Wastegate Modulator - testing and unpowered behaviour?

[OffTrack]

I recently came across a Pierburg service information pdf that contains a short testing procedure for electric switch over valves.
These are the same design as the WGM.

The test procedure is has three short tests to check the operation of the valve, two of which use a hand vacuum pump.

Mine was replaced about 4 years ago seems to operate ok in that I'm not getting any of the classic failure symptoms (yet).

The "tightness" test 1.3 consists of applying a 50kPa differential pressure to the top outlet with power off and checking that the valve holds pressure.
I'm seeing the pressure drop to zero over 10-15 seconds.

If I apply pressure to the inlet (at the bottom) and cap both the waste gate and ambient pressure outlets the value sits on 50kPa for 2-3 minutes. As soon as one of the outlets is uncapped pressure starts to drop.

Assuming the service information is applicable to the Td5's WGM this seems to indicate that the valve is beginning to fail.

Before I replace, can anyone verify the correct behaviour when pressure testing the WGM?

122599

[sierrafery]

Did you test it with ignition on?.... be aware that the modulator is working with pwm command on the earth path to open and gets 12V ignition live to become "active"(which means closed at the beginning), without ignition on it's like a simple open diaphragm.... it's made so to not get overboost if fuse F2 is blown... sorry i didnt read the document before i posted

[OffTrack]

Did you test it with ignition on?.... be aware that the modulator is working with pwm command on the earth path to open and gets 12V ignition live to become "active"(which means closed at the beginning), without ignition on it's like a simple open diaphragm.... it's made so to not get overboost if fuse F2 is blown... sorry i didnt read the document before i posted

Hi Fery,

As I understand the WGM is an open from port 1 ( turbo to intercooler pipe connection) to port 3 (waste gate actuator) when it is "de-energised".
So when power is off and when power is supplied but earth path not completed by PWM there is flow 1 -> 3.

When the WGM is "energised" by the PWM providing earth, flow is open from port 2 ( turbo intake trunk) <-> port 3 (waste gate actuator).

In other words if the WGM is activated by a PWM signal it prevents boost pressure reaching the waste gate actuator.
With 0% PWM the waste gate actuator gets full boost pressure.

Anyway I may as well buy a new WGM to check how they work out of the box.

cheers
Paul

[OffTrack]

Got a new Pierburg WGM from Roverlord this afternoon.

The WGM has been sitting on the desk with 0.5bar vacuum on Port 2 (WGM -> inlet trunk ) for about 30 minutes and the needle didn't budge.
Checking with 0.5, 1.5 and 2.0 bar pressure applied to Port 1 ( turbo outlet -> WGM ) with Port 3 (WGM -> waste gate actuator) capped there is no sign of pressure loss after 20 minutes (at 2.0 bar).

So the Pierburg test procedure appears to be correct for the TD5 WGM.

[OffTrack]

Posted up some information on how the WGM unit works here:

Wastegate Modulator: Operation | DiscoTD5.com (http://www.discotd5.com/ecu-reverse-engineering/wastegate-modulator-operation)

It's probably a bit long winded but may be of some interest.

[steve_a]

It makes sense as if there is a failure the turbo pressure is sent to the wastegate actuator. Safe by design.
I wondered about the third pipe and was curious why they didn't just terminate to atmosphere, but I guess rather than trying to find a place where it won't allow gak back in it makes sense to use the air intake.
As a further benefit I guess it helps to sharpen the system up as you can go from +X psi to -Y psi on the PWM cycle, reducing lag.

This also explains why adjusting the actuator rod gets you boost pressure, if the ECU was controlling boost pressure getting to it then you could set the rod to any old value less than actuation and it shouldn't matter as the ECU would basically be opening and closing the waste gate as it reach desired boost and keeping it on desired boost.

[OffTrack]

It makes sense as if there is a failure the turbo pressure is sent to the wastegate actuator. Safe by design.
I wondered about the third pipe and was curious why they didn't just terminate to atmosphere, but I guess rather than trying to find a place where it won't allow gak back in it makes sense to use the air intake.
As a further benefit I guess it helps to sharpen the system up as you can go from +X psi to -Y psi on the PWM cycle, reducing lag.

This also explains why adjusting the actuator rod gets you boost pressure, if the ECU was controlling boost pressure getting to it then you could set the rod to any old value less than actuation and it shouldn't matter as the ECU would basically be opening and closing the waste gate as it reach desired boost and keeping it on desired boost.

I think one of the main goals was to reduce wastegate creep and improve mid-range response. As you say it's reducing lag.

The modulator is dissipating around 6W when powered so it wouldn't be practical to have it on by default.
It gets pretty hot if left on for a few minutes.

I suspect the intake is used because the pressure is below atmospheric under boost - you see this in the drop in the AAP reading.
That should make bleeding boost away from the actuator more effective than dumping to air.

[sierrafery]

Hi, nice work, i have only one comment for a statement of your's(with red) not to contradict you just to say what i found...
it's about that:
So what does RAVE actually say to support the "activate to limit boost" theory?The following sentence is often pointed to as evidence:
When full boost is reached a control signal is sent to the wastegate modulator, and a vacuum is applied to the wastegate valve.
Interestingly this sentence incorrectly states that vacuum rather than boost operates the waste gate valve. This is often excused as a typo.
But lets examine this from a slightly different perspective.
The turbo intake is subject to suction and we know from the AAP sensor readings that the intake pressure drops below ambient under boost.
We also know from the Pierburg service information that the waste gate actuator is connected to the turbo intake when there the solenoid is energised. Under these conditions the waste gate actuator has a slight vacuum applied.
It's not conclusive either way but there is clearly an error in that statement.it was quite a while but i stripped a modulator in pieces just to see what's in it and my conclusion is that if the statement in RAVE is not a typo it's a complete error cos when let's say the modulator starts opening the internal valve of it works like a kind'a EGR/ILT valve, it should be no effect of suction from port 2 against port 3 so IMO no "vacuum" can reach that, port 2 can be only an ''exit'' for boost from port 1... "vacuum" from port 2 can eventually affect boost from port 1 which might reduce the boost applied to the wastegate valve when the modulator is in a middle position but if there's no restriction in the intake the drop in AAP is almost inexistent 1-2 KPa which hardly can be named "vacuum" cos the boost which is directed to port 2 is much higher from the beginning when the valve toward port 3 is completely closed if you see what i mean

this theory of mine is 100% supporting your last statement here
I suspect the intake is used because the pressure is below atmospheric under boost - you see this in the drop in the AAP reading.
That should make bleeding boost away from the actuator more effective than dumping to air.

So IMO your description(except the red part) is spot on and RAVE is 100% wrong(unfortunately not the only case[wink11])

[OffTrack]

Hi Fery,

I'll remove that reference to vacuum because it doesn't really add much to the discussion.

I may be misunderstanding your post, so apologies if that is the case.

The WGM has two paths for airflow.

I've enhanced the flow annotations in the pierburg information to make this a bit clearer.

First is with the solenoid de-energised, where pressure can flow between port 1 and port 3 in both directions.
That corresponds with the connections from turbo-intercooler pipe to the WGM and from the WGM to wastegate actuator.

122958

Second is when the solenoid is energised by the PWM signal completing the earth path.
In this case pressure flows from the wastegate actuator (Port 3) to the air intake trunk (Port 2).

122957

In both cases port 3 is open, and the solenoid switches between port 1 and port 2 .

cheers
Paul

[sierrafery]

Hi Paul, You might not have misunderstood my post it's just that maybe we approach it differently. AFAIK with energised WGM it's not like in your second scheme but something like this

https://www.aulro.com/afvb/images/imported/2017/05/791.jpg

i mean that when the ignition is turned on the full flow from 1 to 3(from the de-energiseed status when actually boost is not present as the ignition is off) will be redirected entirely to port 2 and at that stage it's no comunication between 2 - 3, then as boost will start rising, at a point the WGM managed by the ECU will start splitting the boost from 1 between 2 and 3 ending up directing it all(not 100% though) to 3 at full load... but there is a stage where boost is present at lower load then port 3 is closed and all goes out through 2 to achieve better throttle response untill boost is high enough ... that's how i understand that but i dont insist it's correct

[OffTrack]

Hi Paul, You might not have misunderstood my post it's just that maybe we approach it differently. AFAIK with energised WGM it's not like in your second scheme but something like this

i mean that when the ignition is turned on the full flow from 1 to 3(from the de-energiseed status when actually boost is not present as the ignition is off) will be redirected entirely to port 2 and at that stage it's no comunication between 2 - 3, then as boost will start rising, at a point the WGM managed by the ECU will start splitting the boost from 1 between 2 and 3 ending up directing it all(not 100% though) to 3 at full load... but there is a stage where boost is present at lower load then port 3 is closed and all goes out through 2 to achieve better throttle response untill boost is high enough ... that's how i understand that but i dont insist it's correct

I'll stick with what I posted.
I've verified that information with testing and it matches Pierburg service information and patent documents.

[sierrafery]

I'm not gonna contradict that just explain me please according to your second drawing do you mean that when the WGM is energised by starting the engine port 1 is completely closed and only 2 -3 are comunicating? ... if yes i can understand that and i stand corrected.

[OffTrack]

I'm not gonna contradict that just explain me please according to your second drawing do you mean that when the WGM is energised by starting the engine port 1 is completely closed and only 2 -3 are comunicating? ... if yes i can understand that and i stand corrected.

Hi Fery,

A solenoid is energised by current flowing through the coil.

As I'm sure you know Current = Voltage / Resistance.
From the data sheet and measurements it's known the solenoid coil has about 28.3 ohm +/-1.5 ohm resistance.

Unpowered you have +ve = 0V, -ve = 0V, so 0V difference, 0V/28.3 = 0 amps.
Ignition on but no PWM +ve = 12V, -ve = 12V, again 0V difference, and 0V/28.3 = 0 amps.
(I can't confirm but it's likely the ECU has an internal pull up to supply V when -ve connection is not earthed).

In both cases the solenoid is de-energised, and the flow is port 1 <-> port 3.

With ignition on, and PWM completing earth path, +ve = 12V, -ve = 0V, so 12V difference. 12V/28.3 = 0.424amps,
In this case the solenoid is energised and the flow is switched from port 2 <-> port 3.

This is what the bitscope capture is showing. The reference lead is attached to 12V and the +ve terminal, the probe is attached to -ve / pin A21 of the ECU.
When the PWM completes the earth path the voltage drops on the -ve terminal, and the voltage difference across the WGM terminals increases to approx 12V. The solenoid energises and the WGM switches.

So starting the engine does not alter the flow. It remains port 1 <-> port 3.
Only when PWM completes the earth path does the flow switch to port 2 <-> port 3.

The operation of the WGM is basically:
With 0% PWM boost flows to the wastegate. At low boost and full boost the waste gate operates as normal. Port1 -> Port3
With mid-range boost where the wastegate can creep open the ECU sends a PWM signal to the WGM to enable pressure to be bled away from the waste gate actuator. Port3 <-> Port 2.

Because the WGM is rapidly switching the connection to the actuator on (3) between intake pressure (2) and boost pressure (1) the actuator sees a blended pressure which depends on PWM %.

cheers
Paul

[sierrafery]

Sir, i'm happy we had this discussion cos i've learned something again.... not about voltages/PWMs cos IMO i undertand that but i have to admit that i was wrong about the way the WGM works from the air flow point of view cos after the last post of mine i had some remorse and made just a simple test to prove myself that i'm right and surprise, i proved myself that i was wrong and you was right :), turned ignition on and used a bycicle pump on the hose which i unhooked from the intercooler(direct feed to port 1), and YES, your second pic is spot on and i apologise for contesting it, port 1 is closed with ignition on without engine running...which after a deeper thinking i realised that it's logical to be so.... i'll not erase the wrong theory from the previous post of mine to let it as a lesson for others to investigate more before making statements... anyway without this "debate" i could have died with a wrong conception in my head so i feel lucky anyway.... please forgive me for contesting your statements, i can just promise that from now on i'll investigate deeply before i'll doubt anytrhing you say
[wink11]

[OffTrack]

Sir, i'm happy we had this discussion cos i've learned something again.... not about voltages/PWMs cos IMO i undertand that but i have to admit that i was wrong about the way the WGM works from the air flow point of view cos after the last post of mine i had some remorse and made just a simple test to prove myself that i'm right and surprise, i proved myself that i was wrong and you was right :), turned ignition on and used a bycicle pump on the hose which i unhooked from the intercooler(direct feed to port 1), and YES, your second pic is spot on and i apologise for contesting it, port 1 is closed with ignition on without engine running...which after a deeper thinking i reasised that it's logical to be so.... i'll not erase the wrong theory from the previous post of mine to let it as a lesson for others to investigate more before making statements... anyway without this "debate" i could have died with a wrong conception in my head so i feel lucky anyway.... please forgive me for contesting your statements, i can just promise that from now on i'll investigate deeply before i'll doubt anytrhing you say
[wink11]

Don't apologise.
I don't accept things presented on the 'net without verifying for myself, and I wouldn't expect anyone to do otherwise.

Both pics are taken from the Pierburg service information and match the information in the 1994 Pierburg patent.

If you have a look at the patent there is a detailed cross sectional view of the switching valve.
Patent EP0633415A1 - Electromagnetic switch valve - Google Patentsuche (http://www.google.ch/patents/EP0633415A1?cl=en)


123048


In the patent description "terminal duct 11" and "channel 11" is same as port 2, "channel 12" is port 1, and "channel 24" is port 3.


Between central portion 17 and rotor core 7 having the one connecting channel 12, a spring 23 is clamped, by means of which the armature 9 is loaded against the valve seat 13 of the other terminal duct 11.The passageway 10 is connected to a third connecting channel 24.Depending on whether the solenoid coil is energized 3 or not, the connecting channel 11 is connected to the connecting channel 24 and connecting channel 12 closed, or of the connection channel 12 to the connecting channel 24 and the other is closed (as shown).

Deciphering patent speak is always entertaining. But at least you know it's going to be precisely worded.

What the patent describes is basically same as the service information.

The armature has valve seats at points 13 and 14. When the coil is de-energised a spring presses the armature toward port 2, causing the valve seat to seal port 2.
When the coil is energised the armature is pulled toward port 1 and the valve seat 14 seals port 1.

It also makes clear the flow. With solenoid energised port 2 is connected to port 3 and port 1 is closed. With solenoid not energised, port 1 is connected to port 3, and port 2 is closed.

Anyway, this labouring the point. [smilebigeye]