How is this possible (650cc idle)

[97moneyhole]

Heres my situation, about a year ago someone broke into my car and ripped out my SAFC. Ever since then, (after replacing my SAFC w/ an Apexi Neo) i have been having the problem of keeping a consistant A/F at WOT. I was unsure if my Neo might have a problem so i disconnected it and fired up the car. To my suprise, the car idles as if it came stock with 650cc injectors 14.6:1 A/F. I let it run for 15min in the driveway waiting for it to choke itself out and it never did. Any thoughts or ideas would be greatly apperciated.

 

[kenamond]

But I bet your fuel trims are very negative. At idle, the car is in closed-loop mode and using the O2 sensor to compute STFT and LTFT that are required to get a stoich AFR. It's working. But when you go full throttle or otherwise enter open-loop mode, you'll be pig rich.

If you disconnect the battery for 10 minutes and start the car back up, I bet it starts out rich and eventually settles back at 14.7-ish.

 

[97moneyhole]

yes, it does go pig rich at start up and then reads normal 14.7:1 after a few seconds. I understand what you are saying about the fuel trims but shouldnt the car die at idle with 650cc injectors and no fuel management system to compensate for that much fuel.

 

[kenamond]

Maybe it would before the trims get adjusted by the ECU, but the O2 is going to be constantly saying "RICH!!!" right from the get-go, and the STFT will probably adjust pretty quickly. If you can log STFT, I'd start it up and watch how it evolves. But if it's in learn mode, it'll shift the STFT value into the LTFT and reset the STFT back to zero. I guess it'd be good to watch both if you want to verify that this is what's going on.

 

[97moneyhole]

okay, but i thought the stock pcm would be limited with it's ability to adjust the ltft. i hate to ask a hypothetical question, but my pcm alone with no aids wouldn't be able to run the car with, say 1000cc injectors, would it?

i ask because i have the same set up right now as this guy(http://www.dsmtuners.com/forums/pro...r-wont-idle-after-650cc-injector-install.html), and i don't understand how his car could idle like crap(how i believe it is supposed to run with no fuel controller), and mine is idling perfectly fine.

 

[97moneyhole]

any ideas?

 

[kenamond]

I'm not sure how high/low the ECU will let the trims get before it gives up and/or throws a code. Whether or not you could idle on 1000cc injectors isn't really the point, because you couldn't drive the car. As for the other thread, I don't know why he has problems at idle and you don't, but there are a few differences; the 1g ECU may be slower to adjust STFTs (a pure guess), the 190fp will overrun the stock FPR at idle making it even richer, and he has a maf-t that he claims is set up correctly but might not be. Or maybe your trims were already negative from your tuning attempts and that helped when you pulled the neo.

If you have a logger, check your STFT and LTFT and see what they are.

 

[97moneyhole]

i hate saying this but i do not have a logger at this time, (i wish i did) but i am just wondering if there is any componet (IAC, PCM, ect) that could have failed or be out of range that would allow me to idle normal with my set up.

(side note)
can a 2gen pcm learn the fuel trims from a SAFC over an extended period of time and start to depend on the numbers?

I know this is a real off the wall subject but i do appreciate your help.

 

[kenamond]

This is long. Prepare thyself!

I'm honestly not suprised that you get 14.7AFR at idle. That's what the ECU is supposed to do, and it's apparently doing it correctly. Maybe something else is screwed up and that was what was giving you fits when trying to tune with the Neo, but the fact that the idle AFR is correct tells me something is working rather than not working. Don't know what else to say.

As far as trims, if you keep power to the ECU (battery still connected), it remembers the trims. So when you yanked the Neo, if you left the battery connected, it would still have the trims from when the Neo was installed.

However, the trims can change constantly when the ECU is in learn mode. Let me explain all of this a bit more - to the extent of my understanding, that is.

When the ECU is at low load, it will run in closed loop mode. In this case, it's "listening" to what the O2 sensor is saying. Otherwise, it's in open loop mode and isn't using the O2 sensor for **anything**.

The stock O2 sensor is a narrowband sensor. It's great for telling you when your AFR is 14.7, because it's calibrated to put out 0.5v at 14.7AFR. However, if you go a bit lean or rich, the voltage changes quite a bit, and it's fairly unreliable for judging AFR other than 14.7.

So when the ECU is in closed-loop mode, it wants to run at an AFR of 14.7. To do this, it tries to get the fuel mix so that the O2v alternates above and below 0.5v. If it stays above 0.5v, you're richer than 14.7. If it stays below 0.5v, you're leaner than 14.7. That's all the O2 sensor is used for: above or below stoich.

So if the O2 sensor is reading consistently rich or lean and the ECU is in closed loop mode, the ECU adjusts the fuel via the short term fuel trim (STFT). STFT will be positive if the O2v were consistently lean; the positive STFT means, "add a bit more fuel". If STFT is negative, that means you were rich and needed to subtract fuel. All of this adding and subtracting is done relative to the fuel map. So the ECU looks at the fuel map, finds how much fuel to inject the next time around, looks at the STFT and adjusts what it got from the map to either add a bit or subtract a bit of fuel. Then it checks the O2v signal to see if the STFT needs to be adjusted some more. So all of this time, the STFT can be changing a bit here and there so that the AFR stays stoich. If the STFT is a bit too high, the O2v will read rich (and not switch back and forth rich/lean/rich/lean), and the ECU will decrease the STFT a bit. That's why this is all called closed loop mode; there is a loop of (1)inject fuel based on the map and STFT, (2) check the O2v, (3) adjust the STFT based on (2). Repeat. The whole system has feedback - cause and effect - that it relies on to keep the AFR at 14.7.

Now if the conditions are right, the ECU goes into learn mode. I forget the exact conditions, but they make sure that "all is well enough that I trust the STFT value I've got and want to use it long-term". In this case, it will *add* the STFT value to either the low or mid long term fuel trim (LTFT Lo and LTFT Mid). The LTFT Lo is used when the airflow is below some threshold while the LTFT Mid is used above that airflow threshold. I guess they called it Mid because at high airflow, the ECU goes to open loop mode where no trims are used (just "trusts" the fuel map as-is). If trims were used in open-loop mode, they'd then be called LTFT Hi. Note, on 1gs, they have a Lo and Hi, but no Mid. Anyway, back to learn mode...

So if conditions are right, the ECU adds the STFT value to the LTFT (Lo or Mid depending on airflow) and then sets the STFT back to zero. Now I lied to you a bit ago, but I'll fix it now...when the ECU computes how much fuel is needed, it actually looks at the fuel map and adds both the STFT and LTFT corrections and uses that for delivering fuel. So when in learn mode and the STFT has just been copied to the LTFT and the STFT goes back to zero, immediately after that, the LTFT now stores the old LTFT plus the STFT value just before it was reset. If the car continues to behave the same way, the STFT will stay at zero, because the new LTFT now includes what had been "learned" via the STFT. So the ECU has "learned" or memorized what fuel adjustment is needed to keep the AFR at 14.7.

Now you can be in learn mode quite a lot of the time. But if the STFT is zero, it doesn't do anything to the LTFT. But if it's in learn mode and the STFT changes, it immediately adds it to the LTFT and sets STFT back to zero.

So it looks to me like your trims are compensating for the larger injectors; all of the tools are there for the ECU to use to make those adjustments. BUT...that all goes out the window when you tromp the gas pedal (that forces open loop mode, and no trims are used, and you go pig rich all of the sudden).

 

[CTP]

I can tell you from experience that the stock ECU cannot adjust the trims enough to compensate for 680's, so I don't see it compensating for 650's either.

Are you sure you don't have a vacuum or post MAF leak somewhere? That would explain your inconsistant WOT AFR's and also the fact that you're able to idle 650's with no adjustment.

 

[kenamond]

I can tell you from experience that the stock ECU cannot adjust the trims enough to compensate for 680's, so I don't see it compensating for 650's either.

Are you sure you don't have a vacuum or post MAF leak somewhere? That would explain your inconsistant WOT AFR's and also the fact that you're able to idle 650's with no adjustment.

Good points. Was that on bone stock fuel system (aside from injectors)? I'm curious how close to the edge he is. And do you happen to know how far the ECU will let the trims get before it gives up, throws a code, does something "special"?

What is your idle vaccuum when it finally settles on 14.7? If it's off, that would be an indicator of a post-TB vaccuum leak.

I'd get a logger if I were you.

 

[JeffGST]

ECU's capabilities - 2G

LTFT lo = The long term fuel trim is able to take away a maximum of -12.5% of fuel away

STFT = The maximum amount at any given moment the ECU can take away (IN ADDITION TO THE LTFT's) is -16.8%

For a grand total of -29.3% of fuel reduction capability.

450cc stock injectors
650cc current injectors

((450/650)-1.0) = 30% change in injector flow

So the ECU BARELY has enough room to fully pull away all that fuel with the 650 injectors. If you have a boost leak after the MAS, it will help the ECU compensate more for the airflow reduction.

 

[spyderturbo007]

But just remember Jeff, that 1% when talking about the ECU's adjustment of trims, is not the same as 1% when talking about the SAFC's manipulation of the airflow signal. Basically, I'm saying that if your STFT is 0 and the LTFT is -5, the, adjusting the SAFC, at that Ne point, to -5% isn't going to bring the FT's to 0.

If I remember from my SAFC days , 1% on the SAFC corresponds to about 3% when dealing with the FT's.

 

[kenamond]

ECU's capabilities - 2G

LTFT lo = The long term fuel trim is able to take away a maximum of -12.5% of fuel away

STFT = The maximum amount at any given moment the ECU can take away (IN ADDITION TO THE LTFT's) is -16.8%

For a grand total of -29.3% of fuel reduction capability.

450cc stock injectors
650cc current injectors

((450/650)-1.0) = 30% change in injector flow

So the ECU BARELY has enough room to fully pull away all that fuel with the 650 injectors. If you have a boost leak after the MAS, it will help the ECU compensate more for the airflow reduction.

Excellent info! Thanks! So what does the ECU do in learn mode when LTFT+STFT exceeds -12.5%? And what happens when any trim hits that limit? Does the ECU just cap it at the limit, throw a code, go to limp mode?

 

[spyderturbo007]

So what does the ECU do in learn mode when LTFT+STFT exceeds -12.5%? And what happens when any trim hits that limit? Does the ECU just cap it at the limit, throw a code, go to limp mode?

The good ol' P0170 "Fuel Trim Malfunction Bank 1" CEL. When my O2 got beat up with a hammer and my trims maxed, the CEL was tripped and the logger continued to show maxed trims all the way home. I must say, that it didn't seem like I was dumped into limp mode, comparing to when I forgot to plug in the MAS, but you sure could watch the gas gauge move.

 

[kenamond]

But just remember Jeff, that 1% when talking about the ECU's adjustment of trims, is not the same as 1% when talking about the SAFC's manipulation of the airflow signal. Basically, I'm saying that if your STFT is 0 and the LTFT is -5, the, adjusting the SAFC, at that Ne point, to -5% isn't going to bring the FT's to 0.

If I remember from my SAFC days , 1% on the SAFC corresponds to about 3% when dealing with the FT's.

If you don't mind, could you explain why? First, I don't own SAFC, so I don't know what the Ne point is, but I think I understand the general idea of SAFC.

Just to be clear, we were discussing whether the ECU could compensate for 650s in closed loop mode with trims, not what the SAFC settings should be, but the info is still good to know so that someone doesn't assume they can directly use LTFT values in the SAFC.

 

[spyderturbo007]

If you don't mind, could you explain why?

To be honest, I really don't know, but I can speculate which might lead to some interesting conversation. The ECU adjusts trims based on the O2 sensor voltage and the SAFC adjusts the MAS frequency based on the user's entered correction. Apparently they aren't using the same factor.

First, I don't own SAFC, so I don't know what the Ne point is, but I think I understand the general idea of SAFC.

The Ne Point is a user-definable rpm point. Basically each map is composed of 16 Ne Points that can be chosen by the user across the entire rpm range. So, for example, 1k, 1.6k, 2k, 2.6k, 3k, 3.6k, etc. Each Ne Point (rpm point) has it's own correction factor on whatever map you are dealing with. That is where the TPS trigger points come into play, which again, is another user-definable setting.

There are 2 different maps in the SAFC, Hi throttle and Lo throttle. Using the TPS trigger points, the SAFC reads the TPS signal and decides what map it should use at that point in time. Below the lower TPS trigger point and you are on the Lo throttle map. Above the upper TPS trigger point and you are on the Hi throttle map. Anything in between the TPS trigger points and the SAFC uses a linear extrapolation of both the Lo and Hi throttle maps.

Just to be clear, we were discussing whether the ECU could compensate for 650s in closed loop mode with trims, not what the SAFC settings should be, but the info is still good to know so that someone doesn't assume they can directly use LTFT values in the SAFC.

Yeah, but I saw Jeff's calculation and he was using the assumption that the adjustments on the SAFC mirrored the ECU's adjustments of fuel trims, which is incorrect. He was using a 1:1 relationship when in fact, it's more like 3:1.

 

[kenamond]

To be honest, I really don't know, but I can speculate which might lead to some interesting conversation. The ECU adjusts trims based on the O2 sensor voltage and the SAFC adjusts the MAS frequency based on the user's entered correction. Apparently they aren't using the same factor.



The Ne Point is a user-definable rpm point. Basically each map is composed of 16 Ne Points that can be chosen by the user across the entire rpm range. So, for example, 1k, 1.6k, 2k, 2.6k, 3k, 3.6k, etc. Each Ne Point (rpm point) has it's own correction factor on whatever map you are dealing with. That is where the TPS trigger points come into play, which again, is another user-definable setting.

There are 2 different maps in the SAFC, Hi throttle and Lo throttle. Using the TPS trigger points, the SAFC reads the TPS signal and decides what map it should use at that point in time. Below the lower TPS trigger point and you are on the Lo throttle map. Above the upper TPS trigger point and you are on the Hi throttle map. Anything in between the TPS trigger points and the SAFC uses a linear extrapolation of both the Lo and Hi throttle maps.




Yeah, but I saw Jeff's calculation and he was using the assumption that the adjustments on the SAFC mirrored the ECU's adjustments of fuel trims, which is incorrect. He was using a 1:1 relationship when in fact, it's more like 3:1.

More great info, and more thanks! As for Jeff's post, I didn't see anything about SAFC in there...??? I read it to mean that the stock ECU could compensate for 29.3% additional fuel via the STFT and LTFT trims.

EDIT: If the SAFC is directly altering the Hz values and not the airflow values, then that might explain why the SAFC and STFT/LTFT are different. The Hz values may not be linear functions of airflow (as a completely fabricated example, if you double Hz, that might be 4x more airflow). What exactly is the SAFC modifying?

 

[spyderturbo007]

As for Jeff's post, I didn't see anything about SAFC in there...??? I read it to mean that the stock ECU could compensate for 29.3% additional fuel via the STFT and LTFT trims.

Oops. I don't know what the hell I was thinking about when I replied to that. Provided that the ECU's FT % is in fuel and not something stupid like O2 trim or % voltage deviation from stoich, then that is correct. Please disregard my comments in that response. I'm going to go crawl back into my hole now.

EDIT: If the SAFC is directly altering the Hz values and not the airflow values, then that might explain why the SAFC and STFT/LTFT are different. The Hz values may not be linear functions of airflow (as a completely fabricated example, if you double Hz, that might be 4x more airflow). What exactly is the SAFC modifying?

You are correct, which is what I was trying to say above, but I don't think it came out right. The SAFC does indeed alter the raw MAS frequency which is then converted / interpreted / whatever, by the ECU as an airflow value. But to this day, I have yet to find anything that allows for a correlation between the MAS frequency reading and the airflow value determined by the ECU. I've thought about doing something like that before, but the problem is the differences between the 1g and 2g ECU.

You can log MAS frequency on the 1g ECU, but not airflow. On the 2g ECU, you can log airflow, but not frequency. The SAFC can display the frequency in a 2g, but you can't log anything and it would be way to dangerous to try and write down the values while you were driving. So there doesn't seem to be any way to do the correlation. Although I'm guessing I could use the raw ECU values from DSMLink and figure something out. That might be a helpful writeup for the 1g guys so they could actually see how much air they are flowing.

Hummm.....I might have to look into that, although it wouldn't work because I have a 2g and the MAS is different. Any 1g guys have a DSMLink log with MAFRaw?

EDIT -> I love these threads where I actually have to think and new ideas get thrown around like popcorn in a microwave.

 

[kenamond]

You are correct, which is what I was trying to say above, but I don't think it came out right. The SAFC does indeed alter the raw MAS frequency which is then converted / interpreted / whatever, by the ECU as an airflow value. But to this day, I have yet to find anything that allows for a correlation between the MAS frequency reading and the airflow value determined by the ECU. I've thought about doing something like that before, but the problem is the differences between the 1g and 2g ECU.

You can log MAS frequency on the 1g ECU, but not airflow. On the 2g ECU, you can log airflow, but not frequency. The SAFC can display the frequency in a 2g, but you can't log anything and it would be way to dangerous to try and write down the values while you were driving. So there doesn't seem to be any way to do the correlation. Although I'm guessing I could use the raw ECU values from DSMLink and figure something out. That might be a helpful writeup for the 1g guys so they could actually see how much air they are flowing.

Hummm.....I might have to look into that, although it wouldn't work because I have a 2g and the MAS is different. Any 1g guys have a DSMLink log with MAFRaw?

EDIT -> I love these threads where I actually have to think and new ideas get thrown around like popcorn in a microwave.

I'm sure someone out there knows the code used by the 2g ECU to convert IAT, barometric pressure and MAS frequency into airflow. But since you can't log barometric pressure on a 2g (pocketlogger can't anyway), you'd have to take a bit of a guess on that one but then be able to back-calculate Hz for the SAFC adjustment. Hell, I'm sure that's been done. Actually, you might only need to be able to convert the frequency to intake air velocity. I'd bet that is what would be linear with airflow. I'm assuming the ECU uses IAT and barometric pressure to compute air density(rho=f(P,T)), then uses frequency to determine velocity. The known dimensions of the chamber containing the velocity sensor give you the cross-sectional area. Air speed and area give you air volume flow rate. Multiply by density and you get air mass flowrate. I'm such a nerd.

 

[CTP]

Good points. Was that on bone stock fuel system (aside from injectors)? I'm curious how close to the edge he is. And do you happen to know how far the ECU will let the trims get before it gives up, throws a code, does something "special"?

What is your idle vaccuum when it finally settles on 14.7? If it's off, that would be an indicator of a post-TB vaccuum leak.

I'd get a logger if I were you.

It wasn't on a stock fuel system. I had my current setup which is a 255HP and AFPR set at 43psi base. I never threw a code.

 

[JeffGST]

I am a DSMlink tuner, and I haven't touched an SAFC in years. In fact, I own the very first "Twist Knob" Apexi AFC probably 9 years back on my 2G.

I know that when LTFT lo fuel trim is at +4%, a +4% increase in 50Hz airflow slider will bring the LTFT lo down close to zero. It directly affects the MAFraw Hz values in that amount.

50Hz * 1.04 = 52Hz

So the ECU will use 52Hz to calculate Injector Duty Cycle instead of 50Hz. I'm sure the AFC works the same way, as that it intercepts the Hz frequency value from the MAF and adjusts it by % and send the new Hz to the ECU.