new ford 5.0 ti-vct

[ashford]

went to the track last weekend and got the best all time run for the car(has had many different engines in it). i had adjusted the clutch right before the event and it messed up my launch routine, first few tries i reved the motor out too far and ended up doing a wheelie .

timeslip.jpeg

[jsmcortina]

The code handles each tooth on each cam differently due to the irregular pattern. If VVT3 is still showing an intermittent cam fault then I have additional debug ideas using the SDcard.

James

[ashford]

i am up for more testing if needed, i was thinking if the vvt angles were adjusted for vvt3 it may help out. currently the min angle is around 330 then the max will be that plus 50.
i can try swapping vvt2 and 4 in software to see how it handles it.

[jsmcortina]

Swapping the VVTs around will almost certainly break things. The code is designed to work with the VVT allocation you have proposed and the resultant phasing to the crank wheel.

James

[ashford]

Swapping the VVTs around will almost certainly break things. The code is designed to work with the VVT allocation you have proposed and the resultant phasing to the crank wheel.

James

yep confirmed that earlier today. although it did sorta work, it was erratic and the filter that is implemented did keep sane readings in the angle gauges. i did this mostly for my own curiosity,i was thinking that being that all cams have the same trigger wheel that just setting the offsets should work but no. i am guessing the code is expecting a pattern to occur within a certain window. just trying to wrap my head around what is happening here.

something i did notice while the actuators were unplugged

alot of vvt1 errors under accel/decel(free reving) it completely goes away when the actuators are plugged in. i have all cams comanded a few degrees off rest at all times so that there is no lag in control (pid loop actualy doing something) there is about 5 degrees retard on the intakes @6000rpm and 2 on the exhaust. i am thinking that the cam angle happens to fall just out of expected range in the code. the exaggerated retard may be due to the hall sensors on the intake cams.

after swapping out to the mustang cams the intake readings were 5 degrees off leading me to believe that ford aligns the cam reluctors to valve events(ivo evc) rather than relative position to crank(icl or ecl) being the mustang cams have 10* or so more duration. i am not sure if this in itself will present problems on other cams or not.

the big question i had is that my vvt3 min angle is 329(f150) and 334 (mustang cams) having this number would this through the code out of whack like in my experiment to swap vvt 2 and 4?

[jsmcortina]

Please post up your current MSQ. I'll see if I can get you some test code this week.

James

[ashford]

here is the msq

[ashford]

bump

[jsmcortina]

Been down with a cold.

James

[jsimmons]

Any updates? Curious to see how its doing

[jsmcortina]

Test code sent by email. EDIT: bounced... not accepting my email address.

James

[Matt Cramer]

Got a question for you, Ashford - do you have any sync loss issues while cranking with the engine hot? Just checking, as I'm hearing that some competing ECUs' wheel decoders are not up to the task.

[ashford]

Got a question for you, Ashford - do you have any sync loss issues while cranking with the engine hot? Just checking, as I'm hearing that some competing ECUs' wheel decoders are not up to the task.

depending on your definition of losing sync. as for crank there is never a lost sync event. i do detect that cam sync happens after the initial fire like a double sync when going from semi to full sequential at startup cold but that is it.

[ashford]

update, sotra: just finished swapping out the intake sensors from hall to vr, installed another ncv1124 in th proto area. vvt3 works beautifully now had to adjust min angle 7 degrees( would need -2 but 358 works) drift on the vr is 4 degrees around 5000rpm versus 6 with the hall.

the bad: vvt1 is all messed up now, i think i may have a bad ms3x board, it worked great at idle but had 20 ish so degrees retard with rpm. different capture edge made for vvt1 angles all over the place, no vvt1 errors though. playing with hysterisis did not change detected angle drastically(1-2 degrees) so i do not believe i am triggering off the wrong edge. however i installed the pullup jumper and adjusted the zero point so it picks up and the retard closely matches vvt3, min angle had to be changed @ 40 degrees and works well above 1000rpm with no errors but errors at idle.

i still have an open channel on the ncv1124 i can utilize, but i have to research what to do to the ms3x to disable the vr output. this also brings to mind that maybe the vr sensors are wired wrong, but they are wired as per oem, perhaps the factory has a positve going zero crossing, i will have to spend more time with the scope to find out.

[ashford]

i think i need help here james, things are making no sense to me, vvt1 is retarding about 15 degrees @5000rpm and scopes look good

i have a scope log of both the ms3xcam in and vvt3, both have scope a as the raw in and scope b as what is seen on th j2 header to the processor, they look the same but do not act the same. all settings for capture edge are on rising, with the exception of the vvt1 in on the vvt page it is set for falling but is greyed out and main cam is set to rising( conflict here?)

here is the ms3x cam measurement i took were of rise and falling delay where input was at 0 crossing and square wave was near 2.5v on the slope
idle:
rise time delay .173ms or about .9*
falling time delay .195ms or about 1*

@5200rpm 25ms for 720* rotation
rise time delay .179ms or about 5*
fallin delay .049ms or about 1.4*
screen shot of vvt1 on the ms3x conditioner

ms3x5000rpm.JPG

next is vvt3 on the ncv1124
idle:
rise delay .439ms or 2.18*
falling delay .391ms or 1.94*delay

@5500rpm 24.5ms for 720* rotation
rise delay .162ms or 4.7*
falling delay .076ms or 2.2*

vvt35500rpm.JPG

oh if needed i can put one of the old cams in the lathe with a degree wheel to get more accurate readings if you want

[ashford]

here is a screen shot of a log reving with the phasers unplugged

[ashford]

any thoughts on this guys?

i did the pullup and voltage offset adjustment again the scope looked like crap but worked, vvt1 errors at low rpm <700 probably one of the smaller waves is not triggering.

my initial is that there is a voltage offset on the ground, but it is not the case.

how about swapping vr polarity then triggering off the falling edge?

[ashford]

i tried it on the spare ncv1124 channel and it works well!?!?! i have no clue why but the ms3x vr conditioner shifts back the signal, it makes no sense vr scoped look exactly the same.

i still have vvt1 sensor wired backwards and noticed @3000 rpm there is less retard vvt1 shows -1.2 or so and vvt3 shows -2.5 and idle shows the same at 1.5. so would it be a good idea to invert vvt3 as well?

on a side note, using the hall sensors is probably not a bad idea if the factory wiring is not used, vvt2 and 4 have a twisted pair throughout and seem to be ok for me but the hall sensors share a ground and power source and a single unshielded wire that follows the factory harness from the sensor in the rear of the engine forward near the coils to the pcm plug. it is a relatively short distance on the oem since the computer is next to the engine but on my car there is about another 10 feet or so of wiring after the oe plug. if shielded wire is used outside of the main harness this would solve the hall sensor issue i had. however they lag a bit more than the vr sensors

[ashford]

james can you give me a screen shot of this section for vvt1 and 3 on the current beta?

Code: Select all

 /* Special handling for Coyote, bring uneven teeth into usable range */
                            if (x == 0) {/* right intake cam */
                                if (t < 36) {
                                    /* Two groups of 2 */
                                    if (t2 == 1) {
                                        ang -= 3170;
                                    } else if (t2 == 2) {
                                        ang -= 2570;
                                    } else if (t2 == 3) {
                                        ang -= 1370;
                                    } else if (t2 == 4) {
                                        ang -= 770;
                                    }
                                } else {
                                    /* Group of 3 */
                                    if (t2 == 1) {
                                        ang -= 6170;
                                    } else if (t2 == 2) {
                                        ang -= 5270;
                                    } else {
                                        ang -= 4370;
                                        /* a3 */
                                    }

[jsmcortina]

Perhaps try removing C13 on the MS3X card. It is a 0.22u cap which could lead to a phase shift.

James