SR20 running rough

[CodySoFine]

Engine is a freshly rebuilt s13 SR20DET with AEM trigger disc, DW740cc injectors, GT2871r, CP pistons(8.5:1 comp), Eagle rods and a G4x

 

I've been trying to get the car running over the last few days and I'm having some trouble. I've gotten most of the setup correct(I think) with the help of the tech support guys and some forum posts but I cannot figure out how to get it running smoothly. It will start and idle correctly but its extremely rich 10:1 blowing black smoke. I changed the oil and put new plugs in today and wired in the AEM wideband to the stock O2 sensor wires and I believe I've set that correctly and now my wideband is reading extremely lean and still blowing black smoke. The car seems to run rougher now than it did previously without the wideband wired in as it is breaking up on revs. 

 

I tried changing the master fuel MS but it doesn't have any change on my AFR readings and just makes the car idle higher so I've left it set where I was instructed by technical support.

 

I've attached my current tune file and a datalog of me starting the car and revving it a few times. Does anyone have any insight on what I'm missing here? Thanks!

SR20 vol2 AEM.pclx PC Datalog - 2024-03-10 6_29_57 pm.llgx

[Adamw]

The MAP sensor trend doesnt look right to me, it idles at 100kpa and decreases as you open the throttle.  Is the MAP sensor connected to a port in the plenium?

If the map sensor is connected to the plenum then the low vacuum and very lean lambda measurement would suggest the cam timing is out or an intake valve isn't fully closing.  

[CodySoFine]

I believe that's my issue. The source I had for the MAP sensor only saw vacuum when the throttle body was open, I wasn't aware of such a thing but I changed the source to a port on the intake manifold and the car ran for a few seconds at idle and read 35kpa. I think I ran out of gas so I'll have to fill it up tomorrow and try and run it again.

I'll update tomorrow after work with a new log hopefully 

[CodySoFine]

So that was definitely my issue I believe. I pulled vacuum from a port directly on the manifold vs the throttle body and it seems to reading what I would expect it to.

I let it idle for a bit(extremely high not sure how to fix that) but it runs rough, doesnt respond to throttle, and if I let it idle without wacking the throttle it will progressively lean itself out until it stops running all together.

I've attached the log below any insights would be appreciated. 

PC Datalog - 2024-03-11 6;40;17 pm.llgx

[Adamw]

Im not sure if the lambda is reflecting reality in that log.  Assuming it is an AEM gauge, are you seeing similar lambda displayed on the gauge as shown in PC Link?

In the area it goes lean the injector PW doesnt change and neither does any of the metrics that represent air flow so it doesnt appear to be caused by the ECU.  Perhaps fuel pressure or something.  

Assuming the lambda is correct then it appears it needs less fuel for most of the log.

The high idle is probably due to the idle base position being too high.  Possibly the throttle is too far open if reducing the base position cant bring the idle down to an acceptable level. 

[CodySoFine]

Firstly I appreciate taking the time to review my logs and offer advice so thank you Adam my hat is off to you sir.

My AFR gauge is reading exactly the same as what the lambda on the ECU is displaying but I question the gauge because I'm getting black smoke out of the exhaust and black plugs when the gauge is reading 13-15AFR.

My only thoughts at this point are the sensor might be faulty but I'm not sure if that's even possible for them to go bad in the way that I'm seeing or my exhaust might have some type of leak somewhere that is allowing air into the exhaust thus causing the sensor to read leaner than it should be but I'm also not sure if that is possible either. I'll check tomorrow and report back. 

[Adamw]

2 hours ago, CodySoFine said:

My AFR gauge is reading exactly the same as what the lambda on the ECU is displaying but I question the gauge because I'm getting black smoke out of the exhaust and black plugs when the gauge is reading 13-15AFR.

The AFR is showing more like 7-12 in the log.  

[CodySoFine]

Okay so since I last posted I've been going crazy trying to figure this issue out. I've fixed some wiring, bought a new O2 sensor for the wideband(seems to read much better now), new spark plugs(again), and verified I have no vacuum leaks.

So onto my current issues. The car will start and idle at 2000rpm, lambda is all over the place and not consistent at all my AEM wideband shows 12~ at idle most commonly but its not a steady 12 and is all over the place, it breaks up when giving it throttle I believe this is related to the lambda issue because it spikes lean before it breaks up and then revs freely, and overall tune seems not happy.

I've pulled the IACV off and applied voltage manually to the pins to ensure that the plunger is actuating correctly and it is, I've smoked tested my intake and have no vacuum leaks, TPS is adjusted, throttle blade is closing all the way, I've talked to some of the guys on the tech support (top notch dudes) and they recommended that I set "Idle ignition control" to off and change to open loop and change the numbers in "idle base position" and it should change the idle but I wasn't able to discern any difference in the idle when changing the numbers so I reverted them back to the default values.

I've attached the current tune I'm using and a log I just took of the car idling and revving.

PC Datalog - 2024-03-16 8_02_26 pm.llgx current.pclx

[Adamw]

I wouldnt worry about the idle control at this point, get it running correctly with stable AFR first.  Have you confirmed fuel pressure?  Have you set base timing?  What does it have for ign coils & Ignitors?

[CodySoFine]

Fuel pressure is set around 30-35psi. I was told by support that there's no way to account for fuel pressure on traditional injection and that it doesn't matter what I set it to so just leave it where it is. Stock fuel pressure is around there so I've left it alone for now. I've set base timing and it has 15 degrees of timing during cranking but I have no way of checking timing while the car is running but even so the idle is so high I don't think it would be accurate anyway. It has P2M coil packs and the stock ignitor in a wiring specialties harness. 

I'm confused about why the AFRs seem to bounce around so much. I've got the wideband hooked into the ECU on the proper pin and set it in the ECU per tech support. Also I'm not sure if this is evident from the log but when its idling it will progressively start getting leaner and dropping idle until it cuts off if I don't give it throttle every few seconds. 

For what it's worth I've had this car idling correctly on an AEM ecu last year and it idled correctly with stable AFRs and I feel as though I've ruled out most mechanical issues I'm assuming I've calibrated something wrong within the ECU. 
 

I am going to get the car dyno tuned next month by a professional but I'm just attempting to make sure the car is mechanically sound.

[Adamw]

You need to confirm fuel pressure is correct and constant.  Your MAP and Injector PW is reasonably stable when your AFR is erratic, so the variation is not coming from anything the ecu is commanding, fuel pressure would be the main external variable that can affect AFR so this needs to be confirmed first. 

 

7 hours ago, CodySoFine said:

I've set base timing and it has 15 degrees of timing during cranking but I have no way of checking timing while the car is running but even so the idle is so high I don't think it would be accurate anyway.

  How did you check base timing?  

[CodySoFine]

What should fuel pressure be set at? It’s currently somewhere around 30-35. Stock for SR20s is 36 so I got it close to that as tech support told me fuel pressure is not important as long as it’s consistent. 
 

I checked timing by taking apart one of the coils and making a jumper to the spark plug and using a timing light on that. I checked it while cranking the car and got it to 15 on the timing marks. 

[dx4picco]

5 hours ago, CodySoFine said:

 I got it close to that as tech support told me fuel pressure is not important as long as it’s consistent. 

do you have any proof that it is consistent?

[jdniss]

7 hours ago, CodySoFine said:

What should fuel pressure be set at? It’s currently somewhere around 30-35. Stock for SR20s is 36 so I got it close to that as tech support told me fuel pressure is not important as long as it’s consistent. 
 

S15 SR20DET service manual states the following, in blue:

 

I've had a non-X Series AEM UEGO Wideband also wired into AN V1 and the reported PCLink AFR numbers were fairly far off the gauge values using AEM's recommended Lambda calibration numbers - even rewiring the ground wire directly to the ECU - until I re-calibrated the voltage numbers in PCLink.

The AEM UEGO X-Series RS232 Serial reported values in PCLink were even further off compared to the gauge, based on the existing configuration available  - to the point I jump straight to CANBus based comms with that gauge instead - not sure whether the AEM calibrations are just inconsistent..

 

I'm in a similar situation to you - poorly running, rich running SR20 despite various troubleshooting;
I can literally plug the old PowerFC back in with no changes and the car runs perfect as it always did, so I must have a timing or Software config issue somewhere in the Link I've yet to determine - I'll be following this topic with interest.

[CodySoFine]

6 hours ago, dx4picco said:

do you have any proof that it is consistent?

Yes. My regulator has a gauge on it and I can visually see the pressure is not fluctuating.

 

@jdniss

I had some issues with that as well but I bought a new O2 sensor for my wideband and my numbers seem to be much more consistent. It's off by maybe .2-.5 lambda but certainly nothing I'm actively concerned about or attributing to the cause of my issues.

One thing I do find suspicious is the timing table. I've watched videos of other people setting up links on their SR20s and in the trigger timing table their offset is -40 to -100 to get the timing marks in the correct spot but I had to give my car +20 degrees to get it on the right mark. This is a brand new motor and I've verified my cams are timing correctly about 6 times in addition to the machine shop verifying it so I know my motor is timed correctly but I'm not sure why my offset is so wildly different to other peoples. This could have something to do with it but I'm not sure. I'll recheck timing for the 7th time and report back if I figure anything out.

[CodySoFine]

Okay I’ve verified my cams are timed correctly and reset my fuel pressure. It reads 43PSI vacuum disconnected. 
 

I rechecked my timing while running and it only needed +5 degrees offset to get to the correct timing mark instead of 20. I’m not sure if there was a miscommunication between myself and my father the first time we checked timing or if cranking VS running is significantly different but I’ve corrected it.
 

I double checked my AFR gauge and it’s actually significantly off compared to what I thought. The gauge is showing 12:1 AFR which I believe translates to .85 lambda but the ECU is actually reading .67~ Which set of numbers should I be trusting? The wideband gauge or the lambda showing in the ECU?

I plan to upgrade to a link wideband at some point but unfortunately it’s outside of my budget on such short notice before my tuning session. Any ideas on what I should do?

[koracing]

2 hours ago, CodySoFine said:

One thing I do find suspicious is the timing table. I've watched videos of other people setting up links on their SR20s and in the trigger timing table their offset is -40 to -100 to get the timing marks in the correct spot but I had to give my car +20 degrees to get it on the right mark. This is a brand new motor and I've verified my cams are timing correctly about 6 times in addition to the machine shop verifying it so I know my motor is timed correctly but I'm not sure why my offset is so wildly different to other peoples. This could have something to do with it but I'm not sure. I'll recheck timing for the 7th time and report back if I figure anything out.

Yeah your offset doesn't match other SR20 engines I've tuned (formerly on power FC as well).  Something seems off.  The last one I tuned was an offset angle of -83.  

6 hours ago, jdniss said:

The AEM UEGO X-Series RS232 Serial reported values in PCLink were even further off compared to the gauge, based on the existing configuration available  - to the point I jump straight to CANBus based comms with that gauge instead - not sure whether the AEM calibrations are just inconsistent..

The X-Series should be sending data to the Link over Analog signal (white wire) and ground reference (brown wire) both connected to the ECU, or via the CAN (green black stripe and white black stripe) wires to a CAN channel.  It should not be using the blue RS232 serial ouptput to the ecu.

How are you verifying timing now?  Are you using a plug wire from the end of the coil to the spark plug and putting your timing light inductive clamp on that plug wire, or are you putting the timing light inductive clamp on the coil trigger wires?

[CodySoFine]

Here are my cams just to verify that my engine is timed correctly. I just made a jumper with a spark plug wire between the coil pack and spark plug for cylinder 1. I got it to the correct timing mark with only +5 degrees offset. I’m not sure what the difference could possibly be. I can try a large - number and see if that changes anything. 

Edit:

I tried -80 and it will not run at all. My car seems to run best at +7 offset but doesn't necessarily run amazing, it doesn't respond to sharp throttle but if I gradually throttle up it will rev all the way. Idle is still very high(2000rpm) and AFRs are inconsistent. 
 

As for the wideband I’ve connected the white wire from the gauge to pin 19 on the ECU. I haven’t connected any grounds from wideband to the ECU. What pin would you use for that?

Disregard all of that. I'm not using an X series wideband so it doesn't have a reference ground that would explain the lambda difference. I've just disabled the lambda reading until I can get the cal file setup correctly.

[koracing]

I suppose the difference in trigger could be explained by the different trigger disc you're running versus the OEM 360 tooth oem CAS.  Timing looks correct as far as I can tell unless the crank pulley is somehow off (running a long screwdriver down the #1 plug hole would sort of verify it appears to reach peak height at the same time as the TDC mark).    If it's running pretty well or as good as it can at the trigger settings you have, then it may just be a tuning issue.  Hard to say over the internet, but I would try some big swings on your ignition timing and fueling in the ecu to see if it's happier one way or the other to see if the tune is just way out to lunch.  It's sometimes possible that you're so rich that you get misfire which would register lean on a wideband and could be causing false AFR readings as well. Engines will run on a pretty wide range of AFR so for testing purposes you can get pretty far out before the engine won't run any more.  I am assuming tech support had you go through and click test all injectors and coils already.

If you need help doing this remotely, I would be available most evenings for a brief session to try somethings.  (Pacific standard time zone)

[CodySoFine]

2 hours ago, koracing said:

If you need help doing this remotely, I would be available most evenings for a brief session to try somethings.  (Pacific standard time zone)

I'm gonna shoot you a PM!

[jdniss]

7 hours ago, CodySoFine said:

@CodySoFine looks like the crank timing is off by 15° ?

PLMS state 5th mark from the left

https://www.plmsdevelopments.com/index.php/sr20-stuff/sr20-setup-tips

[CodySoFine]

@jdniss

Yes in my picture the crank is marked at TDC or 0 degrees. You use that mark to time the top and bottom ends together. When timing the engine with a timing light the light illuminates the second hash mark from the right so 15 degrees.

[jdniss]

8 hours ago, koracing said:

The X-Series should be sending data to the Link over Analog signal (white wire) and ground reference (brown wire) both connected to the ECU, or via the CAN (green black stripe and white black stripe) wires to a CAN channel.  It should not be using the blue RS232 serial ouptput to the ecu.

 

Yeah sorry, white 5V output - I confused the colours
 

8 hours ago, koracing said:

Yeah your offset doesn't match other SR20 engines I've tuned (formerly on power FC as well).  Something seems off.  The last one I tuned was an offset angle of -83. 

In the the old car, the G4+ trigger offset on my SR20 was -100 with a Falling trigger,
on the same car the G4X trigger offset with the same engine was -85 with a Falling trigger - took me a while to work why the car was so "lazy" after the G4X upgrade

[koracing]

Got online with @CodySoFine last night and got it to idle fairly stable.  His IACV seems to bypass a bit more in the fully closed position than excpected, but ultimately turning on the closed loop idle, changing the base position table down quite a bit, and then enabling idle ignition control was able to get the idle pretty stable and the AFR to act pretty stable around idle.

[jdniss]

On 3/18/2024 at 9:01 PM, jdniss said:

I'm in a similar situation to you - poorly running, rich running SR20 despite various troubleshooting;
I can literally plug the old PowerFC back in with no changes and the car runs perfect as it always did, so I must have a timing or Software config issue somewhere in the Link I've yet to determine - I'll be following this topic with interest.

Just to cover this off - my issue is now fixed - re-running the ECCS Sync and dialed timing offset again on the G4+ got the car back to a better running state;

I then replaced both the fuel filter and fuel pressure regulator - was only getting 25psi at the rail - regulator looked to be blocked with old E85 "gummy" contaminants that must not have burnt off during short cold starts.