Need a clarification on MAP sensor

[kyledooley]

Hi all. I've only recently joined the fray of turbo legacy owners, so first of all I'll say hi. I just bought a 1992 white Sport sedan Auto. It needs some love, but just regular maintenance really, and the body is fabulous. For those of you on nasioc.com, I'm one of the old dogs and hang out in the midwest forum... Same username.

Secondly, I'm trying to resolve a couple of CE codes I'm getting. I pulled them and I'm getting

23, 35, 45 and 49.

The MAF codes don't scare me, the car seems to be running fine, but the one I'm wondering about is the Atmospheric pressure sensor code.

One of the flow charts I read for the APS says its is in the ECU and if you have that code, replace the ECU.

Now I read here on the forums that for NA cars its in the ECU and for turbo cars it is the same sensor used for Manifold pressure and atmoshpere with a solenoid switching between the two lines.

So, does my 92 Sport Sedan (Auto) use the external MAP with the solenoid arrangement? That would make the most sense to me.

I also read here that a few months ago, somebody was going to try a GM 2 bar MAP with a signal conditioner. Does anyone know if there was some sucess with that? Because that is the way I would like to go if I really need to change it. GM MAPs are pretty much free compared to the 200+ I would shell out.

The CE light is off when the car is cold (so no hard failures, I don't think) but it comes on when the car is warm and at a light cruise. So I'm pretty sure the only recurring problem is the purge control solenoid. Does anybody know if that part is shared with any other cars... I'd hate to have to spend the cash they're asking when I can get it from another subaru in a Junkyard or a newer one that one of my buddies has lying around their shop from a WRX or an RS.

My first stop though, is to reset the codes and see which comes back. I just didn't know the secret handshake to do it without disconnecting the battery.

Thanks all,

Kyle

[vrg3]

Code 45 is for the atmospheric pressure sensor on non-turbos, and for the pressure sensor and pressure exchange valve on turbos.

You're right that the pressure is built into non-turbo ECUs and is external on turbos with the solenoid.

I was the one who was gonna try the GM sensor. Haven't gotten around to it, but if you like I could build you the conditioner and you could test it for us. :)

Realize that your problem could be the pressure exchange solenoid rather than the sensor. I think the sensor is the same for newer Subarus too. I have a spare solenoid from a 97 Outback that appears to be functionally the same.

Try clearing the codes first though... Always do that to make sure they're not just spurious codes.

[kyledooley]

Code 45 is for the atmospheric pressure sensor on non-turbos, and for the pressure sensor and pressure exchange valve on turbos.

You're right that the pressure is built into non-turbo ECUs and is external on turbos with the solenoid.

I was the one who was gonna try the GM sensor. Haven't gotten around to it, but if you like I could build you the conditioner and you could test it for us.

Realize that your problem could be the pressure exchange solenoid rather than the sensor. I think the sensor is the same for newer Subarus too. I have a spare solenoid from a 97 Outback that appears to be functionally the same.

Try clearing the codes first though... Always do that to make sure they're not just spurious codes.

Great. Thanks for confirming my suspicion. I agree that it may just be the solenoid. And if that code comes back, I'll start there. Thanks for the info about the similar solenoid on the outback too.

I was trying to also follow the thread about using an NA MAP sensor. I think you said in a previous post that they are only good to 6psi max as far as what they can output.

The question is, is the ECU changing fuel and timing when it sees more boost? What would be the disadvantage to just using that particular sensor which stops reading at 6psi. It has some advantages such as being an improvised FCD, but what is the ECU doing differently as it sees boost rising? In other words, what are you giving up if the ECU is not seeing more boost beyond 6psi.

If I want to try the GM sensor, I'll let you know, it sounds like a good solution and I'd be happy to try it that's for sure. It should just be a matter of getting the output voltage into a range that the ECU is happy with.

I agree though that the first stop is clearing the codes and seeing what comes back.

Thanks again,

Kyle

[boostjunkie]

Great. Thanks for confirming my suspicion. I agree that it may just be the solenoid. And if that code comes back, I'll start there. Thanks for the info about the similar solenoid on the outback too.

I was trying to also follow the thread about using an NA MAP sensor. I think you said in a previous post that they are only good to 6psi max as far as what they can output.

The question is, is the ECU changing fuel and timing when it sees more boost? What would be the disadvantage to just using that particular sensor which stops reading at 6psi. It has some advantages such as being an improvised FCD, but what is the ECU doing differently as it sees boost rising? In other words, what are you giving up if the ECU is not seeing more boost beyond 6psi.

Kyle,

I actually tried the 6psi MAP sensor that vrg3 "conditioned" to use with my car. The problem that I found was that the ecu had no idea that it was getting over 6psi of boost. While fine with a stock setup (I think vrg3 mentioned that this setup worked on another member on here) I was running over 14psi!!

Whenever I would hit anything above 11-12psi I would get some nasty hesitation issues, which I chalked up as the ecu altering fuel/timing for 6psi . . . when it should have adjusted for 11-12psi. I think 10-11psi was the limit that was possible with this MAP/FCD fix. After that, the fuel/timing curves were too screwy. Again, this is based on my mods.

I was able to check my pressure exchange solenoid for proper operation by hooking up my boost gauge to it. At certain times while the car was idling, the ecu would take an atmosperic pressure reading that I could see by a change in the boost gauge reading.

Also, do you notice any irregularity when under boost? My car wouldn't even let me get into boost when my MAP sensor was toasted!

[vrg3]

I was trying to also follow the thread about using an NA MAP sensor. I think you said in a previous post that they are only good to 6psi max as far as what they can output.

That's right; they have a 1.5-bar range. The thing is, though, a 2-bar GM sensor is easier to find and cheaper than a 1.5-bar Subaru sensor, and either one would require a signal conditioner, so it seems like you might as well use a GM sensor.

The question is, is the ECU changing fuel and timing when it sees more boost?

We don't know for sure yet. :) I suspect it is but haven't experimentally verified it or found the portion of the ECU's code responsible for it.

Like you say, limiting yourself to a 1.5-bar sensor gives behavior like an improvised FCD... so you're definitely giving up the factory overboost protection.

If I want to try the GM sensor, I'll let you know, it sounds like a good solution and I'd be happy to try it that's for sure. It should just be a matter of getting the output voltage into a range that the ECU is happy with.

I've actually got the signal conditioner circuit all figured out, since boostjunkie and I were considering this option for his car. All I have to do is assemble it. He ended up buying a used Legacy Turbo sensor instead though.

[vrg3]

Heh, boostjunkie, you replied while I was replying too. :) Remember, on your car we have no idea what was going on with the 1.5-bar sensor because your ECU was not properly handling its power supply. You only found the power supply problem after you'd switched to the 2-bar sensor.

[kyledooley]

No. No changes in behavior. The car heads out just fine. I just got it though, so I haven't leaned on it too much other than leaving a toll booth.

Every day or so, I've been pushing it a little harder and it performs flawlessly. I'm sure the codes I mentioned above are just odditites. I just didn't know the secret handshake for resetting the codes until this morning. I will do that and see what happens.

The CE light isn't on at all until the car is warmed up and I'm at a light cruise. I think that is the purge control solenoid, because my 2000RS did the exact same thing when my PCS went. The other codes, I think are just stored from some freak event and have just never been cleared, because the car runs tip top. If the MAF was messed it would misbehave, and, like you said, if the MAP was gone at least something wouldn't feel right.

Anyhow, I'll let you guys know how it goes.

K.

[Legacy777]

kyle,

Have you cleared the codes, driven the car, and verified the exact same codes come back again?

[kyledooley]

kyle,

Have you cleared the codes, driven the car, and verified the exact same codes come back again?

No. I'm away for work right now, but thanks for checking. I will be back tonight and its the first thing on my list.

K.

[kyledooley]

Good news.

The only recurring code seems to be the purge control solenoid. It won't reset the light, but I drove it around all day and that is the only code flashing in diagnostic mode.

Today, I also did a little tune up. Plugs, oil change, air filter. After reading the forums, I thought a new set of plugs would be a good place to start to cure my occasional cold hesitation, and it seems to have done the trick. It seems to have done the trick, and I felt a whole lot better about the motor after getting a look at the plugs. They all looked great.

I also fixed the blower motor relay. The HVAC fan wasn't working and I traced it to that. FYI it is a bit** to change. It's up above the fuse block under the dash and impossible to get at. It did fix my fan though. All speeds work now instead of none at all. (very important in Chicago)

Huge thumbs up to Legacy777, without the wiring diagrams and the electrical component locator in the service manual, I would never have been able to fix it.

Next will be the Timing belt, crank seal, fix my power steering leak, and flush the coolant, radiator hoses and maybe water pump. The coolant is the nastiest I've ever seen.

I also need to get intimate under the hood with a roll of vacuum line. The lines are all a little brittle, which is to be expected for the car's age.

Anyhow, enough babling. Thanks everyone. I guess it was a false alarm on all the codes except for one, so it's all good. I'll let you know about the purge control when I get it fixed.

Kyle

[ciper]

The purge valve isnt really necessary, its only for emissions. I suggest either getting one from the junk yard (bring a multimeter) or replacing it with a resistor to fool the computer.

[kyledooley]

What type of resistor and where to put it?

Just short the connector?

Kyle