I’m not a qualified tuner and damaged my engine just last weekend doing something stupid while self tuning. The below is bunch of abstracts from posts around this site as well as thoughts and information from other engine tuning sites. I thought I’d share with you what I had so far as it might help you. Remember you mess up and it could get expensive. Its mostly copy and pastes so if you want to find the original post use search and its probably not too far away.
I highly recommend you have a wideband and log AFR’s or forget tuning. You can tune based on fuel trims but I would suggest thats best left for someone who has a better understanding then the newbie tuner. When learning about engines ensure you understand Volumetric efficiency
Tuning Strategie:
Few different ways to attack the tuning strategy. However before deciding on a strategy make sure you take in as much information as possible. Make sure you understand the data you are logging and you understand your fuel AFR’s. Get many data points at different temperatures (air temps, inlet temps, engine coolant temps, head temps).
1.) You can pick a boost level (preferably based on a thing like that other thing, but for your own turbo) and then pick an AFR, and then turn up timing until you start getting knock, then back down a little. (Or turn it up until you stop getting extra power.)
2.) you can pick an AFR and timing, and turn up boost until you get knock, then back down a little.
3.) you can pick boost and timing, and then lean out the AFR until you get knock, then back down a little.
I went with strategy 1 eventually when I was comfortable. What that means still very high level is for starters just set the AFR around 11.1 then drop the timing to a low safe amount, and crank the boost as high as your willing to go (based on loads of data points from around this site), then start slowly putting some timing back in by small increments till you get knock retard then back it off 1-3 degs. Then smooth it all out and keep testing. Remember:
1.) Changes all impact each other
2.) Fuel quality impacts results
3.) Temp impacts results
4.) you will be able to get more aggressive with the timing advance the higher the rpms rise,
5.) in the lower rpms you will want much less timing.
6.) You want to lower your timing around peak torque
Then once timing and boost are good, lean out your AFR a bit more if you can, and that should be a good tune.
================
Boost control:
Most tuners seem to run open loop boost control, however when street tuning its easy to dial in your boost in closed loop using the 'desired boost' table (its in inches of mercury and a rough duty cycle estimate and then log wastegate duty cycle, boost error, desired boost etc in a number of gears and then work out what dutycycle it likes for what rpm and the 'desired' boost level you have commanded. This gives good data on approximate wastegate duty cycle if you switch to open loop. The only thing to look out for is if you go straight open loop (i.e. you're only running off the duty cycle table) is that you may need to command a little less duty cycle at 2500rpm compared to what the pcm calculates in closed loop.....which is normally 100%.
OPEN LOOP: you need to zero the proportional gain multiplier table, zero all integral and proportional gains (on/off boost ect) and wastegate period. The wastegate boost FMEM (failure management effects mode) is set at a 50% reduction from factory so that needs to be set to zero reduction. All torque 12, 23 and 34 shift underboost need to be zero'd. P1227 limit (overboost) needs to be set to it's maximum. The over and underboost for open loop DC and the over and underboost for open loop DC error need to be maxed out
CLOSED LOOP: the system references a value from the wastegate duty cycle table, measures the boost, determines an error against the "desired boost table" and then uses the PI system to correct the duty cycle to reach the target level. The factory PI system is reasonably good so just leave is standard for now. In factory form the throttle body is a fair way from the turbo and too far to use aggresssive proportional gains so just work on the basics to start with. With closed loop all you really need to do is zero out the torque 12, 23 and 34 shift underboost, raise the 'P1227 limit' a little, make the overboost for 'open loop DC error' more negative and raise the 'underboost for open loope dc error'. The latter two won't need much at low boost...say 0.5 adjustment. From here you can set the desired boost table which is in inches of mercury. The PCM will always try and target these values and will adjust the duty cycle to get there. Ignore the manifold charge temps (MCT i.e. y-axis) and set all verticle values to the same desired boost level. Now you have to dial in your wastegate duty cycle that will achieve your 'desired boost'. These are the values that the pcm will use to start controlling boost so it is important to get them right (again ignore MCT and all verticle values can be the same). You will need to datalog the pcm's adjusted 'wastegate duty cycle' value and compare it to your commanded duty cycle. Refine this so it's adding a small amount of dutycycle in low gears and removing a little duty cycle in high gears.
==========================
TIMING:
Torque is power = highest Air Load point at WOT.
= 3250RPM 2.41 21/12/2016
Spark = Lower of = (borderline knock + (IAT Spark Base * Mult) + (ECT Spark Base * Mult) + Lambda Spark Base + (Actual Cam Angle * Variable Cam Spark Base).) or MBT
http://www.efidynotuning.com/saftot.htm
Don't get too greedy on a boosted engine, its been said before that the general rule of thumb to be safe is to pull out 1/2 degree for every lb of boost,Although pulling out 1/2 degree per pound can result in satisfactory results more times than not I always yank out a couple degrees more at first and work my way up rather than over advancing and working my way down.
On boosted engines I typically stay at least 3 to 4 degrees away from MBT regardless, unless I know I have enough octane that detonation is not going to ever occur (I.E. Race Gas, E-85, Nitro).
MBT = max brake torque. Ford engineers got it right so leave it alone unless you have made changes to an engine that have impacted MBT. This is usually something that changes the - Volumetric efficiency of the engine or VE.
The MBT table should be left alone unless there have been modifications that would have actually changed the MBT. Other than raising the MBT table up for higher load values, I seldom touch it.
Copying the MBT table to the borderline spark table is not ok.
There is some modifiers for IMRC, combustion noise, EGR etc but these aren't significant in the turbo's.
The BKT and MBT calc's are run...the lower of the two is referenced and then there is an additional retard for knock sensor activity.
However the ford system has many Spark Sources. Idle feedback, deceleration etc etc...there's about 18 in total iirc.
E85 can support up to 7 degrees of timing over 98
==========================
Fuel Control:
Injector terminology explained:
http://www.myo-p.com/Ford-EEC/EEC%20...tor%20info.htm
==
Understand the difference between open loop and closed loop fueling and when each is used. Log values to further grasp.
Look through the fuelling tabs to understand delays and what triggers make you go closed / open loop.
==
Fueling is typically control from the base fuel table. Get used to reading lambda values and converting between stoich and lamda
==
Best engineering aigned is Speed Density Fueling. BA/BF doesn’t look possible on the FG. MAP per Airmass (High Res)
Next best is Injector scaling to correct and then working with injector high / low slope to get the right commanded AFR.
Find the ID1000 Data on his website also, input all the data.. Low slope, High slope, breakpoint, voltage offsets, (off the top of my head)
Using a wideband, start it up, let it get to operating temp then turn off and start again you have 10 or so seconds when the car will run in open loop so adjust your low slope until your achieving your target lambda (once the fuel starts to trim it will give you a good idea of what percentage to alter the slope by) lower numbers add fuel, higher deducts fuel... The Base Fuel Table is your target fueling.. Adjust your high slope to target what's in your base fuel table when at WOT
What I recommend is disable LTFT and set O2 min temp to 4096?F if you are going to use a wideband. Then use wideband_afr / commanded afr to get your fuel trim error.
==
Make sure when tuning if your high slope is shifted when you change offset/low slope/break point. So if you dial in your high slope then adjust anything at all you put it out, hence the calculator. For example if you change your breakpoint at all, you massively change your fueling on the high slope.
Tune fuel in the following order:
Offset
Low slope
Breakpoint
Highslope
If you tune it in any other order you'll be chasing your tail forever.
==
If you are rich or lean by the same amount at different MAP, then your slope is out. A lower slope value will equate to a higher airmass and therefore lead to a longer injectors pulse width per intake stroke. If you are rich or lean by differing amounts at different MAP then your offset is out. As with the slope a lower offset will ultimately increase the injector pulsewidth.
================
Airmass tuning (Speed Density or SD)
Airmass per intake stroke = (MAP - offset)/slope
LOAD = ((MAP - offset)/slope)/displacement
This is in the simplest form and does not take into account slope and/or VE correction. The actual algorithm is much more complicated.
If your afr is lean at the same MAP then you need to lower your slope. If your afr is correct at one MAP value and then lean at a higher MAP value then your offset is out.
The offset has a huge effect at low MAP values.
Remember that the airmass calc is used in the load calc so your spark advance will be affected as well as many other things.
Also it is essential that you start with the correct injector scaling, something that can't be done with the nizpro's etc. Most tuners simply 'touch up' the SD tables to smooth the afr's.
A full SD tue is very demanding on the vehicle, not too mention time consuming given the unique slope and offset for each cam angle.
Speed DEsnity:
14.
2.75
(14.65-2.30)/22500/(14.64*.80)
==========
Enjoy the brain dump. I hope no-one minds me copying their posts into here.
Reply With Quote
Originally Posted by Molch
