P2121 Code and Erratic Acceleration
07-09-2015, 08:02 PM
#41
Senior Member
SL Member
Join Date: Nov 2005
Location: NJ
Posts: 866
Thanks man. I still don't comprehend what crush's AFR table is for though. I appologize for making you beat this dead horse, but could you explain where I need to use this in my process? Do you mean his AFR vs. O2 voltage table? That's for entering in your desired AFR into the O2 map. The purpose is to skew the reading of factory O2 sensor going to the ecu so that the car doesn't try to tune out the enrichment you dialed in on the fuel table when in closed loop. The ECU is always trying to run lean in closed loop. and will start to pull fuel, making you run lean under partial boost. Your O2 table (and Crushes) already addresses this so you don't have to worry too much. In a nutshell, imagine using the AFR table as a key for your target AFR at each psia vs rpm value
1. Unplug battery and ECM fuse- let stand for 10 minutes
2. Change 02 map to all zero's below boost (all 3.16 values in my table) Please back up your previous tune to another file before doing this so you can go back in case something doesn't work right. I also suggest at least modifying the IRT to 769.5us. It probably won't be perfect but it should be closer than the 486us default.
3. Flash FIC
4. Plug in battery and ECM fuse, start car and let idle for ~10 min
5. Monitor short term and long term fuel trims, as well as AFR through obd2
6. Adjust fuel map below boost to bring fuel trims near zero- nominally between -10 and +10 (on the y-axis, 13.1 should have values near 0; increasing in negative value towards 3, with values near -57?) In theory, you should be around -57, but more importantly, do what the car is telling you to do to minimize the magnitude of fuel trims. Your actual number could be quite different depending on other factors, not limited to injector latency, etc.
^Should I rev the car here to check for trims at higher rpms? How do I verify trims near boost if the car is at a stand still? What am I monitoring AFR for? For now, I wouldn't worry too much about this. You will not get near boost while at a stand still, but you can at least try to adjust the cells that you are operating within. After doing this, "blend in" the fuel values between the idle range through around 14.7 psi in the y-axis. You can use the software's built in calculate function to do this.
7. After fuel map is adjusted correctly and the values below boost are negative (If that's what the car wants) (unlike my current map)- turn off the car and plug the battery and ECM fuse- let stand for 10 minutes (You shouldnt' have to reset again. It's not going to make major LTFT changes in one running cycle)
8. Change 02 map to what it was before
9. Flash FIC
10. Plug in battery and ECM fuse, start car and let idle for ~10 min
11. Monitor short term and long term fuel trims, as well as AFR through obd2
12. In setup, change injector lag (increase value from default 486) until idle is smooth
^Does this update in real-time like the maps, or do I need to save and flash? I believe it does, but start with the 769.5us as I suggested above. It'll get you a lot closer. In fact, you may not really need to mess with this again, as the modifications you made to fuel trims in this range will have "tuned out" your latency issues anyway. Won't really hurt to try though.
Thanks again for all of your help- I owe you more than a few beers.
Another side note- I removed the ECU connector from my "boomslang harness" that contains the accelerator pedal sensor, and connected it directly into the ECU. The only connection going to the FIC from this connector is 12v power, so I simply directly spliced that onto the stock connector. If I get a p2121 or p2123 code again, I know the issue can be completely fixed by replacing the ECU and rewiring the wires going from the sensor to the ECU. It would be a hassle, but those are the only two places that issue could reside.
1. Unplug battery and ECM fuse- let stand for 10 minutes
2. Change 02 map to all zero's below boost (all 3.16 values in my table) Please back up your previous tune to another file before doing this so you can go back in case something doesn't work right. I also suggest at least modifying the IRT to 769.5us. It probably won't be perfect but it should be closer than the 486us default.
3. Flash FIC
4. Plug in battery and ECM fuse, start car and let idle for ~10 min
5. Monitor short term and long term fuel trims, as well as AFR through obd2
6. Adjust fuel map below boost to bring fuel trims near zero- nominally between -10 and +10 (on the y-axis, 13.1 should have values near 0; increasing in negative value towards 3, with values near -57?) In theory, you should be around -57, but more importantly, do what the car is telling you to do to minimize the magnitude of fuel trims. Your actual number could be quite different depending on other factors, not limited to injector latency, etc.
^Should I rev the car here to check for trims at higher rpms? How do I verify trims near boost if the car is at a stand still? What am I monitoring AFR for? For now, I wouldn't worry too much about this. You will not get near boost while at a stand still, but you can at least try to adjust the cells that you are operating within. After doing this, "blend in" the fuel values between the idle range through around 14.7 psi in the y-axis. You can use the software's built in calculate function to do this.
7. After fuel map is adjusted correctly and the values below boost are negative (If that's what the car wants) (unlike my current map)- turn off the car and plug the battery and ECM fuse- let stand for 10 minutes (You shouldnt' have to reset again. It's not going to make major LTFT changes in one running cycle)
8. Change 02 map to what it was before
9. Flash FIC
10. Plug in battery and ECM fuse, start car and let idle for ~10 min
11. Monitor short term and long term fuel trims, as well as AFR through obd2
12. In setup, change injector lag (increase value from default 486) until idle is smooth
^Does this update in real-time like the maps, or do I need to save and flash? I believe it does, but start with the 769.5us as I suggested above. It'll get you a lot closer. In fact, you may not really need to mess with this again, as the modifications you made to fuel trims in this range will have "tuned out" your latency issues anyway. Won't really hurt to try though.
Thanks again for all of your help- I owe you more than a few beers.
Another side note- I removed the ECU connector from my "boomslang harness" that contains the accelerator pedal sensor, and connected it directly into the ECU. The only connection going to the FIC from this connector is 12v power, so I simply directly spliced that onto the stock connector. If I get a p2121 or p2123 code again, I know the issue can be completely fixed by replacing the ECU and rewiring the wires going from the sensor to the ECU. It would be a hassle, but those are the only two places that issue could reside.
See above for my responses.
Last edited by bbsciontc; 07-10-2015 at 04:06 AM.
07-13-2015, 08:38 PM
#43
Senior Member
SL Member
Thread Starter
Join Date: Jul 2013
Posts: 100
So I followed your directions and was able to get the trims near zero at around -12 after the injector lag was set to 769.5. However, when I went to drive the car it was laggy and kept dying. The fuel trims were all over the place as soon as I put a load on it. I played around with fuel map quite a bit and tried the entire sequence a few times and was never able to get the car to drive without dying. I reverted back to the stock tune (with the o2 map slightly changed) and will have to wait to see the tuner. Thanks for your help though BB.
07-14-2015, 03:23 PM
#44
Senior Member
SL Member
Join Date: Nov 2005
Location: NJ
Posts: 866
Sorry to hear that it didn't work out for you. Might as well let the tuner take care of the non-boost region since they will be able to vary the load and adjust each of the cells. I was hoping that simply blending the fuel trims between idle and 14.7psia would be smooth enough, but it appears that's not the case.
I have to say that I'm still quite confused that you only got to -12% on your fuel map. It almost seems like the injectors are not flowing like they should, or that you have a vacuum leak. See if the shop can do a pressure check of the system and verify there are no leaks.
I have to say that I'm still quite confused that you only got to -12% on your fuel map. It almost seems like the injectors are not flowing like they should, or that you have a vacuum leak. See if the shop can do a pressure check of the system and verify there are no leaks.
07-14-2015, 05:35 PM
#45
Senior Member
SL Member
Thread Starter
Join Date: Jul 2013
Posts: 100
Sorry to hear that it didn't work out for you. Might as well let the tuner take care of the non-boost region since they will be able to vary the load and adjust each of the cells. I was hoping that simply blending the fuel trims between idle and 14.7psia would be smooth enough, but it appears that's not the case.
I have to say that I'm still quite confused that you only got to -12% on your fuel map. It almost seems like the injectors are not flowing like they should, or that you have a vacuum leak. See if the shop can do a pressure check of the system and verify there are no leaks.
I have to say that I'm still quite confused that you only got to -12% on your fuel map. It almost seems like the injectors are not flowing like they should, or that you have a vacuum leak. See if the shop can do a pressure check of the system and verify there are no leaks.
07-16-2015, 03:17 AM
#47
Senior Member
SL Member
Join Date: Nov 2005
Location: NJ
Posts: 866
Are you back to using the original tune before any map modifications? I don't mean to bring up the vacuum leak thing again, but that is one possible cause for that code (O2 sensor pegged lean). If unmetered air is getting into the engine, you would end up running lean. If it's so bad that only a small portion of the air is entering through the MAF, that would cause the ECU to not inject enough fuel. Of course, it could also be the sensor itself, but that would need to be diagnosed.
If I have time, I'll try to check the idle AFR and MAF voltage on my car now that I got it back. It's currently N/A with stock injectors, so it would be a good reference point.
If I have time, I'll try to check the idle AFR and MAF voltage on my car now that I got it back. It's currently N/A with stock injectors, so it would be a good reference point.
Last edited by bbsciontc; 07-17-2015 at 02:21 PM.
07-18-2015, 12:11 AM
#48
Senior Member
SL Member
Thread Starter
Join Date: Jul 2013
Posts: 100
Are you back to using the original tune before any map modifications? I don't mean to bring up the vacuum leak thing again, but that is one possible cause for that code (O2 sensor pegged lean). If unmetered air is getting into the engine, you would end up running lean. If it's so bad that only a small portion of the air is entering through the MAF, that would cause the ECU to not inject enough fuel. Of course, it could also be the sensor itself, but that would need to be diagnosed.
If I have time, I'll try to check the idle AFR and MAF voltage on my car now that I got it back. It's currently N/A with stock injectors, so it would be a good reference point.
If I have time, I'll try to check the idle AFR and MAF voltage on my car now that I got it back. It's currently N/A with stock injectors, so it would be a good reference point.
I'll rent a pressure test kit to find out for sure about the vac leak, but I've sprayed everything with soapy water twice.
Sadly I won't be able to get to the tuner till probably two weeks from now because I'm traveling when I was scheduled to drop the car off.
07-21-2015, 03:34 AM
#49
Senior Member
SL Member
Join Date: Nov 2005
Location: NJ
Posts: 866
Here are my AFR values, O2 sensor voltage and MAF flow rate at idle. This is after a 30 minute drive, idling at ~730rpm on a 92 degree summer day.
Your MAF values at a fully warm idle should be very similar to this. If there is a large discrepancy, you probably have a leak.
Your MAF values at a fully warm idle should be very similar to this. If there is a large discrepancy, you probably have a leak.
07-21-2015, 12:27 PM
#50
Senior Member
SL Member
Join Date: Nov 2005
Location: NJ
Posts: 866
That would be great! I'm traveling until the middle of next week so I'll have a chance to play with it then. I reverted back to the tune from the tuner and now it has been staying in that cel. I have an antifowler on the secondary o2 and I replaced my primary o2 like a year and a half ago. Does it make sense to order a replacement for either since I've got time to have it shipped in?
I'll rent a pressure test kit to find out for sure about the vac leak, but I've sprayed everything with soapy water twice.
Sadly I won't be able to get to the tuner till probably two weeks from now because I'm traveling when I was scheduled to drop the car off.
I'll rent a pressure test kit to find out for sure about the vac leak, but I've sprayed everything with soapy water twice.
Sadly I won't be able to get to the tuner till probably two weeks from now because I'm traveling when I was scheduled to drop the car off.
Another thing to check as far as a leak goes is the PCV hose. Someone on here posted about a lean issue at idle, becoming rich under boost. In his case it seems that the pcv hose had blown off when he overboosted a bit and was the source of his vacuum leak. It was drawing in air at vacuum and releasing air under boost.
07-21-2015, 08:04 PM
#51
Senior Member
SL Member
Thread Starter
Join Date: Jul 2013
Posts: 100
Don't worry about the secondary o2 sensor. That just checks for the primary cat and has no effect on how the engine runs. If you changed your primary, I doubt it would have gone bad so quickly. I'd suspect something else. In any case, you should be able to diagnose the o2 sensor values by comparing it the readings on a wideband at the tuner.
Another thing to check as far as a leak goes is the PCV hose. Someone on here posted about a lean issue at idle, becoming rich under boost. In his case it seems that the pcv hose had blown off when he overboosted a bit and was the source of his vacuum leak. It was drawing in air at vacuum and releasing air under boost.
Another thing to check as far as a leak goes is the PCV hose. Someone on here posted about a lean issue at idle, becoming rich under boost. In his case it seems that the pcv hose had blown off when he overboosted a bit and was the source of his vacuum leak. It was drawing in air at vacuum and releasing air under boost.
07-25-2015, 11:21 PM
#52
Senior Member
SL Member
Thread Starter
Join Date: Jul 2013
Posts: 100
Still not seeing a leak anywhere. I checked the o2 voltages at the connector as well as resistance on the sensor and both were within spec. I was driving today and the car was impossible to keep running for longer than 20 seconds. I stumbled home, reset the ecu, and tried playing around with the fuel map to get trims lower.
As I was doing so, the car was still running very rough and even with the fuel trims low, it would die. I reset the ecu again and put the original tune back on. Waited five minutes and tried to drive around the block. The car was surging and dying like it was misfiring and I had to restart the car a few times. The fuel trims were below 10% the entire time and afr was 13-18. Finally I was able to get back in my parking spot so I ran a scan an P0101 showed up. I walked to the engine bay (car off but ignition on) and heard a buzzing noise- it sounds like it is coming from the throttle body:
I inspected the maf and nothing seems out of the ordinary.I then turned the car off and back on and the noise was gone. I let the car idle for 5 minutes and again, fuel trims were below 10% with afr at 13-14. When I gave it some gas, the rpms felt laggy and the car stumbled and died. Fuel trims were below 10% but afr was between 16-20 for all of 1.5k-4k rpms. Trying to drive the car again is obviously a no go.
I unplugged the battery and am letting the car sit for the day. I'm unsure of what to do at this point besides hope it's better in the morning or get a tow to the tuner.
As I was doing so, the car was still running very rough and even with the fuel trims low, it would die. I reset the ecu again and put the original tune back on. Waited five minutes and tried to drive around the block. The car was surging and dying like it was misfiring and I had to restart the car a few times. The fuel trims were below 10% the entire time and afr was 13-18. Finally I was able to get back in my parking spot so I ran a scan an P0101 showed up. I walked to the engine bay (car off but ignition on) and heard a buzzing noise- it sounds like it is coming from the throttle body:
I inspected the maf and nothing seems out of the ordinary.I then turned the car off and back on and the noise was gone. I let the car idle for 5 minutes and again, fuel trims were below 10% with afr at 13-14. When I gave it some gas, the rpms felt laggy and the car stumbled and died. Fuel trims were below 10% but afr was between 16-20 for all of 1.5k-4k rpms. Trying to drive the car again is obviously a no go.
I unplugged the battery and am letting the car sit for the day. I'm unsure of what to do at this point besides hope it's better in the morning or get a tow to the tuner.
07-27-2015, 01:17 PM
#53
Senior Member
SL Member
Join Date: Nov 2005
Location: NJ
Posts: 866
Man, this is very strange. I'll keep trying to suggest things as they come to mind, but I bet you checked many of them already. I feel like there's something else going on here. I'm sure you did the resistor mod already, correct? Have you verified that you're running the appropriate slave version for the FIC (z104.hex)? I just installed my vacuum manifold, gauges and FIC yesterday, running stock injectors). Everything is pretty much zeroed out, and it's running fine except for a stumble I got on two WOT runs. I may just relash the FIC to make sure it's the right version since I bought the kit used.
Also, what value did you get when you checked your maf output at idle on the Torque app? It should be similar to what I posted and be very steady.
Also, what value did you get when you checked your maf output at idle on the Torque app? It should be similar to what I posted and be very steady.
07-27-2015, 02:57 PM
#54
Senior Member
SL Member
Thread Starter
Join Date: Jul 2013
Posts: 100
I'll confirm later today, but I believe it was around 30cfm. I'll also confirm slave version but I'm 90 percent it is correct. And yes I have completed the resistor mod. I tried to start the car today and let it idle and it was idling terribly rough so I shut it down. Probably will have to tow to a shop or the tuner now :/
07-27-2015, 09:06 PM
#55
Senior Member
SL Member
Join Date: Nov 2005
Location: NJ
Posts: 866
So I went and looked back at everything we had talked about, especially after the MAF code you threw, and came across this on page one of your thread. This is your MAF table, and I think this explains a lot of what's going on. Look it over and I'll continue below:
Ok, so remember how I was saying that it made no sense that you were running 0 on your fuel map instead of something like -55% or something like that? Well, that's what led me to believe you had a vacuum leak since it looked like your ECU was not metering a whole bunch of air and potentially running lean. Well my friend, I think this is the culprit!
Look at the lower end of you MAF table around 400 and 800 RPM. It's set to output nearly half of what the MAF is actually reading. I think what your tuner did, was try to tune your car by altering the MAF signal rather than with the fuel map (at least in the closed loop region). This is completely unnecessary and ignores the greatest advantage of the FIC which is that in interrupts and modifies the injector pulse width directly. What they did was what we used to do back in the day when we only had crappy Apexi SAFCs which couldn't do anything besides modify the MAF signal. More still, the change in voltage across the rpm range aren't even linear! It randomly goes up and then down. This tells me that the tuner tweaked your maf voltages to get the desired at AFR, then proabably did some last minute fine tuning with with fuel map.
Your goal with the MAF table is for the output to be very close to the input voltage across the board, except for regions in boost. Here you may want to "clamp" the voltage so the ECU doesn't see the high air flow from the turbo (car might freak out if it's sees say, 5V at 2800 rpm for instance).
Ok, so now what should you do? Well here's the dilemma. Your current fuel map was tuned using this MAF table. If you change this, it's going to completely invalidate your tune, and you may not even be able to start the car with the current fuel map. This kinda sucks, but this is how I would proceed:
1. What you need to do is fix your MAF table first. Use the one that Crush posted exactly. You'll see that his is very linear, only hiding/clamping the very highest voltage ranges (approaching 5V). You'll also see how at idle/low rpm, the in and out voltages are nearly identical.
2. Before you start the car, you'll have to completely redo the fuel map. This may not be fun, since you'll have to start from scratch. My recommendation would be to create your own basemap. It's actually pretty easy to do with AEM built in tool. First, you'll have to completely zero out every cell in the fuel map, then follow this really helpful tutorial: Fried Rice BANZAI: Tuning: AEM FIC Fuel Map Steps!
Read the whole thing, it is very insightful. In theory the software should create a relatively conservative tune to run on, but I would try to avoid boost. I would use the create base map feature and assume the stock injectors are 350cc. This should err on the side of more fuel which is a safer place to be. I bet then, you'll have the -50something% I would expect to see in the fuel table.
3. Your ignition map looks decent enough, albeit a little aggressive. This is moot anyways since it will need to be adjusted after a retune. You can keep it for now, but again, stay out of boost until you're retuned. I would however change your injector response time to around 800-850us or so. It should fall somewhere in between the stock injectors and your 750s.
4. Remember when we changed your O2 map to voltage, then fixed the weird middle region to match Crush's? Do that again.
In summary, copy crushes MAF map--> zero fuel map completely and create new base map assuming stock injectors are 350cc--> change IRT to 800us--> copy crushes O2 map in voltage mode, cross fingers and fire it up! Be sure to back up your original file if you somehow need to go back.
Assuming the car starts and idles, follow what I said before to adjust your fuel map in the idle region using the scan tool. You may need to zero out the idle region of the O2 map so you can see what the car is targeting and bring the fuel trims close to zero. Vary the rpm a little bit with the throttle and do this is a few regions to smooth out the fuel trims.
I'm really hoping this is it!
Ok, so remember how I was saying that it made no sense that you were running 0 on your fuel map instead of something like -55% or something like that? Well, that's what led me to believe you had a vacuum leak since it looked like your ECU was not metering a whole bunch of air and potentially running lean. Well my friend, I think this is the culprit!
Look at the lower end of you MAF table around 400 and 800 RPM. It's set to output nearly half of what the MAF is actually reading. I think what your tuner did, was try to tune your car by altering the MAF signal rather than with the fuel map (at least in the closed loop region). This is completely unnecessary and ignores the greatest advantage of the FIC which is that in interrupts and modifies the injector pulse width directly. What they did was what we used to do back in the day when we only had crappy Apexi SAFCs which couldn't do anything besides modify the MAF signal. More still, the change in voltage across the rpm range aren't even linear! It randomly goes up and then down. This tells me that the tuner tweaked your maf voltages to get the desired at AFR, then proabably did some last minute fine tuning with with fuel map.
Your goal with the MAF table is for the output to be very close to the input voltage across the board, except for regions in boost. Here you may want to "clamp" the voltage so the ECU doesn't see the high air flow from the turbo (car might freak out if it's sees say, 5V at 2800 rpm for instance).
Ok, so now what should you do? Well here's the dilemma. Your current fuel map was tuned using this MAF table. If you change this, it's going to completely invalidate your tune, and you may not even be able to start the car with the current fuel map. This kinda sucks, but this is how I would proceed:
1. What you need to do is fix your MAF table first. Use the one that Crush posted exactly. You'll see that his is very linear, only hiding/clamping the very highest voltage ranges (approaching 5V). You'll also see how at idle/low rpm, the in and out voltages are nearly identical.
2. Before you start the car, you'll have to completely redo the fuel map. This may not be fun, since you'll have to start from scratch. My recommendation would be to create your own basemap. It's actually pretty easy to do with AEM built in tool. First, you'll have to completely zero out every cell in the fuel map, then follow this really helpful tutorial: Fried Rice BANZAI: Tuning: AEM FIC Fuel Map Steps!
Read the whole thing, it is very insightful. In theory the software should create a relatively conservative tune to run on, but I would try to avoid boost. I would use the create base map feature and assume the stock injectors are 350cc. This should err on the side of more fuel which is a safer place to be. I bet then, you'll have the -50something% I would expect to see in the fuel table.
3. Your ignition map looks decent enough, albeit a little aggressive. This is moot anyways since it will need to be adjusted after a retune. You can keep it for now, but again, stay out of boost until you're retuned. I would however change your injector response time to around 800-850us or so. It should fall somewhere in between the stock injectors and your 750s.
4. Remember when we changed your O2 map to voltage, then fixed the weird middle region to match Crush's? Do that again.
In summary, copy crushes MAF map--> zero fuel map completely and create new base map assuming stock injectors are 350cc--> change IRT to 800us--> copy crushes O2 map in voltage mode, cross fingers and fire it up! Be sure to back up your original file if you somehow need to go back.
Assuming the car starts and idles, follow what I said before to adjust your fuel map in the idle region using the scan tool. You may need to zero out the idle region of the O2 map so you can see what the car is targeting and bring the fuel trims close to zero. Vary the rpm a little bit with the throttle and do this is a few regions to smooth out the fuel trims.
I'm really hoping this is it!
07-28-2015, 12:36 AM
#57
Senior Member
SL Member
Thread Starter
Join Date: Jul 2013
Posts: 100
I have the car idling just a little rough. I got everything done mines the create base Cal. I modified my old tune and essentially dropped the values by a percentage and tweaked it some. Afr is around 14. Short term between zero and 4. Long term seems stuck at - 20. Haven't tried to drive it yet though
07-28-2015, 02:54 AM
#58
Senior Member
SL Member
Join Date: Nov 2005
Location: NJ
Posts: 866
So when you use the create base map function, you'll have to zero everything first, then enter 750 for new injectors and 350 for stock (I think you said you were running 750cc). When I did this, I got -53.1% at lower rpm, and increasing to positive as boost increases towards max.
07-28-2015, 05:15 PM
#60
Senior Member
SL Member
Thread Starter
Join Date: Jul 2013
Posts: 100
I will try tonight. Thank you. I was able to drive the car again. This time about 20 miles. Short term fuel trim bounced around -7 to +2. Long term bounced around -35 to - 10. On the interstate it dropped low. Driving up a steep hill and lots of stop lights made long term go up. Currently I have p0172 system to rich and p0102 maf circuit low but it's 95 degrees out and the car has never run good that hot.