S54 dual VANOS VVT

[jsmcortina]

I've got some test code that uses two pairs of outputs.

output 3 = complementary pair to output 1, must be set to opposite direction
output 4 = complementary pair to output 2, must be set to opposite direction

PM me an email address if interested in testing.

James

[AbatelliCristian]

Yes good..... abatellic@libero.it

I think I can test it in the end month days....


very thanks

[AbatelliCristian]

Hi James
"more duty means" VVT1=advance, VVT3=retard?

And in No. Teeth I setting 6 an 8?

[jsmcortina]

That sounds correct.

James

[Cell^]

Hi,

The way S54 vanos is controlled, is one of the two types used in BMW's. First one is the common "family-version": one solenoid per VVT-unit, either on/off (single-vanos) or PWM (dual-vanos). This is used for example in these engines:

M50TU, M52, M52TU, M54, M62TU, N52, N42, N62 etc.

Second type is the "motorsport-type" which is faster: two solenoids per VVT-unit, one solenoid retards and one advances. When both solenoids are off, cam doesn't move. You should not activate both solenoids at the same time, because high-pressure oil pump has limited capacity and this also causes unwanted wear in the system. Solenoids are known to break when they get older even with the original control system. This type is used for example on these engines:

S50 (both B30 and B32, first has only one intake VVT, latter both cams), S52, S54, S62 etc

This is particularly interesting for me as I have S62 engined car

Getting correct patterns for cam-wheels is one thing, but the necessary programming for this motorsport-type control is not available at the moment I believe. There is a clear need, S50 for instance is very popular in many racing series at least here in Finland and there aren't a lot of ECU's to control it properly -> use of vanos delete kits

Attached is the wiring diagram of S54 vanos.

[jsmcortina]

Second type is the "motorsport-type" which is faster: two solenoids per VVT-unit, one solenoid retards and one advances. When both solenoids are off, cam doesn't move. You should not activate both solenoids at the same time, because high-pressure oil pump has limited capacity and this also causes unwanted wear in the system.

Seems there is conflicting information about the double solenoids. What is the source for that information (only operating one at a time.)
What holds the VVT in place when the solenoids aren't energised?

James

[Cell^]

I think it is more misunderstanding than conflicting information.

VVT is held in place by stopping oil flow on the top side. On the shaft side of piston is pressure constantly supplied by HP oil pump, on the top side is the volume which is controlled by solenoids, no other way in or out for the oil. The pressure is the same on both sides at static, so no oil leakage "through" the piston (oops, of course lower on top side in static and can leak) and piston cannot move unless oil on the top side is added or substracted by the solenoids. Oil through the solenoids can move the piston because to area is bigger on top side than on the shaft side, bigger hydraulic force.

Holding one solenoid 50% open means that to keep it in place you have to keep the other one open the same amount and constantly pump oil through.

This topic is about same kind of system: http://www.msextra.com/forums/viewtopic ... 31&t=44750
thokes82 has posted to that thread hydraulic drawing of the system. It's from S50B30 but the system is the same in S54 for one camshaft.

[AbatelliCristian]

When solenoid 1 and solenoid 2 vanos is 0 position in start engine.
For move vvt you open only solenoid 2, but for return in 0 position you close solenoid 2 and open solenoid 1.......
See the figure

[jsmcortina]

OK. The current implementation in MS3 isn't doing that. The two solenoids are the complement of each other.

i.e. if VVT1 is 45% then VVT3 = 55%.

Consider this replacement code.

vvt1_raw = base position commanded by code 0-100%
midpos = middle duty % (typ. 50%)
vvt1_out = duty output to solenoid 1
vvt3_out = duty output to solenoid 3
(1&3 are the pair)

if (vvt1_raw > midpos) {
vvt1_out = (100* (vvt1_raw - midpos)) / (100 - midpos)
vvt2_out = 0;
} else {
vvt2_out = (100* (midpos - vvt1_raw)) / (100 - midpos)
vvt1_out = 0;
}

So with vvt1_raw = 60% and a 50% midpoint, that would give
vvt1_out = 20%
vvt2_out = 0%

and with vvt1_raw = 5% and same 50% midpoint would give:
vvt1_out = 0%
vvt2_out = 90%

Does that sound correct?

James

[Cell^]

If I understood correctly, then no.

Dutycycle doesn't affect the end position, only the speed at which you will move towards the end. For example applying dutycycle of 10% to solenoid 1 (I haven't checked which solenoid retards and which advances, but the idea is this) will cause the cam to move towards retard no matter what position it was and will only stop once maximum retard is achieved unless it is turned off.

Could it work if low frequency PWM would be applied to the solenoid as long as: target - actual > hysteresis ? Target - actual sign (+ or -) would define which solenoid is used out of the two.

[jsmcortina]

If I understood correctly, then no.

Dutycycle doesn't affect the end position, only the speed at which you will move towards the end. For example applying dutycycle of 10% to solenoid 1 (I haven't checked which solenoid retards and which advances, but the idea is this) will cause the cam to move towards retard no matter what position it was and will only stop once maximum retard is achieved unless it is turned off.

That's ok. This is only the output section. The closed-loop PID still determines what output duty is required.

James

[hoveringuy]

Does that sound correct?

James

It sounds correct. I would set the unused solenoid to 0 before assigning the duty cycle to the partner.

I would also argue that duty cycle is is still relevant. The alternative would result in oscillations.

[Cell^]

I would also argue that duty cycle is is still relevant. The alternative would result in oscillations.

That's true, when you come to think of it

[jsmcortina]

Explain?

James

[hoveringuy]

Explain?

James

The sampling wouldn't be fast enough. The S54 uses high pressure oil to drive VANOS, not just 60psi that we're used to. The slew rate is VERY fast. A very high plant frequency needs a high sampling frequency. PWM is effective because it lowers the plant frequency.

[jsmcortina]

How does that impact my plan on controlling the two solenoids?

James

[hoveringuy]

How does that impact my plan on controlling the two solenoids?

James

It doesn't. I think you've got it right. I was commenting to illustrate that the Duty Cycle is still important.

[hassmaschine]

didn't realize so much had happened on this topic!

James, do you have the information required to write a wheel mode for the jimstim? I could use that to verify that the wheel mode on MS3 works the same as the stock S54 DME.

I know it's an 8-1 pattern, but I don't know where the missing tooth should go vs the crank sensor.

[golf-2]

In the other hand,
The S50B30 EURO engine has single vanos in the intake cam , with full range about 40 degrees
this vanos has 2 solenoid : one to advance and one to retard (don't know exactly which valve to retadrd or advance??)
every solenoid has 2 pin one is for 12v and other is from ecu ( i think it pwm ground switching unlike the b32 dual vanos , the switching is 12v)
Is this system works like the 2 pin idle valve ?? ,and any one know which solenoid advance, upper o lower one??

any suggestion ??
thanx

[AbatelliCristian]

are you sure do you have 12v fixed and ground pwm?