Running REALLY rich!!

[HugeFrankV]

My car is dumping fuel and it finally killed my 02 sensor. The guy I bought the car from put cams and a bunch of other mods in it, but no fuel management. Also, a friend of mine has a code scanner that can give you real time info while your driving around and according to his scanner my timing is retarding like whoa at full throttle. Any thoughts or input? Thanks guys

[SideshowBob]

Sounds like you need fuel management ASAP. I dunno enough about the 2.5 to comment intelligently, but you should post a complete mod list and any other info so that the dudes who know what's up on here can better help you.

[HugeFrankV]

Good call man!! Thanks!! It's a 99 2.5 with a Cobb cai, stainless header and exhaust with hi flow cat. It has TWE race cams and titanium springs. The dude told me that the cams were somehow messing with the knock sensor, so he bypassed it with a resistor. That's all the mods as far as making power.

[hydrochloric]

Aggressive cams cause a phantom knock at idle, because they idle really roughly.  The PO clearly did the WRONG thing and fucked with the all-important knock sensor.  The proper response is to up the idle (can be done with a 1.3U or whatever that cable was).  If you can't find anybody local with one of the few cables that works on our early cars, you can turn on the A/C (I'd disconnect the connector at the compressor), because the ECU will up the idle.  It will eventually adjust around that, though.

The timing is controlled mainly from the knock sensor, so the resistor could really easily cause a retard at throttle, especially since the ECU now has a weird reading compared to the rest of its readings.  Basically, the ECU was probably seeing a weird A/F or some other sensor without a corresponding reading from the knock sensor.  All of this paragraph is conjecture, but I found out alot when I was considering Delta regrinds for my '00.

The '99s were the first year of the "smart" ECUs, IIRC, meaning fuel management is going to be more challenging.  As in, you'll need piggyback control, not just a FPR.

From what I can tell, the mods shouldn't cause any issue (although you'll probably have a P0420 constantly due to the hi-flow cat).

[TheBigChill]

  ^   I'm no RS pro, but I've heard that mainly the 99's have a very sensitive knock sensor.  Just throwing a resistor in there is not a smart thing to do, regardless of however many people have done it.  I don't think I need to explain why it's bad to blindly reduce the signal output of a knock sensor in an untuned motor.   
The other bit is:  If your primary O2 sensor is dead, that alone will cause very rich fueling and retarded timing. 

[crAppAchic0]

This doesn't happen to be Bill's (homurcules') '99 white RS that he sold a little while back, does it?

[peter32892]

No need really for engine/fuel management a reflash would help you out. The resistor is a bad idea, if you want to reduce false knock you can move the knock sensor to another spot further back, I believe onto the transmission house but I'm not exactly positive on its location, but that shouldn't be necessary. All you should need to do is up your idle, get a new 02 sensor and a reflash if you want to. What other mods are done to the car?

[hydrochloric]

^ Unless the harness has been swapped, and the ECU along with it, the only thing you can change on the old ECUs is the idle speed.  I know, I looked over and over for a way to tune the stock computer....  And it can't be done on the old ones.

[phong]

^ Unless the harness has been swapped, and the ECU along with it, the only thing you can change on the old ECUs is the idle speed.  I know, I looked over and over for a way to tune the stock computer....  And it can't be done on the old ones.

You didn't look hard enough...not sure if he is still around but it can be done.
http://www.rs25.com/forums/f56/20558-sts-preperation-4.html

http://www.rs25.com/forums/f72/t77819-yoshi-rs.html

[hydrochloric]

I'm pretty sure those are services where you send out the ECU and they swap the chip itself, or have a large complicated hookup via the main harness plug.  I'm talking for open-source ODBII cable-type flashing, the type most of us can do.

[phong]

I'm pretty sure those are services where you send out the ECU and they swap the chip itself, or have a large complicated hookup via the main harness plug.  I'm talking for open-source ODBII cable-type flashing, the type most of us can do.

No harness but it does require more than a simple Tactrix cable...you said it couldn't be done.

[hydrochloric]

No harness but it does require more than a simple Tactrix cable...you said it couldn't be done.

....did you ever consider becoming a lawyer? 

[peter32892]

^ Unless the harness has been swapped, and the ECU along with it, the only thing you can change on the old ECUs is the idle speed.  I know, I looked over and over for a way to tune the stock computer....  And it can't be done on the old ones.

You can send your ecu out for a reflash...
http://store.crawfordperformance.com/store/categories/249

[hydrochloric]

As I said to phong, I meant using a cable in the convenience of one's own home.  For that, there's no real options.

[HugeFrankV]

This doesn't happen to be Bill's (homurcules') '99 white RS that he sold a little while back, does it?

It is Bill's old car!! Is that a good thing? Lol!! Thanks for all the responses guys!! So, I don't know much about reflashing and ecu related stuff, how do I go about reflashing?

[HugeFrankV]

No need really for engine/fuel management a reflash would help you out. The resistor is a bad idea, if you want to reduce false knock you can move the knock sensor to another spot further back, I believe onto the transmission house but I'm not exactly positive on its location, but that shouldn't be necessary. All you should need to do is up your idle, get a new 02 sensor and a reflash if you want to. What other mods are done to the car?

Got a Cobb cai, stainless header, exhaust and high flow cat, TWE race cams and titanium springs, Cusco front strut bar, Whiteline adjustable rear sway bar, KYB AGX adjustable struts front and rear, RacingBrake slotted rotors front and rear and Hawk HPS pads front and rear, Exedy lightened flywheel and organic clutch. and although there's no performance benefit, a version 5 STI wing!!!! Lol!! 

[hydrochloric]

Got a Cobb cai, stainless header, exhaust and high flow cat, TWE race cams and titanium springs, Cusco front strut bar, Whiteline adjustable rear sway bar, KYB AGX adjustable struts front and rear, RacingBrake slotted rotors front and rear and Hawk HPS pads front and rear, Exedy lightened flywheel and organic clutch. and although there's no performance benefit, a version 5 STI wing!!!! Lol!! 

Ooof, cams AND a ltwt flywheel, with no idle bump?  I'm surprised the thing idles...

[HugeFrankV]

Lol!! So, I did a little research. I was thinking maybe an A'pexi S-AFC. Will that be able to bump the idle high enough to ditch the phantom knock and take care of all the fuel management issues?

[hydrochloric]

Uh... No.

The S-AFC only controls the what the ECU sees from the MAF.  That won't change idle one iota.

Plus, as long as you have the resistor in there as opposed to the proper knock sensor, there's no such thing as "phantom knock."  Your ECU is simply seeing a non-running engine, because the signal from a resistor is just a voltage.  A knock sensor output looks like this:



With just a resistor, the graph would simply be a line.  I have no clue what that's making the ECU do, but it's not right...

The S-AFC will lean the fuel out at high RPMs, though.

http://www.alamomotorsports.com/apexi/safc.html

[HugeFrankV]

Damn!! So I suppose I should plug the knock sensor back in and see how things go? It's still there the resistor is just plugged in instead of the sensor. So do you think a reflash will take care of upping the idle after I plug the knock sensor back in and this "phantom knock" returns and the running rich issue?

[hydrochloric]

You said you could read ignition timing.  Can you also read the fuel trims, and such?  I would read with and without the sensor plugged in, but...  The mods you have should only cause a problem at low RPM, unless those cams are race-ground or some other super-aggressive lobes.  However, from what you're saying, it sees a knock at all RPMs?

[HugeFrankV]

No, unfortunately I can't. My friend has the "high end" code reader from advance or wherever but as far as I know it only shows your ignition timing. Do you think an A/F gauge would be a wise investment? The guy I bought it from said they're race cams, but it's not as lumpy of an idle as I would've thought, so I really don't know for sure. He said shortly after having the cams installed it started running like shit and after a ton of asking around and research he was told about the false knock that the sensor can pick up from the cams, so he put in the resistor.That's all he told me, so I'm not sure if it was only at idle or certain rpm's or what.

[hydrochloric]

Well.... It's a 99, so any OBDII code reader that supports live data ought to be able to read those parameters.  I have an OBDII WiFi module that hooks up to my iPod, so I can see all that (and a lot more).  It's been a huge help, though I'm not necessarily suggesting you run out and get one.

I suspect that your problems stem from a combination of factors.  The cams cause misfire and knock codes at idle, because of the roughness, but once at speed they shouldn't have a problem.  On RS25.com, there's a HUGE thread where people traded a stock set of cams so they could still run their cars while Delta ground them, and while most of the people got mild or aggressive street grinds, even the ones who got the super-aggressive cams only reported misfire codes at idle.  Most of the guys who got street grinds never had any problem at all...  IIRC, the guys with lightweight rotating mass (flywheel, pulleys, etc) had more issues with idle, and some said they HAD to up it or the car would stall, but I don't remember anything about richness....

I'm suspecting there's a separate cause behind the rich.  First off, I would reconnect the knock sensor and see if you can feel hesitation at high RPMs (or read timings if possible).  If you can, there's a different issue than the cams, as they should cause no issues away from idle.  Thing is...  Lean has many causes in fuel injection.  Air leaks, etc, but richness is harder to figure out, the car can't be losing air somewhere after the MAF very easily...  This is a MAF car, right?  I can't ever remember if they switched in 98 or 99.

I suppose it's possible the MAF/MAP is bad.  It could read more air than there actually is, though IIRC you usually hear about them leaning out as opposed to richening.

Now...  How do you know it's running rich?  This should have been asked earlier, come to think of it...  Killing the O2 sensor isn't a great indicator, they die eventually anyway.

[crAppAchic0]

I've heard it happen and seen it mentioned in a few spots, but sometimes your lightweight flywheel can trigger the knock sensor to trip, causing the ecu to compensate, thinking it's a lean condition, making the engine run rich.  The best way to check that out, of course, would be to log the data from your knock sensor somehow- I mean, easiest/cheapest/quickest- hook up a multimedter and have a passenger in the car keep an eye on the readings- and go from there with whatever your findings might be...

Just my $0.02...

On a complete sidenote, if things don't work out, I'll trade you a stock head, clutch, and flywheel assembly for yours... free of charge- to get those things out of your hair...

[TheBigChill]

 This resistor "fix" reduces (not eliminates..) the knock sensor's output signal level while still maintaining an impedance with which the ECU is happy with.  You've essentially thrown away your ECU's ability to perform active fine knock correction.  Awesome.  Rather than totally eliminating an output signal, it just tones the voltage/frequency of the signal down, which will make large amplitude knocks still pass through to the ECU, but small knocks will get filtered down to near nothing.  The knock sensor is a simple, passive piezo microphone..  It's capable of outputting signals across most of the audio range.  Splicing in this resistor will not eliminate signal output to the ECU, I repeat.

 When the previous owner decided to splice in a resistor, did he use the right one ?  Did you check the banding yourself to be sure ?   When he mounted the sensor back on the motor, did he over-torque the sensor, changing the characteristics of the piezo element, resulting in erroneous output signals ? 

 Have you tried checking the resistance of the knock sensor itself to make sure it's good, and/or removed it to make sure it's in good physical shape (cracked ?), and there's not oxidation between it and the motor ?

 Bottom line: If your car is retarding timing so much at even normal partial throttle operation, this "mod" or "fix" is not doing what it should be doing...    Change the oxygen sensor.  Source a known good knock sensor, and install that one.  Run it a week and see where you're at.  Then and only then, proceed with your resistor application if you're still having troubles.  Better yet, stop fucking around with splicing shit in-line with critical sensors, and buy a MegaSquirt or something, and tune your car.