I did measure the resistance between pins on the egr valve connector.
assuming pins are numbered
1-2-3 (top row)
4-5-6 (bottom row, closer to mechanical parts of valve)

I'm getting 20.x and 21.x OHms between 1-2, 2-3 and 4-5, 5-6

 

If you remove the battery you often need to do an “idle re-learn” to get the idle stable again. I am not certain if this will solve your problem but it is probably worth doing just to eliminate this as a possibility.

Do a forum search and you should be able to find the exact procedure but from memory you turn off all electrical accessories including internal fans and lights, disconnect the battery for 10 minutes and then after re-connecting the battery, start the car and without touching the accelerator or turning on any electrical accessories, let the car idle for 10 minutes.

After 10 mins turn the car off, wait at least 1 minute and then restart the car, again without touching the accelerator or turning on any electrical accessories and let the car idle for 5 minutes.

Turn off the car and wait 30 seconds.

I had to run this procedure on my daughters 2007 Legacy/Liberty after changing a battery and it resolved the problem straight away.

Seagrass

 

@sub2008

I shone a flashlight on one side and could see just a little light on the other side. . . . When I removed it this time, the valve was stuck open about a 1/4". . . . I closed the valve by prying with a screwdriver and it took a lot more force than I expected.

The EGR valve is controlled by a stepper motor which varies the opening in about 50 very small steps. The EGR valve is normally opened only when the car is being driven. It's not open at idle, nor when the outside temperature is below 5 C (or thereabouts). If the valve remains open, even a small amount, when it shouldn't be open, it will cause roughness and a tendency to stall at idle. If after the engine is allowed to idle, it's turned off, then when the EGR valve is removed, it should be closed.

The P0400 code is related to the EGR valve operation. That OBD test is run when the car is being driven at >50 kph, and the accelerator is released (fuel cut). With the accelerator/throttle released, the EGR valve should be closed. The system checks the manifold absolute pressure (MAP), then opens the EGR valve for a second, then closes the valve. It then compares the MAP readings. If the difference isn't as great as it should be, in two consecutive drive cycles, it triggers the code. (The idea being, that if the valve is stuck, or not moving properly, there won't be as great a change in MAP as there should be.)

The resistance measurements appear to be good. This might be of interest: http://mmc-manuals.ru/manuals/lancer_x/online/Service_Manual/2010/17/html/M117305020078100ENG.HTM

 

plainOM: thanks for that link. that's exactly what I was looking for! and will try to test when it's not -18 in my garage...
the "operation check" says to observe whether the motor shudders, but it doesn't indicate if it should open or close in either scenario (1&3 or 4&6 connected to gnd).

what I gather from your points and the article then, is that there's no "passive rest" position -- it will stay at whatever setting it was at when power removed? (rather than always returning to closed).
also, that the valve should not move "freely" or with little resistance in either direction, since it's connected to a stepper motor.

I thought there was a solenoid that controlled the opening of the egr valve. Am I mistaken?


The other item I'm curious about is whether I have a normal amount of carbon deposits in the ducts that go to/from my egr valve.

 

. . . the "operation check" says to observe whether the motor shudders, but it doesn't indicate if it should open or close in either scenario (1&3 or 4&6 connected to gnd).

The original purpose of linking that article was for the resistance measurements. I noticed that "operation" part and I'm not sure just what it's doing. But Figure 1 in this document might be helpful. I think it shows how the unipole EGR stepper is wired; four of the wires go to the ends of the two coils, and two wires are battery supply. Each of the four wires is connected to "switches" in the ECM, which, by changing which end of which coil is grounded in time, determines which way the rotor turns, and how much. Each switch is closed for only a very short time, in the range of 5 - 10 ms. (See also Figures 5 and 6.)

When, in that "operation" test, the positive is connected to the center and the negative to the two outer connections of each coil, the magnetic fields of that one coil oppose each other. The rotor, I guess, starts to move in one direction, but then moves back. So basically, it just shakes in place. I haven't seen any other suggestions to test a stepper motor this way, so it might be something I would leave until pretty well everything else has been exhausted. It also warns against keeping the power on for any length of time. The original purpose of linking that article was for the resistance measurements.

EDIT: given your result above, the two parts of each coil might be wired so that the fields are in line. When the power is applied, the rotor aligns itself with the field and stays there. That's probably the "shudder" mentioned. The motor won't continue to rotate, so no movement of the valve. As we see in figures 5 and 6 of the linked paper, the power to the coils is switched, rapidly, in sequence, to cause the rotor to rotate.

what I gather from your points and the article then, is that there's no "passive rest" position -- it will stay at whatever setting it was at when power removed? (rather than always returning to closed).

Although I haven't found any information specific to the valves in our Subarus, I believe that is typical of the stepper motor type of EGR valve. However, as I noted earlier, at idle the valve has to be closed, so if after the engine idles, it's shut off, and the valve is found to be open when it's removed, that's not correct. The ECM would have normally driven the valve back to closed.

also, that the valve should not move "freely" or with little resistance in either direction, since it's connected to a stepper motor.

That's my understanding of a typical construction. Some time ago I found some design diagrams. At a very elementary level, the rotating stepper motor has a threaded output shaft. There's a bolt that's threaded onto the shaft that's connected to a rod going to the valve itself. The bolt is mounted in a holder that prevents it from turning but doesn't prevent it from moving up and down the threaded shaft when the shaft turns. Consequently, as the stepper motor moves step by step, the bolt moves up or down and this opens or closes the valve. The use of the threaded shaft means the shaft has to turn more than the distance the bolt will move linearly along it. This provides for very fine control of the valve opening.

If you think of a bolt with a nut on it, when the bolt is turned, the nut moves up and down the bolt's threads. But if the bolt is held and the nut is pushed up or down the threaded section, the bolt is not likely to turn, and allow the bolt move along it. It's not so much that a stepper motor is used, but that the motor uses the threaded shaft to move the valve via the bolt arrangement. It's that which means the valve won't move "freely" in either direction. (This is only to explain the system; the actual valve mechanism might be somewhat different in actual design, but the approach is probably typical.)

[Correction: The Subaru valve is spring loaded, and if no power is being applied to the motor coils, the spring will move the valve to the closed position. More in this thread.]

I thought there was a solenoid that controlled the opening of the egr valve. Am I mistaken?

There indeed are engines with EGR valves that use a solenoid driven by the ECM. It uses a duty-cycle drive signal, where the higher the % duty cycle the more exhaust gas is allowed through. Like the old vacuum-operated valves, these are spring-loaded, so that the valve closes (or at least, should close) when the solenoid drive is removed.

The other item I'm curious about is whether I have a normal amount of carbon deposits in the ducts that go to/from my egr valve.

I'd have to leave that to others who have removed the valves on a number of cars and have a better idea of what might be expected of a 10 year old car.

But given the way the valve works, any deposits that might limit the movement of the valve rod, or the bolt along the threaded shaft, could prevent the valve from operating properly.

At some point I'll see if I can get hold of a valve and test and disassemble it to get a better idea of how it works.

 

So I tried to follow their bench-top test procedure.
I used 10 NiMH batteries as a powersource ~14v

I connected pin 2 to the +'ve; and "simultaneously" connected pins 1 and 3 to the -'ve.
I hear a single cluck/click, like the sound of a relay closing and that's it.

I get the same result when I test pin 5 +'ve and 4&6 -'ve.

Should I expect the valve to move at all under these conditions?

 

The carbon that is in the valve looks pretty normal. If you were having an oil problem; like bad rings, then the insides of the valve would be wet, ikky and carbony. The dry light layer you have does not look unusual.

You should see at least some sort of movement in the valve one way or another when you pulse the coil. From what I have seen of over valves (not Subaru) you can detach the coil from the actual valve and make sure that the valve can fully seat and close off exhaust flow and also move without restriction to being fully open without binding. If it binds up then the valve stem may be gummed up or scored.

Part of the wiring you are seeing is the coil, but also feedback on valve position so the engine management system knows that the valve has done something when commanded. You can have a valve that moves just fine but doesn't give feedback to the computer, so the computer will continue to send commands... moving the valve even further.. and further... and further. Then the valve is wide open but the computer thinks it hasn't moved at all. (see the problem there?).

Feedback is part of how a loop control system works (loop, meaning process-loop, every commanded action should have feedback saying that the action really did happen).

 

plainOM: thank you for that in depth explanation of the stepper motor's working.
So the analogy would be like 2 guys sawing a log -- they each take turn powering the saw blade.
So to test the motor extending and retracting this valve, i'd have to alternate power back and forth between the 1-2-3 side and 4-5-6 side; probably faster than I could do it manually.


Tishers: when I removed it for cleaning the first time, the lower chamber did look a bit wet and oily.


is the next step to connect the debug connector and try the freessm egr valve test?


my current problem symptoms are:
when warm: hard to start (sometimes have to floor the gas pedal)
when warm: rough idle and sometimes dies.
feels a bit sluggish at acceleration, but seems to drive ok above ~1200 rpm.

when its experiencing the rough idle, if I keep it in low gears and drive >3krpm for a bit (minutes, or sometimes just between stop signs), it will be ok at the next few stop signs/minutes of driving. I'm usually at destination by this point, and it gets parked for a while.

 

. . . So to test the motor extending and retracting this valve, i'd have to alternate power back and forth between the 1-2-3 side and 4-5-6 side; probably faster than I could do it manually.

Yes, that's the way I see it at this point.

is the next step to connect the debug connector and try the freessm egr valve test?

Unfortunately that doesn't seem to work with these stepper motor valves. I believe this was the conclusion in https://www.subaruoutback.org/forums/109-gen-3-2005-2009/475881-egr-valve-alive.html

Incidentally, if you have FreeSSM (or better still, Romraider), the number of "exhaust gas recirculation steps" can be tracked. Normally, again, at idle, it should be zero, and when very cold out, it will remain there when driving. But otherwise, as the vehicle and engine speed increase, the steps will go off zero and will vary up to about 45. Note: I believe the displayed "steps" is the ECM "target" rather than an actual, measured, position. Also, what each "step" should amount to in terms of visible movement of the valve isn't clear.

 

https://www.subaruoutback.org/forums/66-problems-maintenance/494523-egr-can-go-bad-symptoms.html might be of interest. @heyhar 's 2004 has what looks like the same EGR valve (according to on-line catalogs), and it was separated for part of the testing. Interesting. Makes me wonder how. Seems there's no complicated linkage, and it's described as a solenoid, not a step motor.

(Note the reference at the end to replacing the MAP sensor. . . )

Is it a solenoid, not a step motor? Is there no special linkage between the solenoid/motor and the valve stem itself? Are the resistance measurements that were made here not what I think they are, i.e., measuring the windings of a step motor? Makes me wonder if I have it wrong . . .

[Correction: Concluded it is a stepper motor. More in this thread.]

EDIT: I had a PM from heyhar indicating that the use of "solenoid" was not meant to be definitive -- it could well be a step motor.

Local auto recycler has apparently closed shop, so I'll have to find another where I might pick up a valve, or two to look into, literally. But I'm pretty confident at this point it's a step motor drive.

 

. . . Local auto recycler has apparently closed shop, so I'll have to find another where I might pick up a valve, or two to look into, literally. But I'm pretty confident at this point it's a step motor drive.

I was able to get two Subaru EGR valves at a u-pull. Both are from 2005-9 Legacy or Outback. (I didn't check the cars more closely. It was -20 C and everything was under about two feet of snow and ice). One valve has now been taken apart to see how it's made and to run some electrical operating tests.

I'll try to put together a post (or a separate thread) with photos, explanation of how it works (mechanically) and measurement [see EDIT below], but in the meantime:

The drive is definitely a stepper motor; the unipolar type described in https://www.infineon.com/dgdl/Infin...r_motor_with_TLE8110ED-AN-v01_00-EN.pdf?fileId=5546d46261ff5777016209fd50284300

The motor can be made to operate by applying 12 V sequentially to the coils via the terminals in the six-terminal connector. The above paper, and http://mmc-autoelectric.org.ua/manuals/eur/lancerx/2010/17/html/M217100030043400ENG.HTM are helpful in this regard. (It will take a lot of momentary contacts with the terminals, in a specific order, to rotate the motor rotor enough to see movement in the valve.)

The spring does return the valve to the closed position when there's no power to any coil.

EDIT: Details of the valve's construction, how it works, and possible ways to test it, are now in https://www.subaruoutback.org/forum...intenance/505085-subaru-electronic-egr-valve-how-s-made-works-might-tested.html

 

Thought this explanation of the stepper motor EGR valve (Mitsubishi) might be of interest: EXHAUST GAS RECIRCULATION (EGR) SYSTEM <4B10-Low CO2 specification, FFV> . I think this is pretty close to the Subaru.

 

. . . I eventually brought it to a shop that diagnosed it as an EGR valve problem. He cleaned the valve, and I was off driving for a few months.
Then the problem reappeared.

. . . So when it was dry, I shone a flashlight on one side and could see just a little light on the other side.

. . . When I removed it this time, the valve was stuck open about a 1/4".

. . . I closed the valve by prying with a screwdriver and it took a lot more force than I expected.

. . . 1) how smoothly/easily should the valve move open and closed? mine is quite stiff . .

Based on the RockAuto catalog, the Subaru EGR valve and the Mitsubishi EGR valve in the linked paper (post #12) look very similar. The following diagram of the valve from the linked paper above provides a good idea of how it works, mechanically. I've added some notes.

The [bolded] observations in the first post suggest that the valve stem is probably binding in the bushing. This can be verified by removing the upper motor assembly from the valve base (4 screws).

With the motor part off, it should be possible to press down on the top of the Valve stem (at the spring) and check that the stem moves down and up (the latter by spring pressure) easily.

If the stem is binding, washing with various cleaners might not adequately clean out the area between the stem and the bushing. Moreover, if the bushing/seal is worn, then exhaust deposits could reform there again, as might have been the case after the mechanic cleaned it the first time. In this case the EGR valve probably needs to be replaced.

There's one other possibility, namely, that the stepper motor is not reversing enough to retract the Shaft fully, thereby holding the Valve stem down. If, when the EGR valve is removed from the intake the valve is open, then if the four screws are retracted, the spring acting on the Valve stem will push the upper part away. If this is the case, then either the Rotor/Shaft mechanism (which changes the turning of the rotor to up-down motion of the Shaft) is binding, or, the ECM is not controlling the stepper motor properly. However, I think this is less likely.

 

Since reinstalling it after prying closed the valve, my car has been driving fine for daily commutes ~6kmx3 and 50km trips.

On one of those trips I had freessm logging in the backseat (it would be distracted driving to have it on the passenger seat?), the egr valve steps was at zero the whole time.
Granted it was colder than -18 outside, so maybe it wasn't being commanded to open.

plainOM: in the "detailed valve operation" thread, under what conditions did you make those step measurements? it was mentioned somewhere that the egr valve isn't opened/operated when ambient temps are below 0'c.
btw, that thread is beyond awesome on many levels (linked resource documentation, thorough explanations and experiments). I shared it with some engineer friends who were very impressed too. thanks for your great work there, I'm very glad to have piqued your interest.

also based on your disassembly, do you think the return spring is strong enough to close the valve when power is removed? or is it simply helping the stepper motor?
since mine was stuck open a fair bit, I wonder if my valve stem is sticking/too much friction and will soon need to be replaced.

 

Since reinstalling it after prying closed the valve, my car has been driving fine for daily commutes ~6kmx3 and 50km trips.

On one of those trips I had freessm logging in the backseat (it would be distracted driving to have it on the passenger seat?), the egr valve steps was at zero the whole time.
Granted it was colder than -18 outside, so maybe it wasn't being commanded to open.

Was it FreeSSM or Romraider? I presume Romraider because FreeSSM only displays data.

My understanding is the EGR stepping isn't functional when the Outside Temperature is at or below 5 C. This is based on the "Enable Conditions" for the OBD to test the EGR (which I described in the valve operation thread), as indicated in the Detection Criteria for the P0400 trouble code, namely, that it requires the ambient air temperature to be above 5 C before the test will be run. I've been monitoring EGR Steps for several years now, and invariably during winter, it never goes above zero if the outside temp is below 5 C.

plainOM: in the "detailed valve operation" thread, under what conditions did you make those step measurements? it was mentioned somewhere that the egr valve isn't opened/operated when ambient temps are below 0'c.

All the disassembly and measurements (heights, rotation, number of steps etc.) were done in my workshop, room temperature.

. . . I'm very glad to have piqued your interest.

That you did, and I learned a lot about stepper motors, and this EGR valve.

also based on your disassembly, do you think the return spring is strong enough to close the valve when power is removed? or is it simply helping the stepper motor?
since mine was stuck open a fair bit, I wonder if my valve stem is sticking/too much friction and will soon need to be replaced.

Absolutely. Because the stepper motor pushrod isn't mechanically attached to the valve stem (at the spring), it can only push the valve down against the spring. It has no way to pull the valve itself back up to the closed position. As the stepper motor is reversed, in steps, by the ECM, and the pushrod retracts, the spring has to move the valve up to the closed position. It's a necessary function of the spring.

If I recall correctly, you found the valve open. That's clearly a mechanical fault. In my tests, where I was making quick contact with the connector terminals to cause the stepper motor to push the valve open, if I inadvertently removed power from the motor coil(s), the spring immediately pushed the valve closed, (which means, at the same time pushing the motor pushrod back up). Without power the stepper motor rotor isn't locked in place -- it can turn. I found that if the pushrod is out, it can be pushed back in, but pushing the spring down took more effort, so the spring, again, should always be able to close the valve when the motor is not powered.

It's possible for the spring to break, or become weak, which could lead to the valve staying open. If that were the case, the spring can't be easily replaced in any event, so when an EGR valve doesn't close on it's own, a new one (or at least a replacement lower part consisting of the valve housing, the valve itself, and the spring) would be needed.

If I might ask, where are you located? (PM if you prefer.)

 

Stepper motors usually don't require mechanical latches, pawls or gears to make a "step". If you look at the motor you will see "poles" on the stator (the fixed, outer part) that line up with poles on the rotor (the spinning inner part). The electrical pulse will magnetically pull the rotor to line up with the stator by a certain percentage of turn (depends upon the number of poles and their angle to each other).

The pulses do not need to last very long (just a few tens of milliseconds) for each step and by switching what set of coils are used the rotor can turn in one direction or the other.

The coil that is continually energized is known as the "field".

So, if things are working correctly then you can do 48 or 50 (or whatever number) of pulses to turn the rotor and shaft enough positions for the lead-screw to pull the valve from seated and closed to fully open.

Things can happen that prevent rotation of the stem and seat. Probably the most likely is fouling/ gumming or scoring on the stem where it runs through the seal to get in to the exhaust gas stream (a pretty corrosive and hostile environment). Another would be if it became bound up (due to scoring or obstruction) and it damaged the lead screw threads or the nut (ripping off threads).

Unless there was a leak and water (or exhaust gases) got in to the motor side of things those components should last essentially forever.

I bet that 95% of the EGR replacements could of been fixed with a disassembly and cleaning of the valve, seats and valve stem.

** my personal opinion is that the entire idea of EGR being used to deliberately contaminate the intake air is stupid, and an obsolete approach given to us by 1960's engineers who were trying to reduce NOx emissions by suffocating engine performance. There are other ways now of reducing NOx with dual stage catalysts, combustion chamber design, variable valve timing or even urea injection. **

 

Stepper motors usually don't require mechanical latches, pawls or gears to make a "step". . . .

I covered much of that in this thread.

I bet that 95% of the EGR replacements could of been fixed with a disassembly and cleaning of the valve, seats and valve stem. . . .

I would agree, in general. However, I doubt the particular EGR valve that is the subject of this thread can be disassembled to that level to properly clean the valve, seats valve stem and guide, and then be reassembled reliably.