might give you a call sometime tomorrow evening if I have a chance to fool with it. For everyone else reading this, I'll be sure to post any solutions that we find.
Right now my MAF wire is hooked up just like it was stock. I was referring to the AEM MAF in / out wire that is not hooked up. That's fine with me to leave it untapped, I was just wondering if that was limiting me from tuning the fuel properly. Do you think if I were to copy the values I saw on a youtube instructional video to the MAF table, that it could fool the MAF into reading those voltages without having to tap the actual MAF wire leading to the ECU? I think I will try that just for s**** and giggles.
Right now my MAF wire is hooked up just like it was stock. I was referring to the AEM MAF in / out wire that is not hooked up. That's fine with me to leave it untapped, I was just wondering if that was limiting me from tuning the fuel properly. Do you think if I were to copy the values I saw on a youtube instructional video to the MAF table, that it could fool the MAF into reading those voltages without having to tap the actual MAF wire leading to the ECU? I think I will try that just for s**** and giggles.
Ricktalife
Member
- :
- 2012 Subaru WRX
Ok so jdwk informed me that my long term fuel trim is probably all screwy because of the way that the AFC operated. Like any AFC user, I leaned out the fuel by adjusting the MAF voltage and tricking the ECU. Though because of that, the ECU automatically added fuel in the spots where it saw the fuel was leaning out. So because I had the AFC installed for about a year or so, the long term fuel trim had gathered its data completely from this method. Once I removed the AFC, the ECU continued to add fuel in those spots due to the LTFT data and hence the massive rich condition and bucking. So to solve this, jdwk told me that we can reset the LTFT by unhooking the battery and holding the brake until all the 'juice' is out. I'm guessing the pedal should go to the floor.
@jdwk - If I didn't explain that right please correct me lol.
@jdwk - If I didn't explain that right please correct me lol.
Yeah, your are just draining any electrical charge left in the car by having the brake lights turn on... you shouldn't notice any difference in the brake pedal. When the ecu finally loses all power, it will forget its long term data that it has stored. The car may act a little funny for the first few minutes after you start it back up.... but it should clear up as the ecu relearns everything.
Ricktalife
Member
- :
- 2012 Subaru WRX
Ahh okay that makes sense. I was looking at scan tools online and found a craftsman at sears for $180 that does pretty much all I need it to. Read / clear DTC's, Freeze Frame data, live PCM data stream. Only thing it doesn't have is an ABS scanner. That model is $300, and I don't think it's really worth that much extra. I think I'll wait to clear the LTFT after I either buy a scan tool or borrow one so I can hook it up and see what it tells me.
Hey I remembered this article and found the corresponding dyno plot.
Check out the purple line, the line after the reflash of the ECU.
See what happens when the fuel goes super lean right at 4k rpms. Massive dip in power.
I believe the idiots at Mazda or Callaway who did the reflash tried to lean out that transition just like I did, by changing the maps. Only it doesn't work. You only end up leaning out that last second before it transitions to open loop.
Here's a fun AFR plot with my O2 voltage clamped to 0.015. See how it goes really lean (17:1) even with my O2 voltage clamped, and then immediately drops off the face of the planet (9.7).
Check out the purple line, the line after the reflash of the ECU.
See what happens when the fuel goes super lean right at 4k rpms. Massive dip in power.
I believe the idiots at Mazda or Callaway who did the reflash tried to lean out that transition just like I did, by changing the maps. Only it doesn't work. You only end up leaning out that last second before it transitions to open loop.
Here's a fun AFR plot with my O2 voltage clamped to 0.015. See how it goes really lean (17:1) even with my O2 voltage clamped, and then immediately drops off the face of the planet (9.7).
Ricktalife
Member
- :
- 2012 Subaru WRX
I'm leaving work now so I will compare that with my datalog when I get home. I think it was pretty similar. So what the **** do we do with the hesitation then? Is it incurable?
It should be curable. We need to offset the O2 voltage enough to bring AFR's below 4250 rpm down to about as rich as the open loop rpms, and then use the scan tool to make sure we don't have any STFT. I'll just keep bringing it down until the open loop is smooth. This could take a while and requires I get the O2 manipulation to work flawlessly.
Once it is smooth and the trims are down to 0, then we can remove fuel above 4k to bring our AFR's back up.
I have a feeling that the PCM makes one last read of the O2 signal before transitioning into open loop mode. It will add whatever correction it needs to the fuel in the table for open loop mode, and why it goes crazy stupid rich for a second. This is the only explanation I can come up with for that graph I posted above. See the nice little spike of fuel, before it comes back down to make a more reasonable 11.6 AFR.
I think if I can bring the AFR down before that last read, it won't add any extra fuel and should go smoothly into open loop mode where I actually have control of the fuel. I think we'll always have a little bit of transition, but at least I should be able to bring those 16:1 AFRs down to a safe place.
Most of the zoom zoom boom stories I have read have not been high rpm explosions, they happen because the car is boosting in closed loop mode and trying to maintain a 14.7 AFR and failing miserably.
It's all theory. Hopefully, it works.
Once it is smooth and the trims are down to 0, then we can remove fuel above 4k to bring our AFR's back up.
I have a feeling that the PCM makes one last read of the O2 signal before transitioning into open loop mode. It will add whatever correction it needs to the fuel in the table for open loop mode, and why it goes crazy stupid rich for a second. This is the only explanation I can come up with for that graph I posted above. See the nice little spike of fuel, before it comes back down to make a more reasonable 11.6 AFR.
I think if I can bring the AFR down before that last read, it won't add any extra fuel and should go smoothly into open loop mode where I actually have control of the fuel. I think we'll always have a little bit of transition, but at least I should be able to bring those 16:1 AFRs down to a safe place.
Most of the zoom zoom boom stories I have read have not been high rpm explosions, they happen because the car is boosting in closed loop mode and trying to maintain a 14.7 AFR and failing miserably.
It's all theory. Hopefully, it works.
Ricktalife
Member
- :
- 2012 Subaru WRX
I wish you luck.
Ricktalife
Member
- :
- 2012 Subaru WRX
So I was just browsing the MSP section as usual and came upon a thread about MAF relocation and remembered when I tried relocating but had problems with running rich at idle and compressor surge. Another member pointed out rather quickly that the compressor surge was mostly likely caused by a weak vacuum source, since my vacuum block runs off of the intake manifold. That makes complete sense to me, and I plan on relocating the block before long.
Though I could not figure out why I was running rich at idle until now. As jdwk was telling me about the long term fuel trim, I was thinking about how long I had the AFC installed and how it was so screwed up when I removed it. I would bet that the reason i was running rich after relocating is also because of the fuel trim. I made a pretty drastic change in the flow of air that the MAF sees, and the ECU doesn't really like drastic changes.
The reason I'm posting these findings in here is because I think it is related to my problem with the FIC, and that resetting the fuel trims would be the best thing to do when changing air/fuel. Yes I already mentioned it, but I am just informing others that when installing a new tuner and removing a previous tuner it may be necessary.
Though I could not figure out why I was running rich at idle until now. As jdwk was telling me about the long term fuel trim, I was thinking about how long I had the AFC installed and how it was so screwed up when I removed it. I would bet that the reason i was running rich after relocating is also because of the fuel trim. I made a pretty drastic change in the flow of air that the MAF sees, and the ECU doesn't really like drastic changes.
The reason I'm posting these findings in here is because I think it is related to my problem with the FIC, and that resetting the fuel trims would be the best thing to do when changing air/fuel. Yes I already mentioned it, but I am just informing others that when installing a new tuner and removing a previous tuner it may be necessary.
Ricktalife
Member
- :
- 2012 Subaru WRX
Excellent, I can't wait to switch it back and put my vacuum block on the brake booster. I must be maxing out the port on the intake manifold cause I'm getting a little surge even though I'm running the dual setup
So I haven't given up yet. I bought a cheap (sort of) scan tool from Harbor Freight and watched my STFT's and LTFT's.
The LTFT's are updated much quicker than I thought. It takes just a few seconds of decent sized (+/- 5) STFT to start updating the LTFT. I can play with my MAF voltage on the fly and watch it update the tables. Kind of interesting.
It actually turns out to get zero trim, I just leave my MAF adjustment at 0, but it requires the exact resistance I came up with using the potentiometer. This is somewhere around 1060~1100 ohm. 1220 is shockingly, way too high.
So really the right way to get the correct resistance is to sit at idle and keep adjusting your potentiometer until the trims go to zero. I still highly recommend forgoing the MAF completely and just running it straight into the PCM with no tap, but if you are ambitious, this is the way to do it.
So the reset may not help (although if you are bucking it may not be in that rev and load long enough to update to tables).
In theory the MAF values shouldn't change regardless of whether its near the TB or the filter. The hotter compressed air should have the same effect on the MAF wire as the cooler uncompressed air. However, you do have a lot of turbulence there and pockets or gaps of air could be passing over the wire. I'd like to keep the benefits and merits of MAF relocation out of this discussion though.
What I have done though is manage to finally get the PCM to run rich. I don't think the offset and clamp have quite the effect they say they do, and as soon as the narrowband O2 reads the slightest bit rich, the PCM starts pulling fuel and under full throttle boost is too slow to correct back. A clamped .5v for instance fluctuates around .5v and the PCM is still capable of reading rich and leaning out.
Out of frustration I switched to percentage mode just to see, and I put everything at -49% above 15psia. To my amazement (I almost broke into tears), here is what I saw.
The PCM never leaned out the fuel mixture once I was in boost and the transition to open loop was silky smooth. The -49% kept the O2 reading rich the whole time. Why -.5V didn't work I don't know. I also played with a number of resistance values and currently it's at 20kohm, but I don't think that much is necessary.
Once I saw this I decided to back it off to -40% in the 3600rpm rpm range and see if I could get rid of the 9.5:1 dip.
The results were terrible. It reverted to the old 16:1 down to 9:1 fuel spike. So what happened was the O2 pegged to 1, and 40% of 1 is .60 and the ECU saw rich and pulled my fuel.
It appears the PCM has no concept of more rich or less rich, it's just rich or lean, and if it sees rich you are going to pay for it at high loads, because it's not going to be able to recorrect in time. At idle however it can, so the offset works at idle and light loads.
What I haven't tried is is just clamping the O2 voltage to something really lean like .2, so it never sees a rich value in boost. I think this is how the SSAFC works. My concern with that is the PCM catching on to a static O2 signal and throwing a code and going into limp mode, but it's probably worth a try. I think my 49% reduction is still on the hairy edge of being read as rich.
On the way to work I did quite a few 4th and 5th gear pulls, and it was perfectly smooth although definitely slow at 9.5:1 AFR. So I am not sure I buy that the PCM is capable of leaning it out again like AEM suggests. I did feel a small transition after coming off the highway and doing a 2nd gear pull, but then when I went to lunch I couldn't reproduce it in any gear. Slow and smooth.
So my goal this weekend is to lean it out, by pulling the fuel with respect to MAP.
There is also the inital spike of fuel that seems to occurr about a second after you put you cross 75% throttle mark if you are below 3k rpms. It could also be the transition from negative relative pressure to positive, but since the PCM has no MAP, it seems unlikely this would have an effect. The MAF also seems to cross 3.5V at that transition as well, which means it could be where the table start in the PCM. Not sure how I can fix that, but at least the AEM gives you a lot of options.
The LTFT's are updated much quicker than I thought. It takes just a few seconds of decent sized (+/- 5) STFT to start updating the LTFT. I can play with my MAF voltage on the fly and watch it update the tables. Kind of interesting.
It actually turns out to get zero trim, I just leave my MAF adjustment at 0, but it requires the exact resistance I came up with using the potentiometer. This is somewhere around 1060~1100 ohm. 1220 is shockingly, way too high.
So really the right way to get the correct resistance is to sit at idle and keep adjusting your potentiometer until the trims go to zero. I still highly recommend forgoing the MAF completely and just running it straight into the PCM with no tap, but if you are ambitious, this is the way to do it.
So the reset may not help (although if you are bucking it may not be in that rev and load long enough to update to tables).
In theory the MAF values shouldn't change regardless of whether its near the TB or the filter. The hotter compressed air should have the same effect on the MAF wire as the cooler uncompressed air. However, you do have a lot of turbulence there and pockets or gaps of air could be passing over the wire. I'd like to keep the benefits and merits of MAF relocation out of this discussion though.
What I have done though is manage to finally get the PCM to run rich. I don't think the offset and clamp have quite the effect they say they do, and as soon as the narrowband O2 reads the slightest bit rich, the PCM starts pulling fuel and under full throttle boost is too slow to correct back. A clamped .5v for instance fluctuates around .5v and the PCM is still capable of reading rich and leaning out.
Out of frustration I switched to percentage mode just to see, and I put everything at -49% above 15psia. To my amazement (I almost broke into tears), here is what I saw.
The PCM never leaned out the fuel mixture once I was in boost and the transition to open loop was silky smooth. The -49% kept the O2 reading rich the whole time. Why -.5V didn't work I don't know. I also played with a number of resistance values and currently it's at 20kohm, but I don't think that much is necessary.
Once I saw this I decided to back it off to -40% in the 3600rpm rpm range and see if I could get rid of the 9.5:1 dip.
The results were terrible. It reverted to the old 16:1 down to 9:1 fuel spike. So what happened was the O2 pegged to 1, and 40% of 1 is .60 and the ECU saw rich and pulled my fuel.
It appears the PCM has no concept of more rich or less rich, it's just rich or lean, and if it sees rich you are going to pay for it at high loads, because it's not going to be able to recorrect in time. At idle however it can, so the offset works at idle and light loads.
What I haven't tried is is just clamping the O2 voltage to something really lean like .2, so it never sees a rich value in boost. I think this is how the SSAFC works. My concern with that is the PCM catching on to a static O2 signal and throwing a code and going into limp mode, but it's probably worth a try. I think my 49% reduction is still on the hairy edge of being read as rich.
On the way to work I did quite a few 4th and 5th gear pulls, and it was perfectly smooth although definitely slow at 9.5:1 AFR. So I am not sure I buy that the PCM is capable of leaning it out again like AEM suggests. I did feel a small transition after coming off the highway and doing a 2nd gear pull, but then when I went to lunch I couldn't reproduce it in any gear. Slow and smooth.
So my goal this weekend is to lean it out, by pulling the fuel with respect to MAP.
There is also the inital spike of fuel that seems to occurr about a second after you put you cross 75% throttle mark if you are below 3k rpms. It could also be the transition from negative relative pressure to positive, but since the PCM has no MAP, it seems unlikely this would have an effect. The MAF also seems to cross 3.5V at that transition as well, which means it could be where the table start in the PCM. Not sure how I can fix that, but at least the AEM gives you a lot of options.
