LTFT on eManage Blue

[shnazzle]

Ah... That'll explain it then. Need to reduce the number of PIDs I'm logging.

And yes I definitely meant 10ms haha.
Time for sleeps methinks.

Reconfigured my dash again. Added 21B2.
So two screens; one for timing and one for fueling.
Timing has misfires, knock timing, timing advance and some general parameters.
Fueling has stft, ltft, o2 readings and fueling status.

[shnazzle]

No progress today.
Constant "Faulty Obd2 Detected" issues, takes ages to connect since a few days ago.
Today's issues could be that its not happy with the new PIDs I put in.
They didn't display anything today once it did connect. That's either because it's  not knocking or it's not working  s:) s:)

[Cap]

Late to the party

Some random thoughts..

Here is a logging I did of my Stock USDM 1ZZ Spyder With MAF Mod on a Flat Ground Pull
The Lower Pink line is the Stock Fueling..  Starts at ~12.5 AFR  and peeks out at 12.8afr..

Something is wrong with the RPM scaling..  as it should be topping out at 7K..  its showing RPM by about half..  Green Line is the Ignition Advance.. Orange is Air Flow.  Blue line is the TPS..

This is with 2ZZ Injectors and the MAF mod.. I was able to 'pull' the fuel in Open Loop, by offsetting the 'Starting Point' of the Fuel Trim Line by setting it at a LTFT -7 by offsetting the MAF air flow Unit.. The Variable Vane Intake I built allowed me to Mess with the MAF Calibration and watch the resulting AFR tracings..  




Here is a log with my Stock ECU running in Closed loop 15.2 AFR..  then half way through the log I switch it back to 14.7 AFR Closed loop Cruse..  I was able to get up to 17 AFR closed loop Fueling..  I did not try the other way..  looking for 13.5..  but it's possible..  Switching is cumbersome.. The ECU is still using the Narrow Band Sensors..  just an Offset Voltage is used to get the required AFR



I experimented with the LC1 Analog Outputs to offset my AFR's but the AFR's were not stable..  they were swinging wildly..  I spent many hours of trial and error trying to get it sorted out..  but I could not..  so I had to use the Voltage Offset on the Standard Narrow Band Sensors


Cap

[shnazzle]

Thanks Cap.
Few questions;
1) Did you offset the voltage on the MAF only (by raising it) or also the o2 sensors themselves? Not sure how you switched the commanded AFR from 15.2 to 14.7 mid-drive
2) What would be the equivalent action in the emanage (change fueling, timing advance, injector size compensation, air flow changes, etc) to the MAF mod?
3) What are the effects of raising the MAF on timing, if any?

With regards to the 2nd question; the 2zz injectors are 310cc/min and stock are 260cc/min right? So 19% bigger than stock
So two options:
- I can pretend I've put in 210cc injectors in the emanage settings (19% smaller). So emanage should over-compensate by 19%, mocking the 2zz injectors. That's assuming that the 2zz injectors are running nowhere near 80% duty cycle. Then adjust the airflow map to mock the raising of the MAF (or using the dials in the front of the EMB). Leaving the timing table blank
- Add 19% fueling across the range via the injection table and then adjust the airflow to mock the raising of the MAF. Leave timing table blank.

Both these "solutions" allow the car to have full control of timing. As is the case with the MAF mod but mock the higher amount of fuel going in as per the MAF mod. I assume that the changes in airflow cause the car to adjust its timing accordingly, assuming you're running a higher octane fuel that burns slower and doesn't detonate at the advanced timing.

This contradicts the idea that any gains from N/A tuning are to be had from advancing timing (and using a higher octane).

I had a look at two other maps available on the forum; one from Alex (northernalex) and one from Phil (evileye_wrx). Both take a very different approach.
Phil's does all his work in fueling by adding massive amounts of fuel (from 10% up to 55%) depending on rpm and throttle. He actually retards timing by -2 above 6k rpm and 70% throttle. With up to 10% airflow adjustment in top RPM/WOT.

Alex on the other has NO fueling changes (guessing he didn't have the harness) and has big timing adjustments. starting at 2 at just above idle to 6deg from about 5000rpm onwards with medium airflow adjustments across the board.

Phil's approach seems quite akin to the MAF mod. Adding lots of fuel and adding a lot of airflow at high load.
Alex's approach is entirely around timing. The timing map advances while it is undone a bit by changes to airflow.

[lamcote]

My understanding is:

1) The voltage offset is on the O2 signal, not MAF. (This is a separate thing from the MAF mod)
2) The simplest eManage equivalent to the MAF mod is to use the airflow adjustment and alter injector sizes as you describe above (but presumably Noble have already done the same thing with the timing on your car, but by using the ignition harness? (If so, you can't do it twice)
3) The MAF riser works by simulating lower than actual airflow mass, thereby causing the timing to be advanced by the ECU, the 2zz injectors and fuel trims then return the fuel flow back to normal 1zz levels. Thus you get the standard 1zz AFR but with advanced timing.

I'd be very interested to see Cap's comments though.

One other thought, if the MAF mod works with 2zz injectors and a LTFT of c.-7-8%, wouldn't it also work with standard 1zz injectors and LTFT of c.+10-11%?? (As long as you drive in closed loop for a while before going into open loop to minimise the risk of running lean.)

[Cap]

Thanks Cap.
Few questions;
1) Did you offset the voltage on the MAF only (by raising it) or also the o2 sensors themselves? Not sure how you switched the commanded AFR from 15.2 to 14.7 mid-drive

I only raised the MAF to produce a lower Engine Load Calculation..  causing MORE ignition advance and LESS total fuel..  Less fuel then the fuel trims can compensate for which is why the Bigger Injectors.

2) What would be the equivalent action in the emanage (change fueling, timing advance, injector size compensation, air flow changes, etc) to the MAF mod?

Likely the same..  but the E-Manage can be tweeked over the range, where the MAF Mod is a Fixed Offset

3) What are the effects of raising the MAF on timing, if any?

Advanced it by forcing the ECU to calculate a lower engine load

With regards to the 2nd question; the 2zz injectors are 310cc/min and stock are 260cc/min right? So 19% bigger than stock
So two options:
- I can pretend I've put in 210cc injectors in the emanage settings (19% smaller). So emanage should over-compensate by 19%, mocking the 2zz injectors. That's assuming that the 2zz injectors are running nowhere near 80% duty cycle. Then adjust the airflow map to mock the raising of the MAF (or using the dials in the front of the EMB). Leaving the timing table blank
- Add 19% fueling across the range via the injection table and then adjust the airflow to mock the raising of the MAF. Leave timing table blank.

Both these "solutions" allow the car to have full control of timing. As is the case with the MAF mod but mock the higher amount of fuel going in as per the MAF mod. I assume that the changes in airflow cause the car to adjust its timing accordingly, assuming you're running a higher octane fuel that burns slower and doesn't detonate at the advanced timing.

This contradicts the idea that any gains from N/A tuning are to be had from advancing timing (and using a higher octane).

I was only able to get performance gains by advancing timing..  my view was that the 1ZZ ECU is dumping plenty of fuel for an NA Machine..  as discussed previously in your threads

I had a look at two other maps available on the forum; one from Alex (northernalex) and one from Phil (evileye_wrx). Both take a very different approach.
Phil's does all his work in fueling by adding massive amounts of fuel (from 10% up to 55%) depending on rpm and throttle. He actually retards timing by -2 above 6k rpm and 70% throttle. With up to 10% airflow adjustment in top RPM/WOT.
Alex on the other has NO fueling changes (guessing he didn't have the harness) and has big timing adjustments. starting at 2 at just above idle to 6deg from about 5000rpm onwards with medium airflow adjustments across the board.

Phil's approach seems quite akin to the MAF mod. Adding lots of fuel and adding a lot of airflow at high load.
Alex's approach is entirely around timing. The timing map advances while it is undone a bit by changes to airflow.

Initially I started playing with Fuel by Installing a wide band (LC1) and using the Output of the Analog Outputs to force the Closed Loop Fuel Ratio to where I wanted it..  besides the fact that the ECU AFR swings went wider than with a Narrow Band Sensor ( Fuel/Air ratio Hunting during closed loop actually is programmed to feed fuel then air to the Cats..  that is why it's there )  but the Hunting swing was causing a lean/rich stumble..  but by messing with the LTFT in closed loop, I was able to force an Open Loop AFR in the range I wanted to test..  I did not notice any real difference on the Butt Dyno..  so I abandon that work and proceeded to attack the ignition advance part of the problem..  That is where I saw the best improvement in Fuel Millage and in performance..

  In testing the AFR theory of Fuel Economy, I did various runs from 14.7 up to 17 and checked the approximate fuel millage..  mostly the same..  nothing remarkable different..  but as the AFR's went up..  the HP produced went down so my Foot was slightly more into the throttle to get the Cruse Speed..  

Thought Change..

By Removing the 'Ground Wire' in the Narrow Band O2 Sensor Wire Harness..  I was able to Inject an Offset Voltage ( Raise the Ground of the O2 Sensors above the Veh Ground Voltage Potential ) in the Narrow Band O2 Sensor Circuit of the ECU. This forced the ECU while in closed loop to Pull Fuel to make the 'Total Voltage' it wants to see as proper..  this same technique can be used to move the Ground Potential BELOW the Veh Ground to cause it to run in Closed Loop Richer than Stoch Mixture..

And you were told that Ground Offsets were Bad!.. No So..

Cap

[Cap]

A little More information to digest on the AFR/Narrow Band O2 Sensor Tweeking..

This is a chart of the Voltage output by a Narrow Band O2 Sensor, and the Approximate AFR that is refers to..
O2 Volts to A/F ratio
volts ............................. A/F
0.1 ............................... 17
0.2 ............................... 16.5
0.3 ............................... 16
0.4 ............................... 15.4
0.5 ............................... 14.9
0.6 ............................... 14.4
0.7 ............................... 13.8
0.8 ............................... 13.2
0.9 ............................... 12.7
0.985 ............................ 12.1

By looking at this system as a feed Back Computer..  the ECU is looking for a Voltage of .52 Volts to be returned by the Upstream O2 Sensor(s) in the Tail Pipe at Closed Loop..  
In Closed Loop, the ECU will Read the Lbs of Air Flow of the MAF, Then Calculate how many Pulses needed to Squirt 14.7 Lbs of Fuel for the measured Air Lbs. ( EDIT..  OOOPs 1 Lb Fuel to 14.7 Lbs Air )
If the Voltage Returned is HIGH then the ECU needs to PULL some of the Fuel Pulses from the Calculation of the Injector System..  ( Fuel Trims go to the Neg Numbers )
If the Voltage Returned is LOW then the ECU needs to ADD some pulses Fuel Pulses to the Calculation of the Injection System.. ( Fuel Trims go to the Pos Numbers )

For practical Numbers, the Fuel trims can only Compensate up to 30% Error Sig before they are Frozen from Moving..  Usually an Error Sig or Greater then 20% ( Plus or Minus ) will toss the Lean or Rich Error Code..

Back to the Chart..  in Closed Loop, the ECU wants to have .52 from its 'Sensor' to get 14.7 AFR..  If I put a little Battery in the O2 Sensor Line and 'Add' 0.1 Volt to the Sig coming to the ECU from the O2 Sensor, the O2 Sensor will only have to be running at .42 Volt ( Plus the Batterys .1 Volt = 0.52) so the the ECU will be happy and 'In Loop'..  Extrapolating 0.42 Volts from the Chart Above gives us about an AFR of 15.2..  Closed Loop 15.2 AFR.

If I take the Same 'Battery' and  wire it 'Backwards' into the O2 Sensor Voltage Circuit, it will 'Subtract' 0.1 Voltage from the O2 Sensor input to the ECU..  The Sensor will be running at .62 Volts.. the Battery removes .1 Volt and the ECU sees .52 Volts..  so the O2 Sensor will have to be running at about 14.1 AFR!..  Closed loop!..

Lets Play with the Numbers..
13.5 AFR is about a Voltage of 0.75 Volts..  That is about 0.23 Volt that needs to be removed from the O2 Sensor Circuit before it gets to the ECU input..  Hey that is about One Foreword Transistor Drop ..  

So by MY Math..  If you put a Signal Diode foreword Biased in the Sig Line going to the ECU you will have a System that will be running at about 13.5 in closed loop all the time..

Wonder if that would work?..  

Cap

[shnazzle]

That AFR/narrowband table is great!
I'll assume your electrical engineering
concepts are correct. Well above my pay grade.

A bit confused about the 14.7 fuel though. Isn't it the other way around? Needs to inject 1 part fuel for every 14.7 parts measured air?

This is starting to seem a bit like a tuning box  in that actual feedback signals are being changed to force the stock ECU to "map" itself instead of applying changing to airflow, fueling and ignition via the emanage and trying to work around the built in tables.

My main concern with the o2 voltage manipulation is; do I always want to run 13.5 AFR during closed loop?
Other than the obvious passing of emissions tests, it seems that cruising around and idling at 13.5 seems a waste of fuel.

Where the "tuning box" would become interesting to me is when it could be coupled with a responsiveness to the throttle and rpm within the closed loop range. Which takes us back to one of the original questions of what are the parameters that determine closed/open loop?

Bringing it back to emanage...
As I can't control the o2 signal, my only raw feedback parameter that I can change is airflow.

But no matter how much I manipulate the airflow, the stock ECU will always respond to the o2 sensor reading off 0.52.
Seems like a dead end to me for the emanage.
The only way this looks to be possible is to bypass the narrowband o2 using a metered power source to 0.52v and then let emanage do its thing in closed and open loop

[lamcote]

Yes shnazzle I agree, if it's possible to bypass the O2 signal and replace it with a mock signal automatically switching between 0.9v to 0.1v and back, telling the ECU it is getting 14.7 AFR, this would provide a solution. The key is being able to get this to kick in at a specific and repeatable point (based on revs and or TPS?). This would allow the mapping to be done at all load cells beyond the point it kicks in and then the fuel trims wouldn't be affected. The two things required are:

1. Define the point at which the signal would kick in
2. Identify all load cells beyond this point which could then have the fuel remapped

[Cap]

Thank you for the Heads up on my mistake..  I was sitting on the wrong side of the equation..  It was late on a Friday nite

A Few more thoughts about the Signals..  

I went back through my notes on the ECU inputs..  and the Switching point for the O2 Sensor input is 0.45 Volts..  The MR2 ECU needs to see the O2 Sensor input Switch Above Below .45 Volt..  It can not be a Stable .45 volt. so it's like a Set/Reset Input triggered at .45 volt to switch from adding fuel to subtracting fuel..  
That does not center at 14.7 AFR..  but that is where my testing showed it to be..

If you run one or two foreword biased Signal Diodes in the O2 Sensor Line to the ECU..  this will cause Richer Mixture when they are in..  

A Plan could be..  Take an Arduno that is controlling a Miniature Relay that Normally shorts out the Diodes, this will cause the Mixture to revert back to 14.7..  Have the Arduno reading the MAF and Calculating the RPM from the Tack Line.. This can give you a Engine Load Calculation.  When the Load is above a certain Calc, have the Arduno Open the relay, forcing the Mixture to go to the 13's

Not the direction the Thread started out..  but it's a plan

Cap

[lamcote]

I've edited my post above to clarify some details.

[shnazzle]

On a tangent;

Today I cleaned out my air filter and reset the ECU. The air filter wasn't really dirty so that can't play a role.

What I can 100% say is that an ECU reset makes way for the eManage and it goes like the clappers! It genuinely made the car very noticeably quicker, particularly in the low/mid range. Also, what's new is that at a certain rev range I can hear a bit of induction noise now. Now as I said, the air filter really wasn't dirty. A hardly noticeable amount of black dust came out when I washed it (it's a K&N). But if I floor it in 2nd or 3rd at >2.5-3k rpm I can hear the distinct noise of intake. It could be that the filter clean helped a bit of flow, but I severely doubt it. Also, I had exhaust burbling. I usually only get that after a hard run out when I've got the exhaust flaming hot.
This must mean that with the ecu reset, the car is sucking harder  s:) s:)  

I purposely toodled out of our estate very slowly as to not trigger any LTFT changes past 2k rpm. When I pulled out in 2nd it was a genuine kick in the back. I coasted back down to 1500rpm in 2nd and put my foot down, wheel spin! Last time it did that was when I first installed the emanage.
I drove up the dual carriage way and it was noticeably quicker in higher revs as well. I didn't have to drop to 4th to get it to go. I put my foot down in 5th at about 3k rpm and it gave me a right kick.

Back off the dual carriageway in a 50 zone I tried the 2nd and 3rd gears again. While now a bit less punchy, it's still absolutely quicker than it was before the reset.

Conclusions (which we already knew but figured it was worth re-iterating haha):
1) ECU reset clears the LTFT tables and the emanage makes the car come alive with intake "roar" and exhaust burbling (de-pre-cat stock mani and TTE exhaust). duh.
2) The LTFT table adjusts very quickly, within 10 mins, but it doesn't adjust fully! There's still a huge gain over the pre-reset condition. This means that it's applying some sort of adjustments across the ranges but doesn't finish adjusting until a good few cycles. I reckon the car will be back to "normal" by the time I get home tomorrow after work.

What I'll do next is pick a "course" that will allow me to do some pulls. I will log those and then reset the ECU and repeat it immediately. Keeping all environmental parameters the same.
It'll be interesting to see the difference overlaid kind of like a dyno chart.

What'd be more interesting is if I could get an actual pre/post dyno chart

[lamcote]

Yes, thanks for that. It demonstrates that there is a lot to gain on a largely standard engine but that the standard ECU is very adept at undoing the remap if it sees any changes in closed loop.

The only way to get this right is either to identify all the load cells in the eManage's map that fall within the standard ECU's open loop parameters, presumably on a dyno, or force open loop by some method such as the one I mentioned above, if this works?

[Essex2Visuvesi]

A little More information to digest on the AFR/Narrow Band O2 Sensor Tweeking..

This is a chart of the Voltage output by a Narrow Band O2 Sensor, and the Approximate AFR that is refers to..
O2 Volts to A/F ratio
volts ............................. A/F
0.1 ............................... 17
0.2 ............................... 16.5
0.3 ............................... 16
0.4 ............................... 15.4
0.5 ............................... 14.9
0.6 ............................... 14.4
0.7 ............................... 13.8
0.8 ............................... 13.2
0.9 ............................... 12.7
0.985 ............................ 12.1

By looking at this system as a feed Back Computer..  the ECU is looking for a Voltage of .52 Volts to be returned by the Upstream O2 Sensor(s) in the Tail Pipe at Closed Loop..  
In Closed Loop, the ECU will Read the Lbs of Air Flow of the MAF, Then Calculate how many Pulses needed to Squirt 14.7 Lbs of Fuel for the measured Air Lbs. ( EDIT..  OOOPs 1 Lb Fuel to 14.7 Lbs Air )
If the Voltage Returned is HIGH then the ECU needs to PULL some of the Fuel Pulses from the Calculation of the Injector System..  ( Fuel Trims go to the Neg Numbers )
If the Voltage Returned is LOW then the ECU needs to ADD some pulses Fuel Pulses to the Calculation of the Injection System.. ( Fuel Trims go to the Pos Numbers )

For practical Numbers, the Fuel trims can only Compensate up to 30% Error Sig before they are Frozen from Moving..  Usually an Error Sig or Greater then 20% ( Plus or Minus ) will toss the Lean or Rich Error Code..

Back to the Chart..  in Closed Loop, the ECU wants to have .52 from its 'Sensor' to get 14.7 AFR..  If I put a little Battery in the O2 Sensor Line and 'Add' 0.1 Volt to the Sig coming to the ECU from the O2 Sensor, the O2 Sensor will only have to be running at .42 Volt ( Plus the Batterys .1 Volt = 0.52) so the the ECU will be happy and 'In Loop'..  Extrapolating 0.42 Volts from the Chart Above gives us about an AFR of 15.2..  Closed Loop 15.2 AFR.

If I take the Same 'Battery' and  wire it 'Backwards' into the O2 Sensor Voltage Circuit, it will 'Subtract' 0.1 Voltage from the O2 Sensor input to the ECU..  The Sensor will be running at .62 Volts.. the Battery removes .1 Volt and the ECU sees .52 Volts..  so the O2 Sensor will have to be running at about 14.1 AFR!..  Closed loop!..

Lets Play with the Numbers..
13.5 AFR is about a Voltage of 0.75 Volts..  That is about 0.23 Volt that needs to be removed from the O2 Sensor Circuit before it gets to the ECU input..  Hey that is about One Foreword Transistor Drop ..  

So by MY Math..  If you put a Signal Diode foreword Biased in the Sig Line going to the ECU you will have a System that will be running at about 13.5 in closed loop all the time..

Wonder if that would work?..  

Cap

Seriously tho this thread is an enlightening read, will be referring to this in the spring when I start my turbo conversion

[shnazzle]

Two new logs from just now uploaded to the share. In zip.
Two logs because my obd2 keeps conking out for some reason. Very annoying.

LTFT stayed at 0 until about 3/4 through the drive when all of a sudden they took a nosedive.
I also noticed a misfire or knock cut when I was flat out in 2nd or 3rd I forgot coming from a traffic light onto a 70 road.
Again shows that timing is quite aggressive when you reset ecu. But it pulls it back.

[lamcote]

Doesn't the knock show that your eManage was mapped for 97 octane and the standard ECU is unable to override timing other than using the knock sensor?

[shnazzle]

Well I'm running Vpower nitro so you'd think that if the advance was on on 97 then it would definitely be OK on 98.

It might not have been knock. But I had my foot flat and there was a little bump in power for a fraction of a second

[lamcote]

OK. I thought I remembered you saying you ran 95 octane, I must have got confused.

[shnazzle]

Taking the discussion from the TPS trickery back to here;

So I'm going to apply the TPS trick which will allow me to send the car WOT at pretty much anything over 50% throttle and 2500 revs or so. I'll tweak that based on where the most work is in the maps I have.
Fact is that even at WOT, the car isn't always in open loop. So the solution isn't perfect, but I reckon it will cover a fair bit of the stock closed loop areas.

The way to tell I suppose will be if LTFT remains around 0 for a longer period of time across rev ranges than before the trickery.

I'm also going to install a Fast and the Furious style button to activate the trickery  s:) s:)
What this will allow me to do is some seat-of-pants testing by taking it to say 70% throttle, holding it there and flipping the switch, then it SHOULD do something. At the very least I should feel a stutter and I should see my TorquePro fuel state go to Open Loop.

[jvanzyl]

I look forward to the video of this!... please be sure to change gear at least 7 times whilst accelerating prior to pushing the button! (just like in the film!).

I will also need to plagiarize some of your work to carry across to mine!

[shnazzle]

I look forward to the video of this!... please be sure to change gear at least 7 times whilst accelerating prior to pushing the button! (just like in the film!).

I will also need to plagiarize some of your work to carry across to mine!

That's the idea! But you've already got all the gubbins set up  s:) s:)
Some detailed pics of the box of tricks would be great in my email  s:) s:)

[shnazzle]

So I guess this thread has come to an end with the conclusion from the open loop trickery thread.

LTFT remain stable when forcing the ecu into open loop via the box of tricks.  End result being that you're always driving on the emanage map when it counts. No meddling from the ecu.

Guess my conclusion is; emanage blue piggyback (and possibly many others) is utterly utterly pointless on N/A setup without this open loop trickery.
You can work around it, but really... No...

[shnazzle]

Another update; not the end  s:) s:)

Did a few logs post-open loop mod. Although I forgot to log TPS, it is obvious when I;m on throttle through the MAF values.

It has been a few days since the mod and the ECU has caught onto my trickery a bit. All this means is that I need to play with the VTEC areas a bit more. But it'll never be perfect.
It hasn't caught on fully though. Once i'm over 3000rpm and on WOT the LTFT flies up from around -12/-14 to -4/-5.  

So that's a massive improvement of almost 10% more fuel vs what it was doing before. No wonder it is massively noticeable.

You spend a fair amount of time around 3k with about 50-75% throttle, which in my map doesn;t kick in the tweak yet. So I'll change that and reset ECU. Should improve things

[shnazzle]

Tweaked map a bit. Bit better as far as I can tell, but obd2 broken so no LTFT readings.

Did install a little switch to turn the trickery on/off. Lights up when active. Quite fun  s:) s:)  
When it's MOT time I'll press the switch to turn it off then drive around a bit so that the car can make the emissions nice and pretty. Then press it again when it's done.



 m https://youtu.be/WVuhl56upqE m

[jvanzyl]

I'd like to place an order for an installation of the box of trickery with aforementioned "GoGo" button!!! @ ding day if possible???