Honda / Acura K20a K24a Engine Forum banner

Running lean under boost after hard driving

306 Views 12 Replies 3 Participants Last post by  Lotus
Im having an issue with my rsx, k20a3 stock motor EMUSA turbo kit 11-13psi walbro 255 return setup. When I first start it up and drive it around it does really good, then out of nowhere after about 30 minutes of hard driving, it leans out under boost. im going to switch back to the factory pump tomorrow and see if that changes anything. any suggestions are appreciated attatched are a datalog of when it runs good and one when it starts to lean out and hits lean cut
1 - 13 of 13 Posts
2
Im having an issue with my rsx, k20a3 stock motor EMUSA turbo kit 11-13psi walbro 255 return setup. When I first start it up and drive it around it does really good, then out of nowhere after about 30 minutes of hard driving, it leans out under boost. im going to switch back to the factory pump tomorrow and see if that changes anything. any suggestions are appreciated attatched are a datalog of when it runs good and one when it starts to lean out and hits lean cut
Also running ID1000's that are a few months old, for some reason the datalogs wont attach so here are some screenshots of whats going on

Attachments

See less See more
Welcome to K20a.org @cbrsx.

for some reason the datalogs wont attach so here are some screenshots of whats going on
Please link them via cloud. From what I can see the operation points are different in load. I miss a lot of information, stated in the calibration and essential parameters like ECT. Could you please serve both, calibration and log.

What is not so fine (left picture) boost, INJ duration and lambda loose connection. That could be from a fuel leak, fuel pressure loss or similar. That is something really need to be investigated.
Welcome to K20a.org @cbrsx.

Please link them via cloud. From what I can see the operation points are different in load. I miss a lot of information, stated in the calibration and essential parameters like ECT. Could you please serve both, calibration and log.

What is not so fine (left picture) boost, INJ duration and lambda loose connection. That could be from a fuel leak, fuel pressure loss or similar. That is something really need to be investigated.
Welcome to K20a.org @cbrsx.

Please link them via cloud. From what I can see the operation points are different in load. I miss a lot of information, stated in the calibration and essential parameters like ECT. Could you please serve both, calibration and log.

What is not so fine (left picture) boost, INJ duration and lambda loose connection. That could be from a fuel leak, fuel pressure loss or similar. That is something really need to be investigated.
I attached the links below, Im checking all connections and fittings as i type this. It so strange that it only occurs after doing pulls for a while, then when i shut the car off and give it time to cool down it acts normal again. I have an AEM wideband attached and it reads very similar to what the OEM o2 sensor is reading, so it is definitely an issue with fuel. I am tuning the car myself so it could very well be an issue with the tune, if thats the case i will have to take it to someone more experienced. Thank you for the reply!

I am tuning the car myself so it could very well be an issue with the tune, if thats the case i will have to take it to someone more experienced. Thank you for the reply!
I assume you are still in the process of tuning, otherwise I would need to state that tune has a numerous of issues which would need some attention not to scrap the engine finally, but the fuel lacking issue is definitely no tuning issue. The engine speed is almost same, boost is almost same, injector duration is a bit increasing, but lambda makes a jump.
  • may first impression was the fuel pressure is gone lost somewhere
  • seeing the calibration an log in detail another option pops up, a too hot exhaust gas, which risks maybe the turbine and lambda probe.
Hence we need to talk about wideband probe location, just to check this option of failure. Could you please clarify for me where and how it is installed?

Regarding the calibration status and setup.
  • You know what the K20A3 is doing regarding valve actuation during low speed cam?
  • You know what is happening to the oil flow during low speed cam?
  • Your tuned lambda strategy is risky
    • for the liner and oil management in WOT and
    • the stock pistons on part load
  • Do you run a thicker oil grade?
  • How do you control ignition without a MAT sensor (manifold air temperature)
  • Looks like you run this setup without boost controller or at least the evtl. mechanical existing one is not willing to open. Is that correct?
  • IGT is quite out of control, ECU is retarding IGT up to 15°, can't see from which correction table it is coming, not from ECT and IAT as well as knock control. Do you have any other system sending a parameter to ECU, forcing the IGT correction?
See less See more
I assume you are still in the process of tuning, otherwise I would need to state that tune has a numerous of issues which would need some attention not to scrap the engine finally, but the fuel lacking issue is definitely no tuning issue. The engine speed is almost same, boost is almost same, injector duration is a bit increasing, but lambda makes a jump.
  • may first impression was the fuel pressure is gone lost somewhere
  • seeing the calibration an log in detail another option pops up, a too hot exhaust gas, which risks maybe the turbine and lambda probe.
Hence we need to talk about wideband probe location, just to check this option of failure. Could you please clarify for me where and how it is installed?

Regarding the calibration status and setup.
  • You know what the K20A3 is doing regarding valve actuation during low speed cam?
  • You know what is happening to the oil flow during low speed cam?
  • Your tuned lambda strategy is risky
    • for the liner and oil management in WOT and
    • the stock pistons on part load
  • Do you run a thicker oil grade?
  • How do you control ignition without a MAT sensor (manifold air temperature)
  • Looks like you run this setup without boost controller or at least the evtl. mechanical existing one is not willing to open. Is that correct?
  • IGT is quite out of control, ECU is retarding IGT up to 15°, can't see from which correction table it is coming, not from ECT and IAT as well as knock control. Do you have any other system sending a parameter to ECU, forcing the IGT correction?
I am still in the tuning process, i have not set up the highcam at all yet and i also havent touched timing since i have no clue on how i would even go about that (i am a noob) . So far i have just been touching the fuel tables trying to get it where it needs to be. During part throttle it is running lean but since i have been having the issues with leaning under boost i havent been touching the tables at all until i get everything settled there. In the picture i attached (not the setup i have, it is just a picture for reference) the red arrow is where i have the factory o2 sensor located and the black is where the AEM wideband is located.

  • I run factory oil weight 5w-30
  • I thought for the IAT sensor was the only one needed for a setup like this. Would it be a good idea to get one and run it to kpro for datalog purposes?
  • I dont have a boost controller i run off of wastegate spring pressure and i have one of those ebay tee boost controllers set to fully closed atm
  • I dont have any other system (that i know about) that would be affecting ignition timing correction. It is possible the timing tables are off since i used a supercharged basemap to create the current map i am running.

High EGT seems like a possibility, i can switch my AEM wideband out for an EGT probe if it is necessary. If timing is the issues causing high exhaust temperatures i will have to take the car to a tuner to get everything set.

Attachments

See less See more
Where is your IAT sensor located? Is it located in the intake manifold, e.g. in one of the runners or the plenum chamber or is it located in the pipe from the air filter to the turbo charger inlet as per OEM setup? If the latter, you need to change that as there is no fixed relation of IAT in the intake to the IAT into the engine with a turbo charger. Such an arrangement only works for a non-charge cooled positive displacement supercharger. All others need to measure the actual IAT into the engine.
Before TB :rolleyes:. That engine has no MAT sensor, only IAT.

Actually the Honda ECU adapt the fuel mass according the fuel calibration only by two tables with respect to density: ECT and IAT over a fuel factor. If that IAT pin at ECU would be used for a MAT (manifold air temperature) sensor, then the correction table would reflect changes in vehicle speed or intercooler heat fluxes related changes to density. That approach would give the former IAT sensor a much higher value for the engine control, of which would not do justice to this. The stock IAT sensor is way to slow for a speed density like approach in a boosted concept.

The stock '02-'04 ECU is very limited, even for NA engines. Boosted engines can be done with it, but a quality calibration on highly transient boost looks differently. From my experience the '05-'06 ECU's are better on the knock system side, has a faster calculation capacity to cover this, but the distance to a real speed density system is still Amazonas river width like.
See less See more
Before TB :rolleyes:. That engine has no MAT sensor, only IAT.

Actually the Honda ECU adapt the fuel mass according the fuel calibration only by two tables with respect to density: ECT and IAT over a fuel factor. If that IAT pin at ECU would be used for a MAT (manifold air temperature) sensor, then the correction table would reflect changes in vehicle speed or intercooler heat fluxes related changes to density. That approach would give the former IAT sensor a much higher value for the engine control, of which would not do justice to this. The stock IAT sensor is way to slow for a speed density like approach in a boosted concept.

The stock '02-'04 ECU is very limited, even for NA engines. Boosted engines can be done with it, but a quality calibration on highly transient boost looks differently. From my experience the '05-'06 ECU's are better on the knock system side, has a faster calculation capacity to cover this, but the distance to a real speed density system is still Amazonas river width like.
I checked over the fuel system and it seems there's a leak between the fuel sending unit and bulkhead coming through it. I wouldnt think this would cause the leaning issue since the hose is attached to the bulkhead fitting itself, is it a possibilty that this could be causing my issue?
I checked over the fuel system and it seems there's a leak between the fuel sending unit and bulkhead coming through it. I wouldnt think this would cause the leaning issue since the hose is attached to the bulkhead fitting itself, is it a possibilty that this could be causing my issue?
Seal it, test it, know it ;).
Seal it, test it, know it ;).
Sealed it up and it made no difference, however I managed to find the problem finally! Walbro was on its way out (I knew it wasn’t suppose to be that loud). I’m looking into moving my IAT to the manifold itself but I can’t seem to find any info on that, since temps don’t move at all when in boost, is that normal? I currently have it located before the TB.
...I knew it wasn’t suppose to be that loud...
These fail quite often and are a mess regarding efficiency and performance. I only made negative experience with the Walbro stuff.

I’m looking into moving my IAT to the manifold itself but I can’t seem to find any info on that, since temps don’t move at all when in boost, is that normal?
Not necessary. The failure happen by not having a correlation of fuel mass and density of air is quite low once you run a lambda of 0.77-0.80 and less then 10° BTDC IGT o_O. The swirl motion a K20A3 in non-VTEC has a strong impact on fuel efficiency, I would aim for 0.82 in lambda and at least for 17° BTDC of IGT. The process of that is to increase lambda at your actual IGT, creating a Pareto Front, consisting of IGT and lambda. With that line you have the max. advance to any rational lambda.

Your first main parameter should be VTC, tune for every VTC and find out the best VE (or torque), your actual tune shows either the turbo is really a bad match for non-VTEC or it is not tuned yet. My recommendation for that is: low speed cam VTC's, high speed VTC's, overlay them and setup VTEC right.

Keep in mind, IGT can't be the last with a turbocharged engine, it has always to be done inclusively, as the enthalpy at turbine is quite different running 20° or 10° of IGT. That means, you tune each VTC with lambda an IGT to a final stage to be able to compare them correctly.
See less See more
2
If the IAT sensor, and you need a quicker reacting sensor than the encapsulated sensor, is fitted in the intake manifold, temperatures will vary a lot.
This is what I did to my CT-E SC intake manifold.
The sensor tip was modified before final fitment. I removed the plastic exposing the sensor element. This increases its reaction time down a lot.



Similar to what is seen here as posted in an other forum:



Or get a GM fast acting sensor and adapt the sensor calibration in the ECU.
See less See more
  • Like
Reactions: 1
1 - 13 of 13 Posts
Top