P66 3.4 F-body PCM VSS Settings

[caffeine]

Just curious if anyone knows how the road speed constant and output shaft speed constant variables work in this computer. (As in how do they affect the speedo output). The variables are different from the other ECMs/PCMs I've worked with.

[robertisaar]

MPH is calculated a bit differently from the older stuff, it seems unnecessarily complex to me, but i think it's due to the 4L60E and its lack of a input shaft speed sensor for clutch slip calcs. there is also a lot of stuff that seems to be calculated for the sake of being calcluated.... never actually used. anyways:

road speed constant is a pretty familiar thing to look at and calculate, take tire height, find out how many revolutions it will make in a mile, multiply that by the number of notches on the VSS reluctor and you're good to go if dealing with FWD(since the reluctor is read after the final drive does its job). if RWD, then you need to take axle multiplication into account, just multiply the final gear with the previous value.

some stock values(these may differ slightly between platforms, since it looks like some/most applications got different calibration depending on factory wheel/tire options):

W: 22,847
N/L: 20,377
F auto: 21,964.5
F man: 8,992

going through a typical calc with a W-body example:

205/70R15 is the base tire size on a 95MC, 766.87 revs/mile. i happen to know that the VSS reluctor has 30 teeth on it, so 30X766.87=23,006.1. not exactly what is in the calibration.... then take a look at the optional tire size of 225/60R16. 757.35 revs/mile X 30 = 22,720.5. that doesn't match it either. the difference between the two calc'd values is 285.6 pulses/mile. take half of that and add it to the lower value(to create the average between the two) and you get 22,863.3, which is pretty close to the 22,847 in the calibration. so, this one cal is meant to cover both, with roughly the same amount of error, but in opposite directions. with the base tire, the calc'd speed is .69% higher than reality. with the optional tire, it's .56% lower than reality.

then, you look at the police lumina, where they had a tire unique to that package, along with a unique calibration. 215/65R15 is 775.58 revs/mile, same 30 tooth reluctor, so 23,267.4. calibrations show 22,818 or 22,544.... which causes the speedo to read about 2 and 3% too low.....

from what i can tell, the code takes care of the pulse timing to get a 4000PPM signal with any input rate, doesn't require setting up divisors. but i could be wrong, it's been a while since i've even had a P66 V6 logging. i need to swap in and log some raw values to see just how they interact to create a few values.





also, finally mapped out the rest of the TIC/TOC stuff since i was in the area. both the E and T side processors monitor the 3X, 24X and cam sensors independantly of each other. the E-side also monitors the VATS signal, while the T-side monitors the VSS. interestingly enough, the E-side is responsible for creating the instrument cluster's 4,000 pulse/mile signal for the speedo. there are 3 TOCs left open on the T-side and 2 on the E-side. the only things the factory calibrations use the TOCs for here is generating the 160Hz interrupt for major/minor loop scheduling and on the E-side, 4000 PPM signal. since there isn't any space left open on the T-side(literally a handful of bytes), those TOCs are basically useless, but the E-side having two left open is kind of interesting..... if there are enough outputs left open, could go full banana and let the PCM run a COP/CNP system or any other timing critical function. the VATS TIC channel could also be repurposed.....

[robertisaar]

have been looking into this more since the MC will be back on the road in a short amount of time and I've got some noticeably different winter tires on it.

so, this code is somewhat complicated, but I've worked out enough information to see what's going on.

Vehicle speed calc is entered.

WHEN THE VSS SIGNAL IS LOST:

the transmission's input shaft RPM is assumed using the engine's 3X crank sensor signal.
the current commanded gear's ratio and the previously assumed input shaft RPM are used to create an output shaft RPM value.
the output shaft RPM value is filtered and saved.

WHEN THE VSS SIGNAL IS PRESENT:

the last cycle's output shaft speed is used in a filter along with a new cycle's output shaft RPM value to create a filtered output shaft speed.


this is where they both pick up:

the filtered output shaft speed gets math'd(I really don't want to explain this section of code) with the road speed constant and a couple of values get created and filtered and you end up with a vehicle speed value that you can use.



notice: the output shaft RPM is essential to the vehicle speed calculation. changing the output shaft speed constant or the road speed constant WILL change the vehicle speed calculated. the code for the output shaft speed calculation is done elsewhere and not really interesting for the most part, all it does is take revolutions per minute of the output shaft and turn it into miles per hour via a divide. for whatever reason, it adds some kind of offset value that I'm not seeing ever get changed/written to, so as far as I know, it has a 0 offset being added in. kind of strange, but..... GM. I guess I could always rewrite this section of code to make it simpler, but it seems to work well enough as-is.

with the transaxles, the VSS doesn't directly measure output shaft speed since it measures after the final gear. so, with a transaxle, you have to account for the final gear multiplication(chain ratio included). the manual f-body calibration has some odd things going on though.... I don't know enough about what kind of reluctor is inside the T5 to try and figure that one out.... my calculation show it to be 3.8 pulses per revolution? for that matter, the automatic's OSS constant of 40 pulses/revolution looks perfect, but the road speed constant doesn't make a lot of sense.... showing just under 22,000 when I would expect something just under 100,000? maybe I'm looking at something wrong that applies to transaxles but not transmissions?

I've got all of the relevant text entered into the XDF item, so when I release this next version, it will be at least a little clearer.