my tuning plan
my tuning plan
I am planning to install and tune a BP turbo this week. Car is a stock 2004.5 G35 6MT with 57K miles. I have never installed a FI setup before nor have I tuned an ECU before. Car is my DD. I realize there are several indicators of a disasterous outcome in what I just said. Nevertheless here is my plan.
1. Install fuel pump. Check for leaks, dtc, etc. Done. This was a sphincter tightening experience. Day was cold and dry and i was deathly afraid of a static discharge bringing an abrupt end to my install. Having a electric pump and exposed wiring swimming in fuel seems like such a bad idea but I'm no engineer.
2. Install injectors and spark plugs. Then will set K mult, latency and min pulse width as described by uprev. Then log city, hwy and wot runs to make sure things look ok. Done...sort of. Installed NGK plugs 1 range cooler. Came gapped to 1.025mm. Service manual says 1.1mm. Left them at 1.025mm. Installed Bosch 0280158117 550cc injectors. According to http://www.boschdealer.com/specsheets/0280158117cs.jpg set minimum pulse width to 0.72ms and set latency such that 14v came out to 0.960 ms and 8V came out to 1.757 ms. Started with a K of 16555 by mult stock K value of 27592 by 330 (stock inj size) and dividing by 550 (new inj size). Cleared learned values and started the car. Required some throttle initially to stay running but then idled fine. Did a cylinder power test to make sure I was getting spark and fuel to each cylinder. Then increased K to 19000 as it seemed the measured AFRs where a little leaner (15.5 to 16.5) than the target AFR (around 14.7). This helped but some more tweaking will be needed as the car is a bit sluggish. However, need to get on with installing the turbo. Will write some scripts to help home these numbers in later but for now the car cold starts, warm starts, idles and drives.
3. Install turbo. Check for intake, exhaust and oil leaks. Done. Started with a base tune by setting target AFR to 11.69 for all cells where RPM > 2400 and BFS > 6.07. Set fuel compensation to 100%. Set high det table to 45 deg in all cells. Subtracted 1 deg from timing map at cell (2.43, 1200) and 7 deg at cell (19.44, 6400) and then subtracted a graduated amount of timing for all the cells inbetween. I am running a 7psi WG spring. I then flashed the ECU and used RTT to set the K value to 24000 so that my AFR correction ranged between 95% and 105% at idle and free rev to 3k. I saved my ROM and then took the car out for a test drive. I stayed out of boost and logged about 20 minutes of data. Car seemed to run great except occasionally died if I revved the engine and immediately closed the throttle. Will need to work on that. Uploaded the data into djamps MAF tool (thank you BTW for that) to see how far off my MAF table might be. Was suprised to find that it was spot on. Correction was within +/- 5% for the cells that had data and most of the time there was no correction at all. Collected some more data at 50-75% throttle under boost keeping RPM < 5k with the same MAF results. No check engine lights were turned on either. Did not interrogate the ECU for DTC's. Not sure how long it takes the ECU to register an emission's related DTC. I am running a random tech metal cat in my downpipe with narrowband O2 sensors downstream of that.
Time to dust off the programming skills and start crunching the data.
Things to do:
1. compare wideband and narrow band voltage difference to my preturbo numbers to make sure the cat is working as effectively as my OEM cats -- or at least nearly so
2. wire some knock phones together as decribed by uprev so I can tune safely
3. write scripts to map hp/tq and mpg onto the fuel and timing maps so I can start playing with the targets.
Any thoughts on whether I am missing something would be greatly appreciated.
1. Install fuel pump. Check for leaks, dtc, etc. Done. This was a sphincter tightening experience. Day was cold and dry and i was deathly afraid of a static discharge bringing an abrupt end to my install. Having a electric pump and exposed wiring swimming in fuel seems like such a bad idea but I'm no engineer.
2. Install injectors and spark plugs. Then will set K mult, latency and min pulse width as described by uprev. Then log city, hwy and wot runs to make sure things look ok. Done...sort of. Installed NGK plugs 1 range cooler. Came gapped to 1.025mm. Service manual says 1.1mm. Left them at 1.025mm. Installed Bosch 0280158117 550cc injectors. According to http://www.boschdealer.com/specsheets/0280158117cs.jpg set minimum pulse width to 0.72ms and set latency such that 14v came out to 0.960 ms and 8V came out to 1.757 ms. Started with a K of 16555 by mult stock K value of 27592 by 330 (stock inj size) and dividing by 550 (new inj size). Cleared learned values and started the car. Required some throttle initially to stay running but then idled fine. Did a cylinder power test to make sure I was getting spark and fuel to each cylinder. Then increased K to 19000 as it seemed the measured AFRs where a little leaner (15.5 to 16.5) than the target AFR (around 14.7). This helped but some more tweaking will be needed as the car is a bit sluggish. However, need to get on with installing the turbo. Will write some scripts to help home these numbers in later but for now the car cold starts, warm starts, idles and drives.
3. Install turbo. Check for intake, exhaust and oil leaks. Done. Started with a base tune by setting target AFR to 11.69 for all cells where RPM > 2400 and BFS > 6.07. Set fuel compensation to 100%. Set high det table to 45 deg in all cells. Subtracted 1 deg from timing map at cell (2.43, 1200) and 7 deg at cell (19.44, 6400) and then subtracted a graduated amount of timing for all the cells inbetween. I am running a 7psi WG spring. I then flashed the ECU and used RTT to set the K value to 24000 so that my AFR correction ranged between 95% and 105% at idle and free rev to 3k. I saved my ROM and then took the car out for a test drive. I stayed out of boost and logged about 20 minutes of data. Car seemed to run great except occasionally died if I revved the engine and immediately closed the throttle. Will need to work on that. Uploaded the data into djamps MAF tool (thank you BTW for that) to see how far off my MAF table might be. Was suprised to find that it was spot on. Correction was within +/- 5% for the cells that had data and most of the time there was no correction at all. Collected some more data at 50-75% throttle under boost keeping RPM < 5k with the same MAF results. No check engine lights were turned on either. Did not interrogate the ECU for DTC's. Not sure how long it takes the ECU to register an emission's related DTC. I am running a random tech metal cat in my downpipe with narrowband O2 sensors downstream of that.
Time to dust off the programming skills and start crunching the data.
Things to do:
1. compare wideband and narrow band voltage difference to my preturbo numbers to make sure the cat is working as effectively as my OEM cats -- or at least nearly so
2. wire some knock phones together as decribed by uprev so I can tune safely
3. write scripts to map hp/tq and mpg onto the fuel and timing maps so I can start playing with the targets.
Any thoughts on whether I am missing something would be greatly appreciated.
Last edited by balaguru; May 12, 2013 at 07:54 AM.
Great write-up. You seem like a pretty smart guy…when you're all wrapped up what if you take it to a professional tuner and get a once-over? just to make sure you caught everything…would be great peace-of-mind and you might learn something in the case you did forget anything...
I'd like to. Just not sure anybody good would be willing to work with me in such an open manner since there might be intellectual property right issues. I'm definitely not installing a tune that locks me out of my own car.
Don't really care about dyno numbers per say except maybe to correlate my numbers to the figures that correspond to the 400whp number that most people seem to start having longevity issues at. I'd like to keep things under 400whp. I think calculated hp/tq numbers from datalogging is a better way to assess the performance of an engine.
Don't really care about dyno numbers per say except maybe to correlate my numbers to the figures that correspond to the 400whp number that most people seem to start having longevity issues at. I'd like to keep things under 400whp. I think calculated hp/tq numbers from datalogging is a better way to assess the performance of an engine.
Great write-up. You seem like a pretty smart guy…when you're all wrapped up what if you take it to a professional tuner and get a once-over? just to make sure you caught everything…would be great peace-of-mind and you might learn something in the case you did forget anything...
Last edited by balaguru; May 12, 2013 at 01:04 PM.
Thread
Thread Starter
Forum
Replies
Last Post
Lt_Ballzacki
Brakes & Suspension
39
Aug 6, 2021 06:19 AM