Yet another mapped points question

Unfortunately the P and I gains aren’t defined in the cal, though I wish they were. Only the max I term is defined and that’s only the intake cam iirc. The time to desired angle and dead-band don’t appear to be used in the logic. The feed-forward tables work but that only gives it a head start in correcting the error. If you try to do too much with feed-forward it will cause solenoid oscillation and result in the dreaded vct-disabled. I spent a lot of time tweaking the FF table to get it quicker response to error without oscillation. But when the dc drops to zero and the cam still drifts retarded 30 deg, there is absolutely nothing left for the control system to do.

Quote:

---

Originally Posted by engineermike

Ah, I see the disconnect. I was referring to the oil actuated phaser on the exhaust cam of the gen3. They are much simpler but the control oil has to follow a more tortuous path from the solenoid valve to the actuator.

Robcat030, is your issue on cam torque or oil pressure actuated phasers?

---

I'm not sure I understand the question. The symptom is the phasers that are moved by the oil pressure solenoids. The cause appears to be unknown since it seems like my proposed explanation just isn't possible. I'm not an engineer or expert in hydraulics at all, I just know the issue I saw and made my assumptions from there. I'm almost certain the timing was not slipping, and I'm certain there are no phaser locks.

In the instance of the solenoid being at either 100% or 0% duty cycle preventing the oil from escaping that side, doesn't that make the solenoid effectively useless at either position? (trying to force hydraulic pressure on one side while the other is unable to escape doesn't sound like it would be very successful). And if so, then I should be able to command say 1% DC and leverage the full (and seemingly excessive) hydraulic force required to keep the exhaust cam as far advanced as mechanically possible?

Robcat030 I was asking if the phaser you are having trouble with is cam torque actuated (CTA) or oil pressure actuated (OPA) design. cta has faster response but opa can impart more torsion, so opa is typically used on cams that drive a hpfp. The opa phaser solenoid at 0% sends oil to the advance side of the vane and relieves the retard side to the oil drain, so it doesn’t trap oil anywhere.

I understand now. It's oil pressure actuated

Quote:

---

Originally Posted by RobCat030

I'm not sure I understand the question. The symptom is the phasers that are moved by the oil pressure solenoids. The cause appears to be unknown since it seems like my proposed explanation just isn't possible. I'm not an engineer or expert in hydraulics at all, I just know the issue I saw and made my assumptions from there. I'm almost certain the timing was not slipping, and I'm certain there are no phaser locks.

In the instance of the solenoid being at either 100% or 0% duty cycle preventing the oil from escaping that side, doesn't that make the solenoid effectively useless at either position? (trying to force hydraulic pressure on one side while the other is unable to escape doesn't sound like it would be very successful). And if so, then I should be able to command say 1% DC and leverage the full (and seemingly excessive) hydraulic force required to keep the exhaust cam as far advanced as mechanically possible?

---

It SHOULDNT trap oil pressure on one side but something may be causing that once it phases to 3-5*.
It shouldn't develope a leak on the pressurized side below the 3-5* range either...

You can't command a constant 1% duty cycle. That would be saying move one way constantly, which is impossible to setup in the calibration. You can command 3-5* and if the cam achieved that it should go back to the 37-39% holding duty cycle. The main thing to me in that log is even on decel it's not reaching 0* phase angle. If it can't do that, any movement of the cam when commanded angles between 0-3 are involved mean nothing, other than the phaser is not working properly.

Borg Warner has YouTube videos if those images don't make sense.
AED posted video last month of a cam phaser taken apart, but didn't really explain its physical details very much.

In the exhaust cam phaser there are three vanes, creating six cavities. Two of them I would call slave vanes/ cavities they just allow more oil volume for control, and don't limit the phase range. The main vane is what houses all the check valves and passageways to have it phase/hold/ return to it's base position.
CTA- TA is just coiled springs to assist/ work against, not be the primary control. There's usually a portion of the phaser range they do very little in.
All phasers are primarily oil hydraulic controlled, pretty much all the same way. Some do more/less check valves or more/less passageways, but the general concept of control remains the same. Oil filling or draining from two cavities controlling pressure on either side of a number of vanes. If you think about the way one of the "slave vanes" work, ignoring the complexity of the rest, it's very easy to understand the control and potential failures.

Holding the cams position should be like a severly hydrolocked situation, so seeing a cam phase the opposite way from commanded, because of something outside of the phaser, just really doesn't sound right. The cam would need a crazy amount of counter rotational force (you would feel this driving the car), The engine would need to accelerate the cam sprocket at an incredible rate, or the most likely the phaser would have to have extremely poor control because the fluid is leaking when it shouldn't or being trapped when it shouldn't be or its getting aerated like crazy.

Hey Murfie, it sounds like you're still describing the workings of of CTA phasers.

The OPA phasers are surprisingly simple.

In the attached pic you can see the solenoid valve. This is in the cylinder head and only controls oil too and from the actuator. The oil enters the center slot on the side of the solenoid. As the spool valve rises or falls, oil is directed to the upper or lower slot and relieves it from the opposite passage. In the attached pic, the spool valve is below the neutral state. You can see that it opens a passage from the middle to the lower slot, which feeds oil, and it simultaneously opens up the upper slot to the top, which relieves to the crankcase.

Attachment 130609

The sketchy part, in my opinion, is how the oil makes it from the solenoid valve to the phaser actuator. This pic shows the oil holes that allow oil from the stationary passages in the head into the rotating passages in the cam. There is no positive seal from the advance holes to the retard holes, nor from either set to the crankcase. You're just relying on the fluid losses through close clearances to keep the control oil passages separated.

Attachment 130610

And finally you have the actuator itself. The oil holes above are barrel drilled and connect to the various chambers in the actuator. There are seals on the tips, but not the base or the sides. Once again, you're relying on close clearance and fluid losses to prevent leakage from the high and low pressure chambers.

Attachment 130611

As you can see there are no check valves and no way to hydraulically trap or lock oil in to totally prevent phaser movement from a known good location, though I wish there was.

Quote:

---

Originally Posted by murfie

It SHOULDNT trap oil pressure on one side but something may be causing that once it phases to 3-5*.
It shouldn't develope a leak on the pressurized side below the 3-5* range either...

You can't command a constant 1% duty cycle. That would be saying move one way constantly, which is impossible to setup in the calibration. You can command 3-5* and if the cam achieved that it should go back to the 37-39% holding duty cycle. The main thing to me in that log is even on decel it's not reaching 0* phase angle. If it can't do that, any movement of the cam when commanded angles between 0-3 are involved mean nothing, other than the phaser is not working properly.

Borg Warner has YouTube videos if those images don't make sense.
AED posted video last month of a cam phaser taken apart, but didn't really explain its physical details very much.

In the exhaust cam phaser there are three vanes, creating six cavities. Two of them I would call slave vanes/ cavities they just allow more oil volume for control, and don't limit the phase range. The main vane is what houses all the check valves and passageways to have it phase/hold/ return to it's base position.
CTA- TA is just coiled springs to assist/ work against, not be the primary control. There's usually a portion of the phaser range they do very little in.
All phasers are primarily oil hydraulic controlled, pretty much all the same way. Some do more/less check valves or more/less passageways, but the general concept of control remains the same. Oil filling or draining from two cavities controlling pressure on either side of a number of vanes. If you think about the way one of the "slave vanes" work, ignoring the complexity of the rest, it's very easy to understand the control and potential failures.

Holding the cams position should be like a severly hydrolocked situation, so seeing a cam phase the opposite way from commanded, because of something outside of the phaser, just really doesn't sound right. The cam would need a crazy amount of counter rotational force (you would feel this driving the car), The engine would need to accelerate the cam sprocket at an incredible rate, or the most likely the phaser would have to have extremely poor control because the fluid is leaking when it shouldn't or being trapped when it shouldn't be or its getting aerated like crazy.

---

I'm less concerned about the 0-3* movement in that log. It's the sudden phasing to almost 10* above 7000RPM that causes issues. If I slipped the exhaust cam by 10*, it wouldn't be able to return to 3*. I agree with Mike that passageway in the cam isn't sealed well. It's not much different from the other cam bearings, I'm unsure if this is intentional or a bad design.

The counter-rotational force is the engine rotation and torque. The exhaust cam is falling behind in the opposite direction, the intake cam does not.

Here's an example of oil temp/viscosity effects on phaser control. These 3 snips were from the same drive with the oil temp getting progressively hotter, which of course lowers viscosity:

Des is desired, Ang is actual.

Pull 1:
Attachment 130616
Pull 3:
Attachment 130618
Pull 4:
Attachment 130617

You can see in each successive pull that the error gets worse and worse as the viscosity drops. The phaser solenoid is appropriately going to 0%, though I did massage the FF tables to get it to 0% sooner. And you can see in spite of the 0% DC, the actual angle falls to 40+ deg even though commanded is a little over 10. The single best improvement I made to address this is increasing oil viscosity, but there is still more work to do in this area.

Quote:

---

Originally Posted by engineermike

Here's an example of oil temp/viscosity effects on phaser control. These 3 snips were from the same drive with the oil temp getting progressively hotter, which of course lowers viscosity:

Des is desired, Ang is actual.

Pull 1:
Attachment 130616
Pull 3:
Attachment 130618
Pull 4:
Attachment 130617

You can see in each successive pull that the error gets worse and worse as the viscosity drops. The phaser solenoid is appropriately going to 0%, though I did massage the FF tables to get it to 0% sooner. And you can see in spite of the 0% DC, the actual angle falls to 40+ deg even though commanded is a little over 10. The single best improvement I made to address this is increasing oil viscosity, but there is still more work to do in this area.

---

The fluid enters expanding the cavity or exits shrinking the cavity through the same hole. That hole is at the base right next to the either side of the vane. It leads down to the center where it is either connected to the supply pressure passage or pressure relief passage based on the solenoid position.
The cavities and vanes are hydralic in nature, they should not be leaking like sieves. A fluid filled cavity with only a small hole and passageway to let fluid out is pretty hydrolocked to me, and should be pretty easy to control fluid in and out of.

In each one of those logs the DC should have went to 0 in an attempt to get the cam to the desired position durring the pull. The first one holding 15, shouldn't have mattered, its not at 10 and hasnt been for long enough for the desired rate to max out.
before all the pulls it looks like DC is way to high and not at a holding position ~50%DC(but i think the gen 3 is more like the eco boost and should target ~40%) even though cams are at desired. So your control is way off from where it should be.
Then after the pulls, it goes advanced very fast overshooting desired position,like there was more air in the cavities than oil, and doesn't seem to put much effort in tring to correct that.
IDK seems like you have a very out of control phaser or two(supply issue), have you looked at bank vs bank?

It hardly takes any change in DC to have the exhaust phaser moving very fast.

Attachment 130664

Car was cold, I can take a log later when its very heat soaked to compare. Oil change is over due, motorcraft 5w20.
Attachment 130666

Car just warmed up

Attachment 130670

Car driven for 5 minutes up hill oil should be at temp, I wish we could log oil temp.

Attachment 130669

Quote:

---

Originally Posted by murfie

The fluid enters expanding the cavity or exits shrinking the cavity through the same hole. That hole is at the base right next to the either side of the vane. It leads down to the center where it is either connected to the supply pressure passage or pressure relief passage based on the solenoid position. The cavities and vanes are hydralic in nature, they should not be leaking like sieves. A fluid filled cavity with only a small hole and passageway to let fluid out is pretty hydrolocked to me, and should be pretty easy to control fluid in and out of.

---

Correct, but I THINK the bulk of the issue that RobCat030 and I are observing is due to that passageway from the solenoid to the actuator that goes through a common cam journal:

Attachment 130672

One plane of holes is the advance passages and the other plane is retard. You can see there isn't even a laby seal between them, and it's just relying on the close clearance of the bearing to prevent cross-talk of the two passages. If this is where most of the "slip" is coming from, then it would explain why higher viscosity oil would reduce the leakage and the magnitude of the problem.

Quote:

---

Originally Posted by murfie

before all the pulls it looks like DC is way to high and not at a holding position ~50%DC(but i think the gen 3 is more like the eco boost and should target ~40%) even though cams are at desired. So your control is way off from where it should be.

---

There is a reason for that I mentioned earlier. The solenoid spool valve has been trimmed and return spring shimmed to give it more span of control in the "advance" direction. I offset the FF table to match after verifying solenoid valve position vs voltage. When I was done, I was able to verify that it behaves just as the "stock" bank does, except for about a 20% offset in duty cycle.

Quote:

---

Originally Posted by murfie

Then after the pulls, it goes advanced very fast overshooting desired position,like there was more air in the cavities than oil, and doesn't seem to put much effort in tring to correct that.

---

What I'm seeing is about a 3 deg undershoot/overshoot in normal circumstances. It looks like your exhaust cam is doing something similar. 3 deg is within the dead-band. As RobCat030 said, I'm not worried about the 3 deg. It's the 10-30 deg errors that need to be addressed and they only occur in very specific circumstances.

Quote:

---

Originally Posted by murfie

...have you looked at bank vs bank?...It hardly takes any change in DC to have the exhaust phaser moving very fast.

---

Both banks act the same except for the 10++ deg error under heavy load, high rpm, high oil temp, where only Bank 1 (HPFP) bank accrues unacceptable error. Otherwise, it works just as yours does, where the exhaust OPA phasers are much lazier than the intake CTA phasers. You can see in yours the actual (yellow) lagging behind the desired (pink) worse on the exhaust than intake. If we were to log the same data rate, scaled the same for both time and angle, I'm sure they would look the same.

Attachment 130673

In case you want to see the difference between a 30% throttle 2nd gear pull and a 100% throttle 2nd gear pull.
You can see the difference in DC for the bank 1 exhasut cam. I am very tight with the scale on the DC to show details.

Cold my intake phasers hold with 45-50 DC, hot they take in the low-mid 60s.
Cold my exhaust take 42, hot they take 40-38, with the exception of the bank 1 cam at wot, when the HPFP load comes on. Creates about a 6-8% difference between the banks.
Im pretty sure my old oil has a significant effect on the intake phasers and only a slight effect on the exhasut.

30%
Attachment 130749

WOT
Attachment 130750

I personally dont thing shimming and trimming a proportioning valve does anything for actual phase-ability. The hydralics at play in the cavities and on the vanes didn't change so there would be no better control.
You would need a different design to the vanes ,cavitys, and passageway sizes and proportions. I can't say when it would be nessecary, but it would depend on how fast your RPMs can accelerate in 1st gear. Which only slightly depends on torque/ hp, it is also a gearing function, tire traction. Maybe short pulls on a weak dyno it becomes a problem.
In my pics and log, I think I clearly show load/ torque output has little to no effect on phaser control, unless it is actually accelerating the RPMs of the engine at a significant rate.

Quote:

---

Originally Posted by murfie

... Cold my exhaust take 42, hot they take 40-38, with the exception of the bank 1 cam at wot, when the HPFP load comes on. Creates about a 6-8% difference between the banks.

---

In your log you can see that oil temp (viscosity) plays a part, as does gdi load. So when you add a 40% larger gdi pump, raise pressures 20%, and then make 4 back to back pulls then the effect becomes exaggerated.

Quote:

---

Originally Posted by murfie

... I personally dont thing shimming and trimming a proportioning valve does anything for actual phase-ability. The hydralics at play in the cavities and on the vanes didn't change so there would be no better control..

---

We?ve already established that its not hydrolocked so leakage definitely plays a role. If leakage plays a role, then an increase in flow rate will sustain cavity pressure better. Remember that these passages arent static/locked for an opa phaser; there is always oil flowing because the control oil also lubricates the front cam bearing. It leaks by design and if it didnt then the bearing would fail rapidly. I measured the passageways in the solenoid valve and everything else I could, enough to conclude that the valve is a system bottleneck. I also realized improved control after the valve was modified so that part isn?t in question.

Quote:

---

Originally Posted by murfie

... You would need a different design to the vanes ,cavitys, and passageway sizes and proportions. I can't say when it would be nessecary, but it would depend on how fast your RPMs can accelerate in 1st gear. Which only slightly depends on torque/ hp, it is also a gearing function, tire traction. Maybe short pulls on a weak dyno it becomes a problem.
In my pics and log, I think I clearly show load/ torque output has little to no effect on phaser control, unless it is actually accelerating the RPMs of the engine at a significant rate.

---

I never believed that rpm rate of change or power level directly affects the control issue. In fact, I only see the issue in the upper gears because that?s when the oil heats up enough to cause it.

To summarize:
- Modifying the solenoid valve measurably incrementally helped.
- Modifying the calibration helped incrementally.
- Increasing oil hot viscosity helped significantly, so long as the oil retains it.

I know that mechanically locking out the phaser would definitely prevent it but like RobCat030, that is a last resort. I was hoping someone might have some idea or solution short of locking them out that I could try.

Someone was explaioning how the oil pressure was controlled, ignore the red box.

Note the seperate oil feeds for bearings and phaser control, and the arrows that point in both directions for the cam phasers. Yellow is high pressure, blue is low return pressue.

Attachment 130761

The front cam journal on a Gen 3 exhaust cam only has 2 oil feed holes. They are both vct control oil passages, one for advance and one for retard. The solenoid is designed such that control may be evenly split between advance and retard but it is never fully shut off due to lubrication needs. There is no dedicated lubrication oil feed to this particular journal. I have a gen 3 head apart in front of me so I can post pics if need be.

Quote:

---

Originally Posted by engineermike

The front cam journal on a Gen 3 exhaust cam only has 2 oil feed holes. They are both vct control oil passages, one for advance and one for retard. The solenoid is designed such that control may be evenly split between advance and retard but it is never fully shut off due to lubrication needs. There is no dedicated lubrication oil feed to this particular journal. I have a gen 3 head apart in front of me so I can post pics if need be.

---

Fine don't accept that there are oil passageways for lubrication, and separate oil passageways for hydraulic control.

I don’t accept that because it’s not true. I’m looking at one now. There are just two ports in the bearing and they both go directly to the solenoid valve. Like I said, I can post pics of what’s in front of me if anyone wants to see. I do think this is part of the control issue though because there is guaranteed leakage and cross-talk.

why do people still read the cookbook. so many better sources out there. Murfie - write a book.

"I have not seen the normal variability in cam phaser control amount to anything meaningful in performance. When they fail or are no where near what is commanded that changes things. "

this is correct - don't chase ghosts. Mechanical is easy to see. When I didn't understand the ECU OS/ software I tried to find mechanical problems everywhere. Car was fine.

As far as snap lines - define your likely cam angle transitions and build your map - it increases accuracy, you don't want to see logs with 10 percent weight across 10 MP. hopefully dialed in MP separately.

I have the scales set for the 6th gear- 60-130 pull. Only real difference is time zoom is 10 seconds for 2nd gear pull and 30 seconds for 6th gear pull:

2nd gear 20-60mph pull that lasted 4 seconds. DC drops about 6% with increase in RPM. Bank to bank difference is 6-7%. degree error is between 1-2*.

Attachment 130798

VS

6th gear 60-130mph pull that lasted 14 seconds. DC drops about 6% with increase in RPM. Bank to bank difference is 6-7%. Degree error is between 0-1*.

Attachment 130799

Confirms what I said, force on the phaser that causes error, is related to the RPM acceleration rate. Even with over due oil, thats probably 1/2 quart(more?) low. If you really care about a 1* difference.

Im going to put 0-16w in it see if it makes a huge difference from 5-20w.

Quote:

---

Originally Posted by murfie

I have the scales set for the 6th gear- 60-130 pull. Only real difference is time zoom is 10 seconds for 2nd gear pull and 30 seconds for 6th gear pull:

2nd gear 20-60mph pull that lasted 4 seconds. DC drops about 6% with increase in RPM. Bank to bank difference is 6-7%. degree error is between 1-2*.

VS

6th gear 60-130mph pull that lasted 14 seconds. DC drops about 6% with increase in RPM. Bank to bank difference is 6-7%. Degree error is between 0-1*.


Confirms what I said, force on the phaser that causes error, is related to the RPM acceleration rate. Even with over due oil, thats probably 1/2 quart(more?) low.

---

Did you change the deadband before doing these tests? Looks like the stock deadband for the exhaust cam is set at 3 deg. At 3 deg, the logic doesn't even consider 1-2 deg to be enough error to try to correct. However, I haven't firmly established that the deadband defined in VCM Editor is used in the logic or what it considers to be High and Low rpm. I think for a true test you'd need to set the deadband to 0 and confirm the logic uses it. I think to confirm it's used, you'd have to disable the feed-forward control because the FF could mask a large deadband.

Quote:

---

Originally Posted by murfie

If you really care about a 1* difference.

---

RobCat030 and I are concerned more about the 10-40 deg error at high rpm than the 1 or 2 deg here and there.

Quote:

---

Originally Posted by murfie

Im going to put 0-16w in it see if it makes a huge difference from 5-20w.

---

Seems like it would be interesting, but I'm not sure you can learn much from this test. For starters, there is so much overlap in the viscosity range for 16 vs 20 grade that most 20 grades actually fall into the range for 16 as well. Add to that the likelihood that your old 20 has lost some viscosity, and there's a strong chance that the fresh 16 is the same or possibly higher viscosity than used 20. The only way this test could yield usable results is if you do a UOA of the old 20 and VOA of the new 16 to determine the actual "hot" viscosity difference between the two. Plus, as mentioned, the deadband would need to be reduced to 0 unless you're just looking at solenoid DC vs cam angle.

The below pic is a comparison of fresh vs "50% remaining life" 5W-50. Both of these snips were the 3rd consecutive pull in a 2-3 minute span. The initial 2 pulls had good control in both, as the oil wasn't hot enough to allow performance degradation. Side note: this wasn't the case with 5W-40, as it wouldn't control adequately even at normal oil temp. Anyway, you can see that with fresh oil the control isn't great but the solenoid is still in a usable range. When it hit about 10 deg error the solenoid was still able to respond and bring the error down. With the used oil, the duty cycle is flatlined at 0 and the phaser still drifts back unacceptably; it's out of control.

Attachment 130858