SOHC V6 tuning

threshold for tip in

Congratulations on finding out what works for stopping the tip in spark retard.
I just looked at mine and it is higher than stock.
It is 0.000299930
stock is .000075101

Maybe you can raise it and have the same effect, but still have spark retard on tip in if needed for some odd reason.

With your trans shifting at different rpms, I cant help but wonder if it will be consistent or not in the future. I have not found mine to be very consistent at WOT.

expected shift point

My expected shift point without PCM anticipation has always been higher than specified in the tune due to the time it takes for the PCM to react once the specified value has been achieved. Unfortunately, until finally disabling spark retard at tip-in the PCM has always anticipated the upshifts. I have increased the Trans TV Pressure for each upshift to reduce the overshoot after the shift point is achieved.

I can't think of any time I'd want the spark advance to deviate from the borderline knock table after being corrected for ACT and ECT assuming the engine is warmed up and everything is functional. However, if the PCM is unable to calculate load (failed MAF sensor) then tip-in adjustment of spark might be useful. I've spent some time adjusting my Load W/failed MAF table to match my Borderline Spark Table. I'm a little concerned about the impact to Torque Converter Lock Up Rate After Tip In since the PCM can no longer recognize tip-in but have Torque Converter Lock Up Rate W/high TP set to the same value.

2 > 3 upshift looks good

Here's a graph of the 2 > 3 WOT upshift that looks about right.

Trans WOT Shift RPM 23 = 5900, actual shift rpm (blue) = 5959
Tip In Shock Control (5) and Tip Out Decel Control (6) still work but there was no Tip In Spark Retard (not shown).
The vehicle acceleration (red) appears to be about max when the shift occurred.

Max Vehicle Acceleration Rate

John, the Max Vehicle Acceleration Rate in my stock tune is 15. I don't have to worry about exceeding that but I noticed in one of your posts that you have. If yours is still 15 you might want to increase it.

Max vehicle rate

2000StreetRod said:

John, the Max Vehicle Acceleration Rate in my stock tune is 15. I don't have to worry about exceeding that but I noticed in one of your posts that you have. If yours is still 15 you might want to increase it.

Click to expand...

Yes, mine was 15. I wonder what the ramifications were.
I have changed it to 63.99 just now.
Don't know when I will be driving it next.

Injector slope AFR tuning

Now that my Sport is quite drivable using MTF AFR tuning I've decided to make another attempt using injector slope AFR tuning. Here are some injector specifics from the calibration summary:

AHISL (high injector slope) = .016741 lb/sec or 60.27 lb/hr
ALOSL (low injector slope) = .017552 lb/sec or 63.19 lb/hr
FUEL_BKPT (fuel mass to switch between high and low slopes) = .00003363 lb/pulse

On my earlier attempt I used the above and an MTF from bench airflow testing of a Lightning 90 mm MAF sensor. The AFR varied from rich to extremely rich as load increased.
Below is a graph of my Siemans Deka 60 lb/hr injectors.


Using the graph I derived the numbers for comparison to the published data.
AHISL = 43 mg/6 ms or 7.167 g/s or .0158 lb/sec or 56.9 lb/hr
ALOSL = 10 mg/1 ms or 10 g/s or .02204 lb/sec or 79.34 lb/hr
FUEL_BKPT occurs at a pulswidth = 1.7 ms x .0158 lb/sec = .00002686 lbs/pulse

Since my derived parameters were fairly close to those published I concluded the published data was accurate. Next, I created three tunes and data logged each during short drives. I limited the throttle (load) to keep the AFR on the wideband meters above 10:1 since that is their lower limit. Then I plotted actual lambda vs MAF AD counts. The commanded lambda was .99 for each plot.


Plot A: AHISL=.0234441 lb/s, ALOSL=.0246 lb/s, FUEL_BKPT=.0000471
AHISL is 2.54 times the published and others scaled accordingly.
Plot B: AHISL=.02981 lb/s, ALOSL=.03121 lb/s, FUEL_BKPT=.0000625
ALOSL is the max allowed by Advantage 3 and others scaled accordingly.
Plot C: The only difference between B & C is The Returnless Fuel Pump, Fuel Rail Pressure was changed from 39.15 to 100 psi in the tune.
It is obvious from the plots that the 60 lb/hr injectors are way bigger than what I need for my modest power increase from 160 rwhp NA to 227 rwhp FI. I should have selected Deka 40s which would have allowed me to double my horsepower. I was surprised that changing the Fuel Rail Pressure in the tune made a difference since Injector Comp for Pressure was disabled.

My next attempt will be to double the fuel injector slopes and breakpoint multipliers from 1 to 2 and restore the Fuel Rail Pressure in the tune to the actual of 39.15. This may not have any effect since my stock strategy is set up for a returnless fuel system with no fuel pressure sensor.

no improvement

So I doubled the fuel injector slopes and breakpoint multipliers from 1 to 2 and restored the Fuel Rail Pressure in the tune to the actual of 39.15 and compared the results to Plot C for changes.

Plot D is with the Injector Comp for Pressure disabled and Plot E with it enabled. There was no significant improvement for actual vs commanded. I think I've tried every parameter available in my tune to reduce excessive richness due to my high flow injectors. I guess the only thing left to try is to reduce the actual fuel pressure.

I found a bench test of the 60s on the internet and the flow was 500 cc/min (47.6 lb/hr) at 30 psi vs 585 cc/min (55.7 lb/hr) at 40 psi. That's only a reduction of about 17% so I'm going to try 25 psi which if the injectors are linear would be a 25% reduction.

Fuel pressure

For what it is worth I came across a formula for changing fuel pressure to change the injector flow.
It goes like this

(New injector size / old injector size)squared * old injectors rated fuel pressure

so just as an example
if you wanted 45 lb injectors when you had 60lb ones...

(45/60) = .75
.75 squared is .5625
.5625 * 39.15psi = 22.021875 psi (new fuel pressure)

Keep in mind that when you have 20 inches vacuum the fuel pressure drops about 10 psi.

I wonder what would the effects of fuel pressure that low have on atomization and would there be a chance for vapor lock at 22-25 psi?

fuel injector delta pressure

Thanks for the formula. It indicates that my reduction to 25 psi would be equivalent to a 48 lb/hr injector. That flow rate may still be too large but it's a good starting point. Remember that I changed my fuel pressure sensor reference port from capped to connected to the intake manifold vacuum/boost port so the fuel injector delta pressure wouldn't vary from the pressure setting.

Fuel Pressure

"changed my fuel pressure sensor reference port from capped to connected to the intake manifold vacuum/boost port "

That's why the pressure would drop by 10 psi with 20 inches of vacuum.
approx .49 psi = 1 inHg

I also wonder how low we could go on fuel pressure before having any other unforeseen problems.

I did have my fuel pressure at 50 psi, and it went to about 64-65 under boost. That was with my old 30 lb injectors that were getting maxed out. I had adjusted the MAF transfer function to achieve the correct A/F mixture but I see now that altering the fuel pressure would change the fuel flow model not the air flow model so the right way would be to adjust the fuel injector slopes. I have no idea how that is done so I am watching this to learn from you.

I have since returned my fuel pressure to approx. 39.15 since that would keep the fuel flow model accurate providing that the injector data is correct or "corrected" in the tune.

Also I tried lowering fuel pressure to try to get my O2 monitors to run with these 60 lb injectors. But I did not change the fuel injector slopes to compensate for the lowered fuel pressure. I wonder if that would work or if I will still need to go with smaller injectors.

adjusting rail pressure

4pointslow said:

"changed my fuel pressure sensor reference port from capped to connected to the intake manifold vacuum/boost port "

That's why the pressure would drop by 10 psi with 20 inches of vacuum.
approx .49 psi = 1 inHg . . .

Click to expand...

I adjust my pressure at the rail with the engine off so there is no vacuum. I remove the fuel pump relay and connect contact 5 to a battery charger. Then I depress the Down button on the pump controller for more LED flashes than my desired setting change. Next I start the engine, bleed off the fuel pressure, and turn off the engine. Then I depress the Up button until the rail pressure reads what I want. I can't reliably adjust the fuel pressure with the engine running because the pump ramps up too fast compared to the fuel needs at idle. The fuel pressure constantly varies above and below the desired by several psi. As load increases and the fuel pump runs continuously (but not at constant speed) the fuel pressure variation decreases. There are still short duration spikes on the rail when the injectors close.

still rich at 25 psi

I needed to bring home a few trailer loads of wood bark mulch for my backyard garden so I loaded a closed loop tune and decreased the fuel injector differential pressure to 25 psi. Even at that lower pressure the actual lambda at WOT is at least 19% less (.81 vs .68 meter limit) than commanded.

It got so rich that the engine cut out at about 3600 rpm so I released the accelerator. Tomorrow I'll drop the pressure to 20 psi but that's not going to be enough to make up the difference. I'm thinking about altering the WOT Fuel Multiplier.

The stock value was .875 (richer) which I changed a long time ago to 1.00 for simplicity. I could change it to 1.2 (leaner) as a starting point and then adjust based on test results. Not having the actual matching the commanded would mean that I could never let the PCM go open loop except for WOT and when in closed loop the difference should never exceed 30% since the limits of STFT is .70 and 1.30.

drastic measures

Since lowering the fuel injector pressure differential to 20 psi was not going lean the mixture enough I decided to try 15 psi. Surprisingly, the idle fuel rail pressure stabilized to +/- 1 psi. I guess the characteristics of the electronic fuel pump controller matched those of the pump.

I decided it is not essential that entries in the base fuel table match actual lambdas. What is essential is that actual lambdas are as desired. So I set all entries in the base fuel table to 1.00 as a starting point.

I'll adjust the entries for large TP values as required to get my desired actual lambdas. I datalogged a test drive and found that in closed loop the STFTs are less than 1 when the injection pulse width is less than the injection break point. I'm going to try decreasing the breakpoint and the ALOSL even though I've read that the ALOSL should never be less than the AHISL.

weird commanded lambda at WOT

So I tried to lean my AFR at WOT by modifying the entries in the WOT Fuel Multiplier table.

For TP greater than 100 I made all of the entries in the Base Fuel Table 1.000.

The results were extremely weird. When the TP (brown, 755) exceeded the TP for WOT value (550) the commanded lambda (blue) dropped as low as .43.

Since the lowest ignitable fuel mixture according to Bosch is 0.7 the engine speed (red) started dropping so I released the throttle and tried again with similar results.

Even though the WOT Fuel Multiplier description states the range is 0 to 1.99 it obviously isn't for my strategy. I know that 1.000 works fine from previous datalogs so I'll use that and try adjusting the Base Fuel Table entries. I need about 25% less fuel and the description states the limit is 1.200. Maybe I can up the WOT Fuel Multiplier to 1.05 to get what I need.

I know that I'm being stubborn about not modifying the MTF but I think that accurate airflow measurements are an important performance indicator. When I pull off things to replace the leaking valve covers I'm tempted to replace the Deka 60s with something in the 30 to 40 lb/hr range to eliminate most of these crazy workarounds.

Edit: Apparently the max WOT Fuel Multiplier for my strategy is 1.00. I found a section of datalog when WOT was exceeded and the engine speed was below 4,000 rpm (cursor at 3,500 rpm).

When the engine was below 3,500 the commanded lambda should have been 1.07 but it was actually 1.00. Above 3,500 the lambda starts decreasing instead of increasing.

2nd Edit: I've confirmed that for my strategy the max lambda for base fuel is 1.20. Above that the commanded drops to .43. So the max combined commanded lambda is 1.20 (max fuel table) * 1.00 (max WOT multiplier) = 1.20

considering different injectors

Bosch makes a 36 lb/hr long style, high impedance injector with EV1 connectors (P/N 0280155868). Using Stan Weiss' tables and assuming a BSC of .55 for a supercharged intercooled engine:

53 fhp * 6 injectors = 318 fhp

With my current tune and configuration I have 227 rwhp which is about
227/.85 = 267 fhp

However, my engine has to produce an additional 30 fhp to rotate the blower so my total fhp is 267 + 30 = 297
If I increase the boost the power required could increase by 5 fhp. That only leaves a growth capability of 16 fhp which I might be able to achieve with a better tune.

Bosch also makes a similar 42 lb/hr injector (P/N 0 280 155 968).

Using the same Weiss table:

61 fhp * 6 injectors = 366 fhp with a potential growth of 54 fhp.

PCM interpolation & extrapolation

I determined long ago that for many calibration tables the PCM interpolates between table entries to determine the value to use for measured parameters. Also, for many tables if the value of the measured parameter is outside of the limits in the table the PCM uses the closest entry. Today after a datalog I realized the importance of making table limits equal or beyond those that might occur.

Below is the Base Fuel Table for my latest tune.

When the PCM is in open loop, the TP is 700 and the RPM is 3000 to 6500 the commanded lambda (STFT) should be 1.10. However, for the portion of the datalog shown below when the TP (brown) exceeds 700 (cursor at 702) the STFT starts decreasing as the RPM (blue) increases from 3636 to 3855 and then decreases to 3405 just before I released the accelerator pedal.

I assumed for an earlier tune that the commanded lambda dropped to .43 because the entry exceeded 1.20 but it may have been due to the TPS value exceeding the limits of the table entries. I don't know how the PCM is arriving at the commanded lambda. It is obviously not extrapolating (using the slope) from the available entries.

Below is another datalog of the same tune with the TP held at 458. I'm encouraged that the STFT is 1.17 from 3324 to 3710 rpm matching the BFT and the actual lambda is about .98.

My Bosch fuel injection manual states that the best power lambda is .86 but with my forced induction I should probably shoot for .82 to avoid detonation.

I am changing my table range to make sure what occurs is always within the table.

The greatest relative TPS value I've ever recorded is 758 so 770 should be safe. The rev limiter is set to 6250 rpm so 6500 should be safe.

.98/.82 = 1.20 so I need to change my upper values to .975 (1.17/1.20). I still have a chance at tuning the AFR without altering the upper half of the MTF.

loop

Are you sure it is in open loop?
It was my understanding that in stock applications when the PCM is in open loop, STFT is supposed to be 0% because the PCM is ignoring the O2 sensor.

Now with SCT LiveLink, STFT at WOT(or 70% or more load) is supposed to show commanded A/F. I guess that is what you are looking at. I am just not used to looking at the Fuel table in TP vs RPM.
What is the set point that your PCM sees TP at WOT?

If it is in open loop, maybe there is another variable effecting it so it doesn't read exactly what is in the fuel table.
Maybe load with failed maf?

open loop

4pointslow said:

Are you sure it is in open loop?

Click to expand...

For that tune open loop is set to TP = 200.

4pointslow said:

It was my understanding that in stock applications when the PCM is in open loop, STFT is supposed to be 0% because the PCM is ignoring the O2 sensor.

Click to expand...

The PCM ignores the O2 sensor but sets the STFTs to commanded lambda from the base fuel table. For some of my early tunes I set the open loop value to 0 and tuned AFR entirely in open loop.

4pointslow said:

Now with SCT LiveLink, STFT at WOT(or 70% or more load) is supposed to show commanded A/F. I guess that is what you are looking at. I am just not used to looking at the Fuel table in TP vs RPM.
What is the set point that your PCM sees TP at WOT?

Click to expand...

I have WOT set to 550.

4pointslow said:

If it is in open loop, maybe there is another variable effecting it so it doesn't read exactly what is in the fuel table.
Maybe load with failed maf?

Click to expand...

I doubt it but anything is possible.

Correct commanded lambda at WOT

Increasing the TP range in the Base Fuel Table fixed my weird commanded lambda at WOT.

At the cursor the TP=756, STFT=.98 (same as commanded in BFT) from 2360 to 5365 rpm.

The actual lambda for the engine speed range varied between .82 and .85 (STFT and actual have different scales in the graph).

I'm still having problems with lambda when MAF AD Count is less than 300.

With the PCM in closed loop the STFTs are richening the AFR by about 20%. When I decrease the fuel injector low slope to make the AFR richer the PCM starts doing weird things like changing the breakpoint and sometimes getting lost because the low slope is less than the high slope. Right now the only time it really matters is after engine start when the PCM is in open loop until the O2 sensors have had time to heat up.

ranges

Dale that is great, you figure out some of the hardest things.
I didn't even know what interpolation was till you mentioned it. I had to look it up on the internet to see what it meant. Now you are figuring out all kinds of other stuff with this PCM tuning.

I will keep this in mind when looking at my tune. Of course I have load instead of TP but it would probably react the same way.
Great stuff here!