Okay, now what? (sperry's motor tear-down thread)

[sperry]

Originally Posted by van

Awesome! when are you going for a tune?

Well, I gotta get some miles on it as is to get the initial break in done (like 20 miles or so). Then I can try to wire up a stock ECU w/ RomRaider to build a map that will pass smog. Then it's back to the Hydra for 1000 miles of real break in time, which is why I need to pass smog, so I don't have to drive it around for 1000 miles w/o registration.

Once the break in is done, I'll probably just head out to a Fernley test-n-tune day and tweak the map that MikeW made for the car, since the car is basically unchanged aside from the last of forged pistons... so I can just dial back the boost and the timing a little, and I should be okay. I'm not looking for a ton of power, I think the car will be plenty fast.

[Kevin M]

You can probably get 1000 miles in in 30 days, which means you can probably get by with just a tweaker tag.

[sybir]

Glad to hear it's living, man

[sperry]

After 6 hours of messing with wiring trying to get the proper voltage to the flash jumper (for whatever reason, lots of '02s seem to have problems seeing the proper 12V at the jumper and of course I happen to have on of the early '02s with the problem ) I've got a map on the stock ECU allowing the car to idle, and *not* throwing any CELs (so far).

All I remapped was the injector latency and the injector scaling for my PE800's using some data from Cobb, and then I turned off all the CELs associated with all the missing EVAP, EGT and O2 sensors that aren't on my car anymore. I'm dying for a standalone wideband O2 sensor... I've got one built into the Hydra, but I can't use it with the stock ECU plugged in. I may have to go pick up a Innovate LC-1 which will work with RomRaider's logger... or better yet an LM-2 which works with the software *and* has a hand-held display making it actually useful without a laptop, but costs like $400 .

Anyway, the wideband is needed 'cause the MAF tables are still bone stock for the '02 WRX, and I'm running a 2.5L block with a FPGreen blowing thru a 70mm MAF housing... Not exactly what the original map was intended for. The car starts right up and runs, but the idle is a little rough (though it does idle better than the Hydra did in the past ). The bottom line is that it looks like I'll be able to put together a map for this thing on the stock ECU. It's going to be a PITA, 'cause this RomRaider software and OpenECU flasher stuff is so damn flaky and underdeveloped (the Hydra is pretty crappy itself, and it blows this free stuff away). But I think I'll be able to get this bitch registered and on the street for a proper break in, even if it means I'll likely miss the remainder of the trials season.

[knucklesplitter]

You shouldn't need a wideband O2 sensor to get the car the run and drive (and pass emissions). The stock one is fine above 12:1 AFR and the ECU uses it for closed loop anyway.

[sperry]

Originally Posted by knucklesplitter

You shouldn't need a wideband O2 sensor to get the car the run and drive (and pass emissions). The stock one is fine above 12:1 AFR and the ECU uses it for closed loop anyway.

Am I going to need the front and rear stock O2 sensors? Or does it just use the second one for checking for cat efficiency?

[cody]

As I understand it, the rear O2 is mainly for DP cat efficiency but I've heard it's used for fine fuel trim or something like that too. I can tell you that I removed mine and replaced it with Ed's wideband last time he tuned my car.

[knucklesplitter]

Originally Posted by sperry

Am I going to need the front and rear stock O2 sensors? Or does it just use the second one for checking for cat efficiency?

The front one is the main one it uses for closed loop. I have read that the rear one does some minor trimming too ( along with cat. eff.), but the front one is the one to log. Also log the fuel learning #1 & #2 along with MAF voltage. In closed loop you want the learning to be near zero and you tweak the MAF table to get there. To tune the lower end of the MAF table (what you need to pass smog) I usually log idle and a few progressive part-throttle runs staying in closed loop (less than 5psi boost or so). Then I put the logs in Excel and make some graphs of total learning versus MAF voltage. Focus on trends and ignore noise. Call me if you want to discuss or need help.

*** Stay out of open loop until the Hydra is back on.

[sperry]

Originally Posted by knucklesplitter

The front one is the main one it uses for closed loop. I have read that the rear one does some minor trimming too ( along with cat. eff.), but the front one is the one to log. Also log the fuel learning #1 & #2 along with MAF voltage. In closed loop you want the learning to be near zero and you tweak the MAF table to get there. To tune the lower end of the MAF table (what you need to pass smog) I usually log idle and a few progressive part-throttle runs staying in closed loop (less than 5psi boost or so). Then I put the logs in Excel and make some graphs of total learning versus MAF voltage. Focus on trends and ignore noise. Call me if you want to discuss or need help.

*** Stay out of open loop until the Hydra is back on.

Great info!

I'm definitely not going to get into open loop at all on the stock ECU... I'm pretty sure it won't like the MBC, even with it set to minimal boost. Besides, I set the red-line to 4000 rpm already, so I won't exactly be hauling any ass.

[knucklesplitter]

The ECU will be fine with the MBC - it will have no clue about it. Keep in mind that even though the MBC is tuned all the way down your turbo will still be able to boost to whatever your wastegate spring is calibrated to (prolly 14-18psi on that FP Green). You can also just jumper the boost hose directly to the wastegate for minimum boost, but the spring still is the factor.

[sperry]

Originally Posted by knucklesplitter

The ECU will be fine with the MBC - it will have no clue about it. Keep in mind that even though the MBC is tuned all the way down your turbo will still be able to boost to whatever your wastegate spring is calibrated to (prolly 14-18psi on that FP Green). You can also just jumper the boost hose directly to the wastegate for minimum boost, but the spring still is the factor.

Yeah, I know. My point was that running the stock ECU w/ the MBC in open loop leads to that whole pesky partial-throttle/full-boost kerblamo situation, specifically because even with the MBC set to minimal boost, wastegate boost is around 16psi which is still too much boost for the stock WRC fuel maps.

[knucklesplitter]

Originally Posted by sperry

Yeah, I know. My point was that running the stock ECU w/ the MBC in open loop leads to that whole pesky partial-throttle/full-boost kerblamo situation, specifically because even with the MBC set to minimal boost, wastegate boost is around 16psi which is still too much boost for the stock WRC fuel maps.

I think that turbo is big enough you won't get that part-throttle/hi-boost going on. But definitely watch out for it - any boost above 5psi.

[100_Percent_Juice]

Originally Posted by sperry

Yeah, I know. My point was that running the stock ECU w/ the MBC in open loop leads to that whole pesky partial-throttle/full-boost kerblamo situation, specifically because even with the MBC set to minimal boost, wastegate boost is around 16psi which is still too much boost for the stock WRC fuel maps.

can't you just put your bcs back in?

[sperry]

Originally Posted by 100_Percent_Juice

can't you just put your bcs back in?

It's actually in there and wired to the ECU to prevent CELs... just not plumbed to the turbo. I just don't need it for boost control, since the car should never see boost on the emissions map.

[cody]

I've heard leaving it connected can cause it to destroy itself and make an annoying ticking noise (like you'd hear it...) so I disconnected mine. The CEL is defeatable with Protuner, that much I know.

[sperry]

Originally Posted by cody

I've heard leaving it connected can cause it to destroy itself and make an annoying ticking noise (like you'd hear it...) so I disconnected mine. The CEL is defeatable with Protuner, that much I know.

Um... okay.

I'm sure I can turn off the CEL with RomRaider... I just haven't gotten that far. It's not like I'm driving the car around. I put a whopping 2 miles on it around my neighborhood on the Hydra right after I first fired the motor to check for leaks and to warm it up under some light load, rather than just at idle like I did the last time.

If I end up tuning this thing myself, I'll probably go to a GM solenoid, assuming the Hydra can control it. I'd rather be able to program my boost, than to rely on a fixed setting with the MBC, even if the MBC give me faster response. With an EBC on the Hydra, I can do stuff like anti-lag, flat-foot shifting, etc.

[cody]

Um... Okay.

[Dean]

If your Wideband uses a Bosch sensor, you can get away with a LC-1 control module only. A LC-1 is basically the same as a LM-1 minus the box. It has analog and digital outputs to feed either type of gauge in addition to a serial line obviously. I run an analog gauge because they are cheaper.

You should be able to rescale any of the maps for your configuration. You don't have as many tables as the STI, but should have plenty of latitude to make a closed loop and even low boost open map.

If you have the injectors mapped, mapping your intake is the next challenge to get it to hunting less or not at all at any point in open loop.

I have a spreadsheet that does double interpolation from logs to try and build the intake map based on trims and targets. It still requires some manual trimming of idle and off throttle data, but works for me.

[sperry]

LC-1 you say? Greg Benson welded a bung for the O2 sensor for it on Sunday, and I did half the wiring last night, but the install is taking a while because I'm cutting a dash piece for the stereo area to mount the display in (along w/ the Defi control box, the DCCD control box, and a few switches). Unfortunately, the Hydra's WBO2 is not a Bosch (it's a WEGO) so I couldn't just get the LC-1 controller. But I figure, having two WB's isn't total waste, now I've got one for the ECU to log with, and an independent one to compare against that's got a display in the dash (plus I can log with it on the laptop as well for OEM ECU tuning).

How'd you run the O2 sensor itself on your install? I like being able to drop the DP w/o having to take out the O2 sensors, but that means installing the "lambda wire" control box under the car where it's exposed to all sorts of water/oil/rubber/rocks/etc just so I can unplug the sensor... so I mounted the control box under the center arm rest and ran the O2 sensor through the shifter boot to the exhaust... so only the sensor itself is exposed to the elements, but I'm gonna have to unscrew it every time I take off the exhaust.

[Dean]

My LC-1 controller is just inside the firewall with the cable coming through one of the bulkhead grommets. Sensor is just behind the turbo in the DP basically under the side of the top mount.

Sensor must be removed to remove DP.

[sperry]

Originally Posted by Dean

My LC-1 controller is just inside the firewall with the cable coming through one of the bulkhead grommets. Sensor is just behind the turbo in the DP basically under the side of the top mount.

Sensor must be removed to remove DP.

So that whole "the sensor ideally should be 2-3 feet from the end of the exhaust, 6" after the turbo is *not* a good location" recommendation got ignored?

http://www.innovatemotorsports.com/v...r_content.html

[Dean]

Originally Posted by sperry

So that whole "the sensor ideally should be 2-3 feet from the end of the exhaust, 6" after the turbo is *not* a good location" recommendation got ignored?

http://www.innovatemotorsports.com/v...r_content.html

That is where the bung in the DP already was and it has to be in the DP somewhere as I still have a cat after the DP. If I get a new exhaust, it will likely move down the DP further. I did make one of the large copper heatsinks as described in the manual which should be getting plenty of airflow being right under the IC.

Replacement sensors are about $50... It is a standard VW part.

[sperry]

Originally Posted by Dean

That is where the bung in the DP already was and it has to be in the DP somewhere as I still have a cat after the DP. If I get a new exhaust, it will likely move down the DP further. I did make one of the large copper heatsinks as described in the manual which should be getting plenty of airflow being right under the IC.

Replacement sensors are about $50... It is a standard VW part.

I wasn't really concerned with the location... I figure, if the stock narrowband can survive in the header 6" from the block for 100,000 miles, the WB can survive anywhere in the DP... but the hotter it runs, the more often you should pull it out and recalibrate it to free air. As long as the EGTs are <1300F and pipe temps are <900F it's within operating range, right? Either way, you gotta be ahead of the cat... since I'm catless, I just used the rear O2 bung that was already in the back of the DP for my WEGO... then had a second bung welded right next to it for the Bosch.

And Greg was telling me the Bosch WB are like $38 from the dealerships. That's cheaper than a narrowband sensor for the SVX.

[Double Phister]

I concur with the LC-1 vote.

My controller box thingy is mounted to the side of the transmission brace. Been that way since the LC-1 came out. No problems. It's a pretty fast reading setup. I use the XD-1 digital gauge (replaced by the XD-16).
One of the analog outs is setup for narrow band O2 sensor simulation and wired to the ECU. LC-1 is in place of the Stock rear O2 at the end of the downpipe.



I think the issue with the rear O2 sensor trim problem people were/are having is only related to the 32bit ECUs.

[sperry]

I noticed that the analog out 1 simulates a narrowband... I wasn't going to bother to wire it up to the ECU... but that may be an option if the car drives oddly. However, changes are with the limited use I'm going to be getting on the stock ECU, ever if things are "off" it's probably not an issue anyway.