7’s Boost Attempt Part Deux
03-24-2015, 05:11 PM
#21
Junior Member
Thread Starter
iTrader: (1)
Join Date: Apr 2005
Location: Bethesda, MD
Posts: 430
Total Cats: 9
Hopefully I can help, as I've just gone down the same path.
A few checks:
*calibrate the wideband. Is it solely the Autometer, or which wideband do you have and how is it hooked up?
*what is your "stoich" set to in the MS3's software? It's under the Basic/Load Settings and then General Settings. It appears TunerStudio wrote the software assuming that gasoline is 14.7:1, so unless you're running E85, just make sure it's 14.7. This is of course different than reality (a different topic), but it puts the software in the same headspace as the code writer and also gives you a foundation that from now on Lambda 1.0 = "14.7 AFR"
*You should not be running closed loop fuel at this point. In the long term, you want the VE table roughly sorted with each cell tuned manually at a steady state. VEAL is nice and all but still a crutch.
A few checks:
*calibrate the wideband. Is it solely the Autometer, or which wideband do you have and how is it hooked up?
*what is your "stoich" set to in the MS3's software? It's under the Basic/Load Settings and then General Settings. It appears TunerStudio wrote the software assuming that gasoline is 14.7:1, so unless you're running E85, just make sure it's 14.7. This is of course different than reality (a different topic), but it puts the software in the same headspace as the code writer and also gives you a foundation that from now on Lambda 1.0 = "14.7 AFR"
*You should not be running closed loop fuel at this point. In the long term, you want the VE table roughly sorted with each cell tuned manually at a steady state. VEAL is nice and all but still a crutch.
-Gauge/ Wideband is Autometer Elite. It has Analog Ground, Sensor Ground and Data all going direct into the MS harness. Power is coming straight from fuse box. Is there a way to confirm the gauge is putting out an accurate reading?
-I'll have to check this next time the car is running. It was reading 14.7 on my laptop gauge but again, I haven't synced the wideband yet.
-I'm going to do my best with this. Can I run VEAL first through the whole range to get me 75% tune and then start manually rounding from there?
Reply
0
0
03-24-2015, 06:49 PM
#22
I recommend powering the wideband with the MS3 Basic's supplemental power supply. That way everything is on the same ground and the risk of offset is minimal. This is likely why your Autometer and MS3 are currently showing different AFR's.
Everyone has different tuning philosophies, so I'm mainly going off how my OEM engine calibrator friend is approaching my tuning. Try this after sorting out the wideband offset issue:
*Turn off closed loop fuel, "No correction"
*Spend a lot of time with the car just at idle. Very gently adjust the VE cells at idle such that the AFR targets stoich or slightly richer, wherever it's stable. Remember, the job of tuning the engine is to figure out how the engine wants to run, not what our personal opinions of how we think the engine should run. Put that BS away and focus on how the engine actually wants to run.
*Only once idle is stable you can apply a couple degrees of throttle. Hold it perfectly steady and then adjust that VE cell until you're at your AFR target.
*A little more throttle angle [should be able to hold it steady every few hundred RPM, but more importantly try to hold the RPM in the middle of each cell's breakpoint], and adjust that VE cell until you're at your AFR target.
*Rinse and repeat for a few more cells, all with the car in neutral in your driveway. At this point you'll be able to calculate a slope. This slope is fairly linear up until about 4500 RPM, you can infer the RPM cells you missed now at the load you were at just revving the engine.
If you decide to use VEAL, remember that it only works at steady state, but also for whatever reason it's not very good at tuning idle cells. Also, any acceleration enrichment's will throw it off. When I use it on the road, I'll hold that load and RPM perfectly steady before hitting "start auto tune" with it set on "easy". If there are any cars that get in the way, click "stop auto tune" before moving the throttle. This still takes a while to be steady and hit the more major cruising cells.
From there I recommend figuring out the shaping of your VE in terms of RPM and load axes. Guessing is not a good idea. Use your datalogs to help figure out how much to change problem cells. If you're running 12.2:1 at 75 kpa at 3k rpm, but you really want to run 14.7, it means you have 17% too much fuel in that VE cell, so you have to multiply that VE cell by 0.83 to get 14.7:1. In the end it'll end up looking just like your torque curve in 3 dimensions, so the highest numbers will likely be from 4500 to 5500 at full boost. The lowest numbers will be below idle.
Of course, this is assuming your MAT density correction curve is appropriate. If you find on a markedly hotter or colder day your AFR's are off, you'll have to adjust the MAT density curve to get more or less fuel at those MAT's. This is a multiplier just like VE, so calculate the difference between cold and hot MAT AFR's are plug it in.
All of this will be masked if you run with O2 correction on, and while the car will probably run OK assuming the AFR correction PID is responsive enough with enough authority, the ECU will always be trying to tug the fueling somewhere else than having the fundamental mapping correct. Put simpler, at this point my VE table and MAT density correction curve are such that running open-loop fuel results in about perfect AFR targets over a tested range of MAT's. When I turn on closed loop it barely needs correction, if any.
Ok, that's enough to chew on for now.
Everyone has different tuning philosophies, so I'm mainly going off how my OEM engine calibrator friend is approaching my tuning. Try this after sorting out the wideband offset issue:
*Turn off closed loop fuel, "No correction"
*Spend a lot of time with the car just at idle. Very gently adjust the VE cells at idle such that the AFR targets stoich or slightly richer, wherever it's stable. Remember, the job of tuning the engine is to figure out how the engine wants to run, not what our personal opinions of how we think the engine should run. Put that BS away and focus on how the engine actually wants to run.
*Only once idle is stable you can apply a couple degrees of throttle. Hold it perfectly steady and then adjust that VE cell until you're at your AFR target.
*A little more throttle angle [should be able to hold it steady every few hundred RPM, but more importantly try to hold the RPM in the middle of each cell's breakpoint], and adjust that VE cell until you're at your AFR target.
*Rinse and repeat for a few more cells, all with the car in neutral in your driveway. At this point you'll be able to calculate a slope. This slope is fairly linear up until about 4500 RPM, you can infer the RPM cells you missed now at the load you were at just revving the engine.
If you decide to use VEAL, remember that it only works at steady state, but also for whatever reason it's not very good at tuning idle cells. Also, any acceleration enrichment's will throw it off. When I use it on the road, I'll hold that load and RPM perfectly steady before hitting "start auto tune" with it set on "easy". If there are any cars that get in the way, click "stop auto tune" before moving the throttle. This still takes a while to be steady and hit the more major cruising cells.
From there I recommend figuring out the shaping of your VE in terms of RPM and load axes. Guessing is not a good idea. Use your datalogs to help figure out how much to change problem cells. If you're running 12.2:1 at 75 kpa at 3k rpm, but you really want to run 14.7, it means you have 17% too much fuel in that VE cell, so you have to multiply that VE cell by 0.83 to get 14.7:1. In the end it'll end up looking just like your torque curve in 3 dimensions, so the highest numbers will likely be from 4500 to 5500 at full boost. The lowest numbers will be below idle.
Of course, this is assuming your MAT density correction curve is appropriate. If you find on a markedly hotter or colder day your AFR's are off, you'll have to adjust the MAT density curve to get more or less fuel at those MAT's. This is a multiplier just like VE, so calculate the difference between cold and hot MAT AFR's are plug it in.
All of this will be masked if you run with O2 correction on, and while the car will probably run OK assuming the AFR correction PID is responsive enough with enough authority, the ECU will always be trying to tug the fueling somewhere else than having the fundamental mapping correct. Put simpler, at this point my VE table and MAT density correction curve are such that running open-loop fuel results in about perfect AFR targets over a tested range of MAT's. When I turn on closed loop it barely needs correction, if any.
Ok, that's enough to chew on for now.
Reply
0
0
03-25-2015, 08:27 AM
#23
Boost Czar
iTrader: (62)
Join Date: May 2005
Location: Chantilly, VA
Posts: 79,501
Total Cats: 4,080
I do not subscribe to Ryephile's tuning methodology whatsoever.
Your wideband also sucks. It's only 0-4v and the manual doesnt even tell you the AFR scale. it could be 10-18AFR or 7-22AFR or 12-16AFR.
you need to calibrate the voltage:AFR scale in TS to match your gauge, but I can't find the scale you need to use. That's pretty ******* pathetic coming from a gauge/metering company.
What I love most is this:
From their website -- 0-5V data output feed for ECU, data acquisition unit, or Laptop tuning. 10:1 to 20:1 AFR Range for Gasoline Applications.
Range: 6:1-20:1 AFR
From the install manual -- The Auto Meter Wideband Air / Fuel gauge has a signal output for supplying information to a Data Logger or engine management system. The signal provided is a linear 0-4v output. See the BGD Range Setting Mode for information about how to define the air/fuel mixture values for low (0v) and high (4v) ranges of the output in order to scale this linear signal output for best operation with your Data Logger or engine management.
I would try 10-20AFR and 0-4v and see if that doesn't get you close.
Your wideband also sucks. It's only 0-4v and the manual doesnt even tell you the AFR scale. it could be 10-18AFR or 7-22AFR or 12-16AFR.
you need to calibrate the voltage:AFR scale in TS to match your gauge, but I can't find the scale you need to use. That's pretty ******* pathetic coming from a gauge/metering company.
What I love most is this:
From their website -- 0-5V data output feed for ECU, data acquisition unit, or Laptop tuning. 10:1 to 20:1 AFR Range for Gasoline Applications.
Range: 6:1-20:1 AFR
From the install manual -- The Auto Meter Wideband Air / Fuel gauge has a signal output for supplying information to a Data Logger or engine management system. The signal provided is a linear 0-4v output. See the BGD Range Setting Mode for information about how to define the air/fuel mixture values for low (0v) and high (4v) ranges of the output in order to scale this linear signal output for best operation with your Data Logger or engine management.
I would try 10-20AFR and 0-4v and see if that doesn't get you close.
Last edited by Braineack; 03-25-2015 at 08:38 AM.
Reply
0
0
04-06-2015, 06:59 PM
#26
Junior Member
Thread Starter
iTrader: (1)
Join Date: Apr 2005
Location: Bethesda, MD
Posts: 430
Total Cats: 9
Scott, You were right I had to calibrate the WB in TS to show 0-4v = 10-16 however for some reason I had to go into the settings on the gauge and also, set that for 16V. It really is a crappy gauge, I know. Next time around I'll know better.
Here's a sight I haven't seen in quite some time Car is on the ground and almost ready to hit the road aside from one more issue:
I'm having a hard time starting with Revs base map. It will crank, turn over and stop, crank, turn over and stop, then eventually on the 5th or 6th try, it fire and run fine no problem. RPM will dip down a bit below idle if I rev it but I suspect I can tune that out. Suspecting the starting issue is with WUE. I took a log of one failed start and one successful start. Any thoughts?
Here's a sight I haven't seen in quite some time Car is on the ground and almost ready to hit the road aside from one more issue:
I'm having a hard time starting with Revs base map. It will crank, turn over and stop, crank, turn over and stop, then eventually on the 5th or 6th try, it fire and run fine no problem. RPM will dip down a bit below idle if I rev it but I suspect I can tune that out. Suspecting the starting issue is with WUE. I took a log of one failed start and one successful start. Any thoughts?
Reply
0
0
04-07-2015, 01:51 PM
#29
... It will crank, turn over and stop, crank, turn over and stop, then eventually on the 5th or 6th try, it fire and run fine no problem. RPM will dip down a bit below idle if I rev it but I suspect I can tune that out. Suspecting the starting issue is with WUE. I took a log of one failed start and one successful start. Any thoughts?
That whole cranking-ASE-WUE-VE handoff has to be tuned for all temperatures. Of course, you really need the VE cells dialed in first before you perfect the ASE and WUE percentages, so once you get the car up to temp, tune the VE cells around idle and then work backwards to get the warm-up right.
Reply
0
0
04-07-2015, 07:54 PM
#32
Junior Member
Thread Starter
iTrader: (1)
Join Date: Apr 2005
Location: Bethesda, MD
Posts: 430
Total Cats: 9
Thanks guys, for the Noobie hand holding. It's much, much appreciated.
Sorry if I wasn't clear, once the car starts its idles fine. It just takes a while to get it to catch. I increased the ASE in each cell by a factor of 10 and that really helped it not dying immediately. How aggressive can I get with the increase off the base tune and does the line have to stay linear? Here's a screen shot for reference:
Sorry if I wasn't clear, once the car starts its idles fine. It just takes a while to get it to catch. I increased the ASE in each cell by a factor of 10 and that really helped it not dying immediately. How aggressive can I get with the increase off the base tune and does the line have to stay linear? Here's a screen shot for reference:
Reply
0
0
04-08-2015, 11:02 AM
#33
For my setup, the "base tune" from Rev was so far off on [IIRC] literally every setting that I now call it the "random guess file". As such I would not assume anything is close enough to be considered a tune in any way.
Ok, regarding crank and warmup, let's see if I can put it in "words". If the engine goes:
-->crankcrankcrankcrankcrankcrankcrankcrank....... ., then you need a ton more cranking pulsewidth at that coolant temp
-->crankcrankcrankBLUBblublubdead....., then you still need more cranking and probably more ASE at those coolant temps
-->crankcrankVROOOOoooooodead....., then you're finally OK on cranking and definitely need more ASE and probably WUE
-->crankcrankVROOOOOOOMMMMmmmmdead...., then your ASE probably needs more, and your WUE needs more too
-->crankcrankVROOOOOOOMMMMmmmmmIdleIdleIdleIdleIdleI dleIdle...dead..., then your ASE is good, but you need more WUE at that coolant temp.
Rinse and repeat at all possible coolant temps until the whole chain of events from crank to ASE to WUE and finally straight-up VE map are satisfactory.
Ok, regarding crank and warmup, let's see if I can put it in "words". If the engine goes:
-->crankcrankcrankcrankcrankcrankcrankcrank....... ., then you need a ton more cranking pulsewidth at that coolant temp
-->crankcrankcrankBLUBblublubdead....., then you still need more cranking and probably more ASE at those coolant temps
-->crankcrankVROOOOoooooodead....., then you're finally OK on cranking and definitely need more ASE and probably WUE
-->crankcrankVROOOOOOOMMMMmmmmdead...., then your ASE probably needs more, and your WUE needs more too
-->crankcrankVROOOOOOOMMMMmmmmmIdleIdleIdleIdleIdleI dleIdle...dead..., then your ASE is good, but you need more WUE at that coolant temp.
Rinse and repeat at all possible coolant temps until the whole chain of events from crank to ASE to WUE and finally straight-up VE map are satisfactory.
Reply
1
1
04-08-2015, 05:49 PM
#34
Junior Member
Thread Starter
iTrader: (1)
Join Date: Apr 2005
Location: Bethesda, MD
Posts: 430
Total Cats: 9
Seems like the car needed a little of everything. I added a little more crank pulse and ASE and it starts now on the first or second try. I'll add a little more and try again after it cools down. I think its 80% there until after I get the VE table dialed in. I actually dove the car up and down the driveway today so promising signs!
Reply
0
0
04-12-2015, 05:05 PM
#35
Elite Member
iTrader: (10)
Join Date: Jun 2006
Location: Athens, Greece
Posts: 5,979
Total Cats: 356
Is this your opinion or your friend's opinion? I mean, wow. Just wow.
Reply
0
0
04-12-2015, 05:58 PM
#37
Junior Member
Join Date: Oct 2013
Posts: 77
Total Cats: 5
2nd endorsement for Rev's base tune. Started first crank on my NB2 w/ a custom turbo setup.
A little auto tune and I was at the track. Eventually got a little more aggressive with top end timing (and loved the pull to red line) but thankful he left that to me.
A little auto tune and I was at the track. Eventually got a little more aggressive with top end timing (and loved the pull to red line) but thankful he left that to me.
Reply
0
0
04-13-2015, 09:49 AM
#38
Propped for the caveman's guide to startup settings for ra tards.
For my setup, the "base tune" from Rev was so far off on [IIRC] literally every setting that I now call it the "random guess file". As such I would not assume anything is close enough to be considered a tune in any way.
Ok, regarding crank and warmup, let's see if I can put it in "words". If the engine goes:
-->crankcrankcrankcrankcrankcrankcrankcrank....... ., then you need a ton more cranking pulsewidth at that coolant temp
-->crankcrankcrankBLUBblublubdead....., then you still need more cranking and probably more ASE at those coolant temps
-->crankcrankVROOOOoooooodead....., then you're finally OK on cranking and definitely need more ASE and probably WUE
-->crankcrankVROOOOOOOMMMMmmmmdead...., then your ASE probably needs more, and your WUE needs more too
-->crankcrankVROOOOOOOMMMMmmmmmIdleIdleIdleIdleIdleI dleIdle...dead..., then your ASE is good, but you need more WUE at that coolant temp.
Rinse and repeat at all possible coolant temps until the whole chain of events from crank to ASE to WUE and finally straight-up VE map are satisfactory.
Ok, regarding crank and warmup, let's see if I can put it in "words". If the engine goes:
-->crankcrankcrankcrankcrankcrankcrankcrank....... ., then you need a ton more cranking pulsewidth at that coolant temp
-->crankcrankcrankBLUBblublubdead....., then you still need more cranking and probably more ASE at those coolant temps
-->crankcrankVROOOOoooooodead....., then you're finally OK on cranking and definitely need more ASE and probably WUE
-->crankcrankVROOOOOOOMMMMmmmmdead...., then your ASE probably needs more, and your WUE needs more too
-->crankcrankVROOOOOOOMMMMmmmmmIdleIdleIdleIdleIdleI dleIdle...dead..., then your ASE is good, but you need more WUE at that coolant temp.
Rinse and repeat at all possible coolant temps until the whole chain of events from crank to ASE to WUE and finally straight-up VE map are satisfactory.
Reply
0
0
04-13-2015, 01:43 PM
#40
I'll admit my setup is unique for a USA 1.6, but literally the only correct setting was the injector size. General settings all wrong, Triggers were wrong, startup/idle pulses all wrong, not a chance in hell it would've even sputtered the 1st crank.
If taking the due diligence to make sure the software matches the hardware is "trippin", then absolutely. No vendor is perfect (some marques less-so than others). Maybe Rev got confused and sent me the wrong file, it can happen. If I'm literally the only customer that fell through the cracks, then his good rep is worth it.
I'm not throwing some hissy fit, I'm saying check your **** and don't assume anything. It's awesome to hear that Sevan's base tune let him drive the car straight away.
Last edited by Ryephile; 04-13-2015 at 02:51 PM. Reason: smoothed it out
Reply
0
0
