When you click on links to various merchants on this site and make a purchase, this can result in this site earning a commission. Affiliate programs and affiliations include, but are not limited to, the eBay Partner Network.
Just wanted to give Bruce7 more props for this thread.
I changed the SAI valve last night and it could not have been easier thanks to these instructions. The most time consuming part is getting the air box back in place. I've done that about 5 times now, and I'm still slow at it.
My experience was different than his in a few ways:
* No need to take off the MAF. I just loosen the near hidden hose clamp, then rotate the top of MAF toward the front of the car a bit to free it from the air box.
* I didn't have interference from the injector and knock sensor cables/clips, so just left them alone. Saved a few minutes dealing with those plugs.
* The vacuum line to the change-over valve stayed attached. His photo made it easy to confirm with a mirror.
My valve must have been changed with the top-end rebuild done by the PO ~30k miles ago. It was not in as bad of shape as some that I've seen online. Whoever changed it last time used a sealing ring and plenty of anti seize (although not the copper stuff Bruce recommended).
Just wanted to give Bruce7 more props for this thread.
Thanks for the update. Coincidentally, my CEL came on yesterday with the
P1411 code. Now sitting at 117,267 miles, I went 42,619 miles since the
SAI port clean-out. The clean-out was done in 4/2013. I had changed the
check valve in 4/2015. The port clean-out is looking like an every 2-1/2 year
exercise at my average of 17k/year mileage.
The good news is that the car is running good with no indication of a need
for top-end job yet. A relative compression test shows all cylinders even.
I changed my SAI check valve a year and 3000 miles ago. The CEL came back on almost immediately. This past weekend, at 98,500 miles, I did the SAI port clean out, including flush from the top end, at the valve.
Upon removing the check valve, I blew through it and it held no pressure. I was surprised and quite disappointed. I took it back to Sonnen and they gave me a new one under the two-year parts warranty - a big shout out to my buddies there!
So, for me, this raises big questions. How long do these $75 valves really last? The new one looked exactly the same as the one I put in a year ago. If they commonly fail within a year, how can we keep this damn foolish system working...
After all the labor (and some expense) to roto-router and flush the SAI ports, it is quite frustrating to know this valve can defeat my efforts. I plan to pull out the valve a year from now and check it for proper operation. I'll probably throw in a new one just to be sure.
I just completed the SAI flush- and the CEL light came back on after 130 miles! Here's what I found when I was in there: The valve was really bad- totally blocked- but the passages were VERY clean We flushed them 6-7 times and had no carbon coming out at the end. The valve seats and ports were so clean, I was seriously wondering if the top end has been done before. The ceramic sleeves around the valves still show white with no carbon at all. So this has me really miffed. What would trigger the light after a shot time and the system is clean? Is there a sensor I need to replace?? HELP!!
What would trigger the light after a shot time and the system is clean? Is there a sensor I need to replace?? HELP!!
You say CEL, but I'm assuming you're using that as shorthand for the SAI codes, yes? If not, what specific codes are you getting?
The SAI effectiveness is determined by reading the O2 sensors, so that's a possible path to investigate.
Looking at the diagram:
Assuming the sensors are OK, clean passages but a fully blocked check valve (5), and the passages blowing clean (so no blockage in the distribution housing (6)) would mean that air is not moving through the system, so the pump (13) or the shut-off valve (4) would be bad. The pump connects to the air box with a knockout, so those connections and the knockout on the airbox itself can be checked.
All of that stuff is upstream of the check valve, so my guess (as some random guy on the internet) is that your O2 sensors are the culprit. But again, that's assuming that we're talking about the SAI codes.
We are talking SAI codes 0411 I believe- My hunch is 02 sensors as well. The codes fire around 120-150 miles of driving which would lead me to believe sensor as well, since an air pump mal-function would send a code earlier wouldn't it?
We are talking SAI codes 0411 I believe- My hunch is 02 sensors as well. The codes fire around 120-150 miles of driving which would lead me to believe sensor as well, since an air pump mal-function would send a code earlier wouldn't it?
My understanding of how this works is that the SAI system is not directly measured; only the pre-cat vs. post-cat measurements from the O2 sensors are compared at startup for the first few minutes. So I wouldn't expect a malfunctioning part in the SAI system to trigger any instant errors. It's all up to the test cycle. 120-150 miles doesn't seem unreasonable, but I don't know the specifics of the computer here.
You could go ahead and replace the O2 sensors (esp. if they are nearing end of life). I think Andy had a thread in the last few weeks on this strategy. Alternatively, if you can get your hands on a test tool, you might be able to get more detailed readings from the O2 sensors to confirm whether they're in good shape or not. I'd rule them out one way or the other before diving in further.
That's where I am at with this as well. I hear the pump coming on and shutting off so I think this has to be a sensor issue. My bet is that they've never been changed and I have 113 K on the clock now.
That's where I am at with this as well. I hear the pump coming on and shutting off so I think this has to be a sensor issue. My bet is that they've never been changed and I have 113 K on the clock now.
Prior to spending the money on the O2 sensors, I would pulse Andy on the results of his strategy.
If 113k mi. on original O2 sensors, I would concur on the thought that it's time for a change.
If you have a durametric test tool, you can validate the performance of your O2 sensors. I don't know for sure if the handheld, store bought OBD2 readers provide real time, voltage read outs of the O2 sensors.
...an air pump mal-function would send a code earlier wouldn't it?
To my knowledge, the air pump only operates for the first couple of minutes after startup, acting as a bellows to assist in heating the catalytic converter. If the air pump was non-op; wouldn't the code be different?
I have a really good OBD 2 test tool - i'll see if it does work with real time o2 sensor readings. I have a crack in the CV boots on the rear CV joints as well and am taking it to my mechanic to get that fixed so i will have them check the sensors as well.
The air pump works- I hear it come on and go off and it's blowing well, so I am thinking sensor.
Outer- and YES I am going to replace both axles. The shop I work with will probably charge me the same for the replacement versus the added work of repacking the outers, etc. with 113K on the clock, I'll sleep better with new bearings, etc anyway! Love this car and it's not going anywhere so let's do it right. This car is in terrific condition- and the motor is SO strong and pulls SUPER hard. Been in a lot of 993's and this one performs like one of the better ones. I wanna keep it that way.
The offset of the socket tool attachment point looks a bit long and may result in the attachment being under the tube next to the valve obstructing the attachment of the socket tool from above. I had this problem when I purchased a generic crows foot on eBay. I guess we must stick to the brands that folks have had success with. Other options include cutting off the valve body and using a socket wrench or taking a crack at it with a 3/4" internal pipe wrench, however I have not tried or heard of others trying the internal wrench approach.
Andy
I just did this 2 weeks ago and the crows foot was not the answer- Could not get enough leverage on the valve - that baby is on there something fierce! Ended up breaking the top of the valve off and using a 29MM socket- I think the bolt is actually a 27 MM, but there was enough metal sticking out around the top of the valve where it broke that the 29 seated well. I pounded it on a bit and then was able to crank it off. It's a bit unnerving to break the valve off as there's no going back at that point- but it worked and I have seen others comment on this approach as well. Now to fix my 02 sensor(s).
12-13-2015, 08:44 AM
