OB99W

The 16 flashes just indicate there's an electrically-related trans problem, and that one or more codes are stored in the TCU. While a C duty solenoid fault might be common, it isn't the only possibility, and reading the code(s) is the way to find out.

It means that there's at least one trans code to be read. Look at the links I gave.

Hmmm, only eight? Anyway, see the following to retrieve trans codes: http://www.subaru-svx.net/forum/showthread.php?t=54004 http://www.svx-iw.com/svxiw/article/?article=79

16 flashes, by any chance?

The engine apparently seems "happier" above idle, although perhaps not "ideal". Certainly wear might have changed valve clearances by now, although it's hard to say to what degree without checking. At it's current mileage, the engine probably should have had two valve adjustments. I don't know if it's had any. I don't believe the pressure was rechecked after the FPR replacement. I did suggest to 2000 Legacy that the vacuum at the FPR be checked, and he also reported that the fuel return line seemed clear to the tank (tested by air passage). However, verifying that the new FPR has brought the pressure within specs is a reasonable thing to do.

Yes, those things could all potentially result in an incorrect determination that could cause the ECU to improperly set fuel trim. However, in the case at hand we've verified that both the MAP and IAT sensors appear to be working and producing data that's within tolerance, an FPR that regulated the fuel pressure too high has been replaced, and the front A/F sensor has been changed. That would leave the ECU's determination to be at fault, if we assumed it was the only remaining possibility, but ECU failures are rare. It's correct for cars like the 2000 Legacy that don't use that sensor for anything more, although others do. And just because it's purpose might be primarily to determine if the cat is functioning well doesn't mean we can't use the voltage reading for diagnostic purposes. If the rear O2 sensor voltage remains somewhat above 0.45 volts, and doesn't fluctuate much, that's usually a good sign. All other things performing correctly, it would indicate that there wasn't excessive oxygen in the exhaust after the cat was done with it, so that likely CO was being oxidized to CO2, etc. But a rich mixture would also be lacking oxygen, and could drive the sensor above 0.45 volts as well. Notice that the rear O2 sensor has been at about 0.85 volts at idle, and that with the MAP and IAT looking good and the FPR and A/F sensor replaced, the ECU is still trying to cut fuel (negative trims). From post #189, though, at 2500 RPM "O2S B1 S2 (V)" drops to "0.525 to about 0.710" volts, and the STFT goes positive (thanks for the additional data, 2000 Legacy). We've noted all along that the engine runs better at higher RPM, both in terms of drivability and what the live data shows. Many things have been ruled out either due to tests or replacement, and yet low speed operation is still not quite right. Since idle and slightly higher speeds are mainly controlled by IAC, parts associated with it should probably be looked at again. The system on the 2000 Legacy uses air-assisted injectors. With an auto trans, there's an air-assist solenoid to control things, but the manual trans version is different. At idle, air via the IACV is split into that needed for intake (bypassing the closed throttle plate) and the remainder, which is used for injector assistance. Since the IACV itself was replaced by the dealer without improving the symptoms, it's possible that there's a partial blockage in a throttle body passage (which I mentioned in post #94, suggesting looking at the P0506 procedure). Previous cleaning might not have been able to resolve this. It could be useful to find out what duty cycle the ECU is applying to the IACV. Unfortunately, that's not an OBD-II PID that the Legacy provides. It can either be read with a Select Monitor, or using an oscilloscope (which is what I'd do). If anyone has an idea about something we haven't already considered, please speak up -- I'm running out of suggestions.

Is that link a secret?

Sorry, time constraints. Hopefully everyone is still alive. There are numerous parameters associated with OBD-II data, such as O2 sensor voltage, long term fuel trim, etc. These parameters are standardized, and have been assigned identification numbers. Since a computer (ECU) is involved, the Parameter Identification (PID) numbers are in hexadecimal, a number base convenient for programming. (See http://en.wikipedia.org/wiki/Hexadecimal if you'd like more on "hex".) As I mentioned previously, at least one of the PIDs in the live data that Legacy 2000 was reporting had me wondering what was going on. For example, there was MAF data, even though there's no MAF sensor on a 2000 Legacy. So, I decided to speak to tech support at Innova (the manufacturer of the Equus 3130 scan tool). I explained about the phantom MAF reading (and other things), and was told that the scan tool just communicates with the ECU using PID numbers (and apparently then converts those into the equivalent "LTFT", "MAP", etc., displaying them that way for easier reading by average humans ). With that in mind, I looked at what was reported by 2000 Legacy in post #187 (I added the matching PID numbers): 03 Fuel Sys 1: CL, CL, CL & CL. 04 Calc Load (%): 5.4, 4.7, 4.3, 4.7 but then dropped to 4.3, then 3.9. 05 ECT (Degrees F): 190, 190, 186 & 195. 06 STFT B1(%): -18.7, -21.0, -17.9 & -18.7. 07 LTFT B1 (%): -1.5, -1.5, -1.5 & -1.5. 0B MAP )inHg): 14.7, 12.6, 12.1 & 11.5. 0C Eng RPM: 667, 673, 653 & 676. 0D Veh Speed: 0, 0, 0 & 0. 0E Spark Adv: 4, 5, 11 & 6. Note it did go as high as 12 but then drop as low as 4. 0F IAT (degrees F): 51, 64, 86 & 109. 10 MAF (1 lb./min): 0.682, 0.615, 0.575 & 0.511. 11 TPS (%): 0,0, 0 & 0. 15 O2S B1 S2 (V): 0.820, 0.850, 0.875 & 0.845. 15 STFT B1 S2 (%): 0,0, 0 & 0. 1C OBDSup: OBD2, OBD2, OBD2 & OBD2. 14 O2S B1 S1 (V): 3.845, 3.796, 3.767 & 3.679. 24 EQ Ratio 11: 0.973, 1.030, 0.999 & 1.011. The same PIDs, from the CanOBD2 (Innova) Website: 03 - Fuel System Status 04 - Calculated Load Value 05 - Engine Coolant Temperature 06 - STFT Bank 1, Bank 3 07 - LTFT Bank 1, Bank 3 0B - Intake Manifold Absolute Pressure 0C - Engine RPM 0D - Vehicle Speed Sensor 0E - Ignition Timing Advance for #1 Cylinder 0F - Intake Air Temperature 10 - Air Flow Rate from Mass Airflow Sensor (MAF) 11 - Absolute Throttle Position 14 - Oxygen Sensor 1 Bank 1 15 - Oxygen Sensor 2 Bank 1 1C - OBD2 Support Requirements 24 - Oxygen Sensor Equivalence Ratio/Voltage (bank 1, sensor 1) So, why is there MAF data? From the 2000 Legacy FSM: 03 Fuel system control status 04 Calculated engine load value 05 Engine coolant temperature 06 Short term fuel trim 07 Long term fuel trim 0B Intake manifold absolute pressure 0C Engine revolution 0D Vehicle speed 0E Ignition timing advance 10 Air flow rate from pressure sensor 11 Throttle valve opening angle 14 Oxygen sensor output voltage and short term fuel trim associated with oxygen sensor - bank 1 15 Oxygen sensor output voltage and short term fuel trim associated with oxygen sensor - bank 2 1C On-board diagnosis system Ahhh, PID#10, as Subaru defines it for the 2000 Legacy, is not "Air Flow Rate from Mass Airflow Sensor (MAF)", but "Air flow rate from pressure sensor". The 2000 Legacy uses a "speed density" system -- the ECU calculates air flow rate using engine speed (RPM), manifold absolute pressure (MAP) and intake air temperature (IAT), so what it's reporting as PID#10 is calculated by the ECU using that data, not read from a MAF sensor. In looking over the FSM PID list, I also noticed some errors and omissions: 0F - Intake air temperature (IAT) - missing from list, but reported in data 14 - doesn't mention this is for sensor 1 (front A/F sensor) 15 - should be bank 1, sensor 2 (rear O2 sensor)(*not* bank 2) 24 - Equivalence ratio bank 1, sensor 1 (EQ Ratio 11) - missing from list, but reported in data In addition, I spoke to Innova about the STFT B1 S2 (%) consistently being 0%. Apparently since PID#15 includes both the rear O2 sensor voltage and its effect on STFT, the 3130 reads and reports the voltage and the STFT info as well. If it displays 0%, that means the ECU doesn't use the rear O2 reading at all to affect STFT (but only to determine that the cat is doing its job). So, that clears up some of the mysteries of the live data readings. It's late and I'm tired, so I'll post again tomorrow with where I think this is all leading.

In many models the connector is black, 6-pin (not all the positions are necessarily populated).

Just as another point of reference, read this entire thread (both pages): http://www.ultimatesubaru.org/forum/showthread.php?t=93767

About the hardest part of the info I linked to in post #8 is finding the connector.

Download this: http://www.subaru-car.ru/for1/for_01_chas_11.pdf . See the section "Read Diagnostic Trouble Codes". The diagnosis connector and ground probes are typically taped to the harness under the dash, above the accelerator pedal.

The tooth count I gave in post #2 is from the factory manual, but if that doesn't seem to align, do what you have to. On the other hand, the 307 belt is the wrong one, according to Cloyes, it should be the 304. See the link in post #7.

Guesstimation.

Agreed -- the parts lookup at http://www.cloyes.com/default.aspx? tabid=378 results in: B304 Engine Timing Belt Subaru Impreza (Car) 2002 2.5L H4 2458cc; EJ251 Desig.; SOHC

See attachment -- © is mark on cam sprocket, (d) is matching surface on cylinder head

Info I have says 46.8 teeth to passenger side, 43.7 to driver's.

Intermittent "weirdness" with the ABS system is sometimes caused by a poor ground connection. You might want to check for corroded ground points near the ABS unit. Dirt/mud/filings on the wheel sensors or tone rings can also cause problems that come and go. Some ABS info: http://endwrench.com/current/summer02/insiderinfo.pdf http://endwrench.com/pdf/brakes/FtSubaruABSS99.pdf http://www.furi-dorifto.com/tech/Subaru_dox/ABSEWFall05.pdf

Agreed, I got the impression that the OP was set on removing the condenser, no matter what anyone else said. Not only would that cripple defrost/defog, but there's also the issue of refrigerant recovery.

The common cause is a stuck relay. See http://www.furi-dorifto.com/tech/Subaru_dox/ABSInsiderEWFall05.pdf

See http://www.ultimatesubaru.org/forum/showthread.php?t=116111 Is the dash ABS warning light lit? If so, codes can be retrieved that may help diagnose the problem.

It was also a perfect excuse to use this smiley:

Congrats! It would probably be a good idea to make the fan wiring repair permanent now. Hot antifreeze typically has that kind of aroma. It's nothing to be concerned about if it doesn't return. It would be a good idea to keep tabs on the coolant level for a while, just to make sure that if there were any air pockets to fill, the level doesn't drop too much. Why do you think the pistons (etc.) need cleaning?

Thanks for posting that. I've been very busy, and may not be able to respond to what I see in the data for a few days. However, at least one of the PIDs interested me enough to call tech support at Innova. When I post next, I'll explain and have some comments and further suggestions. Until then, would you please run a live data check once more? Fully warm the engine (ECT = 190 or so). Bring the engine speed up to 2500 rpm, and hold it there for a full two minutes. With it still at 2500 RPM, note O2S B1 S2 (V) (rear oxygen sensor voltage). If it's fluctuating, make note of the range of values. If steady, note that reading. Thanks.

75 deg C = 167 deg F. What is that the temperature of? Did you mean 75 F? Having the hood open might not allow the temperature to get sufficiently high for the fan(s) to come on, especially if you're not waiting long enough. The gauge reading about half of scale is normal, and the fans might not come on at that point. The smoke could just be burn-off of oil (etc?) from the work that was done. If the fans ran with the test connectors coupled, they're probably okay. Are you sure that the fuse is getting warm?