Still P0401 after cleaning egr !! should we drive it?

I just hold the valve again in my hands, and yes, you probably are right. There is little room for the nipple to jump somehow over the intact ski jump. Maybe the form is only to reinforce the stop point?

In my mashed valve, the rotor gets about 120 degrees past the original stop point - how many steps it would be?

I agree - unlikely.

At my valve, it does not stuck like this, so I would consider it unlikely too.

Mat

 

The number of steps per revolution might be deducible from the number of metal 'fingers' visible in the stator.

The published number of steps from (the correct, ski-jump indexed) closed position to fully open is 110. I don't know how many revolutions that's supposed to be.

 

I just measured the pitch of the thread - it is 2.4 mm, so the 120 degrees end up to be a difference of 2.4 / 3 = 0.8 mm

So you get the valve 0.8 mm less opened, if the ski jump is completely mashed.

And as we know, the system is able to detect that.

Mat

 

I suspect the designers of my furnace's gas valve also considered it unlikely. I hope they were cozy and warm that night.

I agree we've seen in this thread that at least one car, ozmatt's, could fail the flow test and set P0401 because of a mashed ski jump.

Whether what it detected was the 0.8 mm smaller opening of the valve, or a more complete sticking-shut of the valve, I'm not sure was positively determined, in ozmatt's case anyway.

 

There is a third of a turn, when the valve remains completely closed, as the computer expects to see some flow.
No one knows the test procedure, except Toyota itself maybe...

Mat

 

For me at least, best price for a new valve, the revised part no 25620-37120 (old no is 25620-37110), was through Amayama. One expensive way to fix the issue. Very frustrating that you can't just get a "rebuild kit": for the dang thing, namely that wear item with the ski ramp.

There's a Toyota document for using that new valve, involves cutting some of the conduit insulation to allow it more reach, and a software update?

 

Yeah, so the replacement valve is a bit dimensionally different, requiring the extra reach for the wiring.

And the TSB that announces it also touts a certain new firmware revision.

The same new firmware revision is also touted by a different TSB that announces an updated intake manifold for a different issue.

The TSB about the EGR valve doesn't say you need the new manifold, and the TSB about the new manifold doesn't say you need the new EGR valve. They just both say you'll want this newer firmware rev.

That suggests to me that there isn't anything in the new firmware rev that is inextricably tied to either the new manifold or the new valve, and that probably they have revised some control strategies to make some past observed problems less likely to happen again. It might be interesting to see how the car behaves with the firmware update and neither the new manifold nor the new valve.

It may be moot anyway, as by now there are newer firmware revs than that one. I think the one for keeping the EVAP hose from blowing off is newer. (Note to self: I haven't had that one applied yet.)

Yeah, I bet they know it.

We know it in general terms. It involves opening the valve by some amount, while the car happens to be in decel fuel cut, and measuring the increase in manifold pressure.

We don't know how many steps it opens the valve for that test, or whether it's always the same amount, or if not, what it depends on. We also don't know why the 'min' threshold for failing the test isn't a fixed number, which is weird, or what that depends on. (Possibly specifics of the fuel-cut decel driving conditions when the test was snuck in?)

I'm surprised there isn't some electronics hobby site selling little circuit boards for counting six-terminal stepper motor steps and showing the number on a little display, sold six for $9.95. There are so many other things like that available these days. But I didn't have luck finding one of those, anyway. That would make it much easier to watch what the valve is doing.

 

I bet, only a few engineers will know really, what the procedure is.

It would not be very hard to make something like this out of an arduino nano and a bunch of resistors.
For me more of a PITA is to securely connect probes to the valves contacts...
I do not like to pierce any isolation in the engine bay.
I thought about using a few (maybe three) hall sensors to determine the rotors direction of movement and maybe the amount of steps, but did not manage to do it.

Mat

 

Maybe someone could get Toyotas ear, get that bucket with ramp established as an available part.

 

Yea... I saw this picture already, but wondering, how shall I get the probe in and keep the lamelles of the seal bent outwards...

 

Additional Data:
On the disassembled valve, the travel from completely closed to completely opened is 6.5 mm.
That would give us 6.5 / 2.4 = 2.7 turns of the rotor.
The rotor is almost certainly not able to go so far to open the valve, so I would think, there is between 2 and 2.5 turn of the rotor from closed to opened valve.

I'm rather on the 2.5 turn side

So, the 111 steps would have to turn the rotor 2.5 turns, ergo we get something like 44 steps per turn.

One third of a turn are then about 15 steps, where the computer gets no flow.

Mat

 

Not sure if you have an AutoZone in Vancouver but they sell a Duralast version with a lifetime warranty. Thinking of picking one up…

 

@ChapmanF I have the firmware, but old manifold and old EGR valve. Car runs normally…

 

As a way to cross-check that, it probably has a simple relationship (maybe 1:1, maybe 2:1 or 1:2?) to the number of metal 'fingers' visible in the stator. I don't remember ever actually counting those.

 

There are 48 (4x12) visible "fingers" in the stator.
And the rotor has 12 pairs of magnetic poles (12 x south, 12 x north).

Mat

 

My money might then be on 48 steps per turn, and the active travel being about 2.3 turns.

 

Exactly my thought

Mat

 

Methinks the valve would have to open to or near its maximum opening, in order for the test to be most sensitive to changes in other EGR flow restriction.
Those "decel driving conditions" would have to include engine speed, atmospheric pressure, and especially throttle angle (unless that's always the same during the test).

 

Yeah, that's the big challenge for the engineers writing the code for a test like that. It looks for moments in your normal driving when you are decelerating, with the engine not stopped, but spinning with no fuel injection. When it sees such an opportunity, it sneaks a test in. But it has no way of controlling, or even knowing, whether you're going down a long or short hill, or approaching a curve or a stop light, whether you're going to decelerate more strongly or less strongly in another hundred feet, or whatever.

So it has to sneak the test in under whatever conditions it gets, and it can't strictly control all of them. But I wonder if, by the end of the test, it can look back at what the actual conditions were, and then maybe 'grade on a curve' by computing what the min passing value should be for those test conditions.

Most of the other monitor tests just have their 'min' or 'max' thresholds as fixed values, but on the EGR test thread, people keep posting results where the 'min' values shown are not all the same. That's weird, and some explanation like this might be part of the story. (Otherwise, I don't know what is.)