SOHC V6 tuning

[2000StreetRod]

I finally purchased SCT's Advantage III Ford Pro Racer Software package to learn how to generate my own tunes. I decided that with the mechanical and electrical completion of my M90 supercharger installation it was a good time to start. There seems to be minimal information on the forum for how to tune so I'm starting a how-to thread to document my steps and what I learn: Self Tuning My Supercharged 4.0L SOHC V6. That thread will remain closed until the tune is complete (if ever) to improve the readability of the tuning process. Please post any comments (and they are encouraged because I need all the help I can get) regarding the how-to thread in this thread. I know it's rather presumptuous for me to generate a thread on how to tune since I've never done one but I've learned that my logic is less flawed when I document a process for others to review. Hopefully, the thread will be of some help to others.

 

[Dono]

No doubt, this will help many understand whats involved.
I'll be watching.
I'd like to get to the point that your at one day.

 

[2000StreetRod]

Idle speed control?

When my engine was normally aspirated (NA) I could adjust the throttle plate stop so the engine would just barely idle when the IAC valve connector was disconnected. Then with the IAC valve connector connected the engine would idle at any reasonable commanded idle speed. With forced induction (FI) the engine is idling fast (1000 rpm) in Park when the stock commanded idle speed in Neutral is only 656 rpm. I can lower the idle speed by adjusting the throttle plate stop and throttle cable but then if I disconnect the IAC valve connector the engine will immediately die.

The NA engine load at idle (1,000 rpm) was .13 and the MAF AIRFLOW was .7 #/min (155 MAF AD COUNTS). Now the FI engine load at idle (1,000 rpm) is .18 and the MAF AD COUNTS = 180 due to the power required to rotate the blower.

Advantage III provides access to a table named ISC Neutral Idle Air. There is a similar table named ISC Drive Idle Air.

The description states: These functions are the amount of air, in #/min, that it takes to make the engine idle at the RPM in the left column. The PCM uses these values to determine the idle speed duty cycle. If the motor is modified to change its air requirements at idle, like a larger camshaft, then these functions will need to be modified to prevent idle surging.

180/155 = 1.16 It seems reasonable to increase all of the values in the second column by 15% and then adjust the throttle plate stop as I did for NA. Does anyone have any experience or opinions?

 

[2000StreetRod]

Calculating injector pulsewidth?

I've accepted the challenge to determine how to calculate injector pulsewidth. I recorded the number of clock tics between pips and plotted them (bold green) vs engine speed (light purple). I am puzzled by the results.

The clock tic count is inversely proportional to the engine speed (min 924, max 1037) - seems reasonable. The ratio is about 5 tics/rpm.

 

[Dono]

Im betting your on your own on the idle thing (And everything else on tuning)
I had idle issues, and just adjusted my throttle stop plate as you stated above on my ohv.
This worked for me.

I really don't know what else James has done in the tune for my truck.

 

[2000StreetRod]

possible pw calculation

How about this? For a 6 cylinder engine there are 6 pips for every 2 crankshaft revolutions. 1,000 rpm = 16.67 revs/sec and 1 rev = .06 sec or 60 ms. 3 pips/rev = .33 revs/pip * 60 ms/rev = 20 ms between pips
20 ms/4,922 tics = 4.06 us/tic
The pulsewidth I logged at 1,000 rpm was 348 clock tics.
348 * 4.06 us = 1.413 ms which seems quite reasonable.

According to the characteristic injector flow curve a 1.4 ms pulse has a dynamic flow of 10 mg. The specified fuel breakpoint is .00003363 lb or 15.25 mg. Therefore, when the engine is idling in Park at 1,000 rpm the PCM is using the low injector slope.

There's probably a simpler method of performing the conversion. But maybe not since I haven't been able to find one on the internet. If the number of tics is inversely proportional to the engine speed then there should be 9,844 tics between pips at 500 rpm and 2,461 tics between pips at 2000 rpm. I'll verify that during a future datalog. I suspect tics are related to the PCM clock rate so a PCM with a different part number probably has a different conversion. However, there should be a simple conversion factor for determining injector pulsewidth in ms since a tic represents the same amount of time for the injector and the pip.

EDIT: Siemans specifies my injector has a dynamic flow rate of 15.7 mg for a 2.5 ms pulse with an injector delta pressure of 39.15 psi. I have modified my returnless fuel system with a constant 65 psi rail pressure and installed an electronic fuel pump controller that maintains an average 39.15 psi injector delta pressure.

 

[2000StreetRod]

ISC valve transfer function?

There's a table named ISC Valve Transfer Function that's accessible with Advantage III.

The description states: In most cases there is no need to change this table. However, if a different ISC valve is installed on the engine, then the correct transfer function should be entered for best idle control. I'm using the IAC valve for a 2000 Mustang GT because it simplified my remote IACV installation. I have not been able to find a transfer function for it posted on the internet. If modifying the ISC Neutral Idle Air and ISC Drive Idle Air tables doesn't provide the desired results I'll experiment with the ISC Valve Transfer Function.

 

[2000StreetRod]

Injector duty cycle calculation

Injector duty cycle is defined as the time an injector is energized during an engine cycle (intake, compression, combustion and exhaust). Calculating the injector duty cycle is easy:

DC = injector pulse width in tics/(6 * pip to pip interval in tics)

For example: 348/(6 * 4,922) = 1.18%

 

[2000StreetRod]

More clock tic data

The following are from the step test. The numbers are not exact due to changing engine speeds and the resolution of the display.

rpm.....tics between pips

1000.........4922
2000.........2495
3000.........1663
4000.........1147

tics * rpm = 4,872,250 or tics = 4,872,250/rpm is an approximate equation

The minimum pulsewidth usually occurs after the TPS is rapidly decreased to minimum.

I "blipped" the accelerator and the TPS relative value (red) went from 20 to 495 to 0 in 0.5 seconds. The injector pulsewidth (blue) went from 306 to 2274 to 86 clock tics in about .75 seconds. The engine speed (yellow) went from 1978 to 4340 to 1146 rpm in about 2 seconds. There is no indication of clipping with the minimum pulsewidth in the tune set to 0.2 ms. For the minimum pulsewidth (86 clock tics) observed:
pip to pip clock tics = 1366, rpm = 3828.25
tics * rpm = 5,229,389

(3828.25 rev/min)*(1 min/60 sec)*(sec/1000 ms)*(3 pips/rev)*(1366 tics/pip) = 261.5 tics/ms

86 tics/(261.5 tics/ms) = .33 ms min pulsewidth

Since three of the conversion factors will not change the equation can be reduced to:

PW in ms = 1/[(rpm/20,000) * (pip to pip tics/PW in tics)]

 

[2000StreetRod]

Max boost

While I haven't been watching the boost gauge as much as I should (I've been paying more attention to the wideband AFR gauges) the max boost I've noticed so far is about 6 psi. That is less than my anticipated 8 psi with the 2.7 inch dia. pulley. One possible reason is my high flow exhaust system. Another possibility (and more likely) is excess clearance between the M90 rotors and the housing. When I finally sell my Volvo and buy a 4 door Explorer I'll probably have my M90 rebuilt or buy a rebuilt one. Last week I asked the manager of the auto shop with the dyno when I could do some testing and he said maybe the end of this week. On the dyno I don't have to worry about road conditions - I only have to watch AFR and boost. I'm a long way from an optimized tune but my seat-of-the-pants accelerometer indicates a significant performance improvement from the M90 installation.

 

[2000StreetRod]

Upshift points to maximize acceleration?

For decades I've wondered if there is a theoretical way to determine the upshift engine speeds to maximize acceleration. The typical empirical method seems to be to upshift near max engine speed after horsepower has dropped and measure the time for the distance (i.e. 1/8 or 1/4 mile). Then repeat the process shifting earlier or later and compare times.

I wonder if the emphasis should be on torque instead of horsepower. With my M90 the torque at lower engine speeds should be considerably larger than when normally aspirated (NA). I should adjust my shift points to take advantage of the new torque characteristics but don't want to use the typical empirical method. Changing the rear axle ratio from 3.73:1 to 4.88:1 will decrease acceleration times due to mechanical advantage. 1st speed has a mechanical advantage over 2nd speed but at some for the same engine speed but at some point the upshift to 2nd has an acceleration advantage.

Below are torque plots of two NA pulls performed three minutes apart. The red one indicates a slighter larger torque than the blue one due to lower IAT and ECT.

I estimated the red data points and multiplied them by the gear ratios:

RPM -- RWTQ -- 1st ---- 2nd ----- 3rd ---- 4th ----- 5th
2000 -- 182 -- 449.5 -- 338.5 -- 267.5 -- 182.0 -- 136.5
2500 -- 192 -- 474.2 -- 357.1 -- 282.2 -- 192.0 -- 144.0
3000 -- 200 -- 494.0 -- 372.0 -- 294.0 -- 200.0 -- 150.0
3500 -- 205 -- 506.4 -- 381.3 -- 301.4 -- 205.0 -- 153.8
4000 -- 203 -- 501.4 -- 377.6 -- 298.4 -- 203.0 -- 152.3
4500 -- 193 -- 476.7 -- 359.0 -- 283.7 -- 193.0 -- 144.8
5000 -- 175 -- 432.3 -- 325.5 -- 257.3 -- 175.0 -- 131.3
5500 -- 138 -- 340.9 -- 256.7 -- 202.9 -- 138.0 -- 103.5
6000 -- 100 -- 247.0 -- 186.0 -- 147.0 -- 100.0 --- 75.0

1st speed 2.47:1
2nd speed 1.86:1 (2.47 * .75)
3rd speed 1.47:1
4th speed 1.00:1
5th speed 0.75:1

Assuming that an upshift should occur when the mechanical advantage of the next gear exceeds that of the lower gear the above calculations can be plotted to determine the shift points.

For the 1st to 2nd upshift the engine speed is 5280 rpm.

For the 2nd to 3rd upshift the engine speed is 5180 rpm.

For the 3rd to 4th upshift the engine speed is 5470 rpm.

For the 4th to 5th upshift the engine speed is 5280 rpm.

Comparing the stock values to my torque multiplication method doesn't add any creditability to my method:
------- Upshift ------------- Stock -- Revised
Trans WOT Shift RPM 12 -- 5500 ---- 5280
Trans WOT Shift RPM 23 -- 5500 ---- 5180
Trans WOT Shift RPM 34 -- 5800 ---- 5470
Trans WOT Shift RPM 45 -- 5800 ---- 5280

 

[Dono]

I would agree with you. I know my M90 v6 ohv kinda falls on its face at higher rpm's. Different motors, so I don't know for sure.

 

[2000StreetRod]

SOHC vs OHV V6

I'm surprised at how fast I hit 6250 in 1st with the M90 and only 6 psi of boost. I guess the overhead cam and head design improves the higher speed airflow compared to your OHV V6. I rely heavily on the rev limiter to keep from destroying my engine. My first trip to the dyno will be mainly to check my fuel tuning. But I will get a torque curve with non-optimized spark timing to work with. On a later dyno test session I'll perform a time to speed comparison using the typical shift points vs my torque multiplication method.

 

[CDW6212R]

I'm surprised at how fast I hit 6250 in 1st with the M90 and only 6 psi of boost. I guess the overhead cam and head design improves the higher speed airflow compared to your OHV V6. I rely heavily on the rev limiter to keep from destroying my engine. My first trip to the dyno will be mainly to check my fuel tuning. But I will get a torque curve with non-optimized spark timing to work with. On a later dyno test session I'll perform a time to speed comparison using the typical shift points vs my torque multiplication method.

There's the hidden magic. Real acceleration/improvement is how much quicker the engine gains rpm. That's not a typical dyno parameter, and that's what the track times do measure.

You will get it done I'm sure, find a way to track test it to find the best shift points. Keep an eye out for the quality of shifts, for slipping shifts, to keep them consistent and reliable. When those are under control you can narrow down the shift points easily. Well done.

 

[4pointslow]

Tuning

Glad to see the tuning is coming along. I cant wait to see the torque curve.
About the boost being lower than expected, I agree with you that the free flowing exhaust is probably why it is lower. That really means more flow in and out of the engine like it should be.

 

[2000StreetRod]

Calculating new shift points

Using my recent FI dyno plot I extracted rpm vs rwtq and multiplied them by the gear ratios:

RPM -- RWTQ ---- 1st ---- 2nd ----- 3rd ---- 4th ----- 5th
3000 -- 258.4 -- 638.2 -- 480.6 -- 379.8 -- 258.4 -- 193.8
3500 -- 262.0 -- 647.1 -- 487.3 -- 385.1 -- 262.0 -- 196.5
4000 -- 256.7 -- 634.0 -- 477.5 -- 377.3 -- 256.7 -- 192.5
4500 -- 247.2 -- 610.6 -- 459.8 -- 363.4 -- 247.2 -- 185.4
5000 -- 232.6 -- 574.5 -- 432.6 -- 341.9 -- 232.6 -- 174.5
5500 -- 213.9 -- 528.3 -- 397.9 -- 314.4 -- 213.9 -- 160.4
6000 -- 185.4 -- 457.9 -- 344.8 -- 272.5 -- 185.4 -- 139.1

1st speed 2.47:1
2nd speed 1.86:1 (2.47 * .75)
3rd speed 1.47:1
4th speed 1.00:1
5th speed 0.75:1

However, I realized that the previous gear multiplication method is invalid because the vehicle speed doesn't change during an upshift. The engine speed will only decrease according to the gear ratio separation.

For example, shifting from 1st at 5900 rpm results in a 2nd engine speed of 4480 rpm. Assuming that maximum acceleration occurs at maximum mechanical advantage (torque * gear ratio) then the shift should occur when the value in the lower gear drops below the value in the next gear. The optimum engine shift speed can be determined graphically using the gear ratio vs vehicle speed plots and the torque * gear ratio plots.

Interpreting the plots:
1A represents shifting 1st @ 4650 (599 product) which results in 2A (2nd @ 3500 and 488 product).
1B represents shifting 1st @ 5600 (513 product) which results in 2B (2nd @ 4200 and 470 product).
1C represents shifting 1st @ 5750 (492 product) which results in 2C (2nd @ 4370 and 463 product).
1D represents shifting 1st @ 5900 (471 product) which results in 2D (2nd @ 4480 and 460 product).
1E represents shifting 1st @ 6000 (459 product) which results in 2E (2nd @ 4500 and 459 product).
From the graph the optimum 1st to 2nd shift point is 6000 rpm because the product for 1st has decreased to the product for 2nd.

Interpreting the plots:
2A represents shifting 2nd @ 5600 (387 product) which results in 3A (3rd @ 4430 and 365 product).
2B represents shifting 2nd @ 5750 (372 product) which results in 3B (3rd @ 4550 and 361 product).
2C represents shifting 2nd @ 5900 (354 product) which results in 3C (3rd @ 4700 and 354 product).
From the graph the optimum 2nd to 3rd shift point is 5900 rpm because the product for 2nd has decreased to the product for 3rd.

Interpreting the plots:
3A represents shifting 3rd @ 6000 (272 product) which results in 4A (4th @ 4060 and 257 product).
3B represents shifting 3rd @ 6200 (253 product) which results in 4B (4th @ 4240 and 252 product).
From the graph the optimum 3rd to 4th shift point is greater than 6200 rpm because the product for 3rd has not decreased to the product for 4th. This is due to the wide spacing between 3rd and 4th. However, the rev limiter is set to 6250 and the engine may rev at least 50 rpm during the upshift. I prefer to set the shift point at 6200 rpm instead of raising the rev limiter.

Interpreting the plots:
4A represents shifting 4th @ 5750 (199 product) which results in 5A (5th @ 4280 and 188 product).
4B represents shifting 4th @ 6000 (186 product) which results in 5B (5th @ 4470 and 187 product).
From the graph the optimum 4th to 5th shift point is 6000 rpm because the product for 4th has decreased to less than the product for 5th.
With my current axle ratio (3.73:1) I don't anticipate ever experiencing a WOT 4th to 5th shift (even on the dyno) since the vehicle speed would be 137 mph and my tires are only rated for 118 mph.

Comparing the stock values to my torque multiplication method:
------- Upshift ------------- Stock -- Revised
Trans WOT Shift RPM 12 -- 5500 ---- 6000
Trans WOT Shift RPM 23 -- 5500 ---- 5900
Trans WOT Shift RPM 34 -- 5800 ---- 6200
Trans WOT Shift RPM 45 -- 5800 ---- 6000

Comparing the NA baseline dyno plot to the initial FI dyno plot illustrates that the FI torque decreases with rpm at a lower rate (smaller slope) than the NA torque.

I think that supports an advantage to delayed upshifts which adds some credibility to my torque multiplication method. It will be interesting to compare different FI shift points on the dyno in time to speed tests.

 

[2000StreetRod]

theoretical vs actual mph vs rpm

I compared the previously posted mph vs rpm for 1st gear with a datalog and found they do not match. Using my rear axle ratio, 1st gear ratio and the manufacturer's revs per mile for my tires the vehicle speed should be 56 mph at 6000 rpm. However, the datalog shows 51 mph at 6000 rpm - a difference of 10%. The 1st gear ratio in the tune is 2.473 vs 2.47 that I used so that is not the source of the difference. However, the tune rear axle stock values are: Axle Ratio of 3.500; Final Drive Ratio of 3.770; and Tire Revs Per Mile of 800. For my 2000 Sport the rear axle sensor signal goes to the 4WABS module that computes the vehicle speed and then provides it to the PCM. I suspect that the only rear axle value that matters is the Tire Revs Per Mile. According to BF Goodrich my tire revs per mile are 721 which explains the 10% difference. I changed the value in the latest version of my tune but also changed the axle and final drive ratio to 3.73.

I also noticed in the datalog that the 1st to 2nd WOT upshift occurred at 5721 rpm and 49.5 mph even though the Trans WOT Shift 12 was set at 5850 rpm and Trans Shift Schedule 12 for TP > 700 was set at 66 mph. It's my understanding that the shift will occur as soon as one of the conditions is satisfied - in this case the engine rpm. Perhaps the PCM assumes a certain amount of time to accomplish the shift and starts early enough that the shift will be complete at the specified engine speed and my Sport shifts quicker than assumed. I've read that shifting based on vehicle speed is more accurate than engine speed so I'll try that. I'll set the engine speed 150 rpm greater than desired and the vehicle speed according to the desired engine speed. The rev limiter will prevent the engine speed from exceeding 6250 rpm.

Trans Shift Schedule 12: 56 mph, Trans WOT Shift RPM 12: 6150
Trans Shift Schedule 23: 73 mph, Trans WOT Shift RPM 23: 6050
Trans Shift Schedule 34: 96 mph, Trans WOT Shift RPM 34: 6350
Trans Shift Schedule 45: 137 mph, Trans WOT Shift RPM 45: 6150

 

[joecrna]

Don't forget you are assuming a fully locked torque converter. I've never tested anything in our power output range but at higher levels they are seldom truely locked (no RPM difference crank shaft to transmission input shaft). The slip is likely very small but may account for some of the difference. Then as you stated above there is RPM fluctuation as one clutch band is partially engaged and the other is partially disengaging.

 

[RangerOZ]

SCT Forum

How do you get access ?
I've joined the forum but can't read anything
I get no response to the email to the moderator.

 

[RangerOZ]

TP vs TP_counts

Do you know the difference between
TP and TP_Counts ?


The Base Fuel Map has TP while the Fuel Open Loop TP uses TP_Counts

thanx