DaveT

Here is where I got some: http://www.por15.com/

I have had good luck buying used ones from a local Subaru specialist repair / used car shop. Buy boot kits, clean & regrease, etc.. My other source has been parts cars. As far as replacing the other one, only of the boots look like they are as bad as the one that failed. You don't get to save much work doing both together.

This made me think of the temperature sensor. It would screw up the ECU mixture / etc. If it were reading "normal operating temp" constantly. I don't get the smoke without finding melted insulation somewhere. Kind of hard to get smoke w/o burning something.

Moving the harness causing a change. I had a problem with one of my EA-82s a while back. After traceing & testing, etc. I found that one wire in the harness was intermittant. Bend it one way, Ok, bend the other way, open! I un taped, looked over the wire in question. Nothing obvious. But I felt a funny spot - where the wire, inside the insulation, was broken off! All the strands, at the same place. I spliced the bad spot, and all's well.

If you mean 40 Meg ohms, it will be an error - open circuit. Find the pin in the connector for the ECU that fires the solenoid in question. check with an ohm meter that you read the resistance of the solenoid or "test" resistor. Risky if you make a mistake: apply 12V to the wire, listen for the solenoid to click.

I have to say, I never thought of cleaning the MAF until I happened to see this thread. I gave the MAF in my 90 a shot of brake cleaner. The engine seems to run better cold - it was almost stalling out when put in gear (3AT), and stumbling until warmed up. After the cleaning, the engine seems much smoother. Almost seems like I have to use a little less throttle while crusing down the highway. It seems to have smoothed out my 92 also.

At about the center of the side facing the driver's seat, there is a hole in the sheet metal sheild of the ECU. The LED is visible through the hole, but it does not stick out or point out the hole.

Take a look at the lower arms, etc. for damage, bent, etc. The only adjustment is the tie rod ends. There is a jam nut against the tie rod end, and a flat on the tie rod. Loosten the nut by wrenching the end & nut. Adjust the toe in / out by turning the tie rod. Once you do a turn or 2, check the tape measures again to get an idea of which way / how far to go. If you adjust both sides, you can get the steering wheel to be straight when the car is going straight. It will probably take a test drive or 2 to get that right. Tighten the jam nuts before driving.

Here's a variant on the original: Would it be better to: Add a seperate oil cooler. Find a bigger radiator to make the cooling not so marginal. Find a way to modify the front / grill / etc to let more air flow throught the existing rad?

I had alingment trouble a while back. I found this worked pretty well: 4 pieces of 4x4 bocks. 2 pieces of unistrut (or other straight bar, with moderate weight). 2 tape measures. Check air pressure in tires. Tires the same make / model / miles. Drive car straight forward to a gentle stop on flat smooth surface. Put the blocks just ahead and behind each front wheel. Lift the bar / strut by the center, and place it on the blocks, center aligned with the center of the axle. Push on the center of the bar, against the tire. Do this a few times to make sure it is evenly against the front and back edge of the tire. Place a bar on each side. Hook the tape measures on the bar on the far side, just touching the leading & trailing edge of the tire. Read the 2 measurements. They should be very close, like under 1/8". The FSM actually has specific ranges. If you use long bars (Unistrut comes 10' long) you can check how the rear wheels align with the fronts also. I found using the 10' unistrut even for the front check handy, because the weight of them kept them from moving around unintensionally while fiddleing with the tape measures.

Before ripping apart the harness, look at this thread: Test LED - for intermittant troubles http://www.ultimatesubaru.org/forum/showthread.php?t=64831&highlight=test I have used this technique to find this intermittant starting problem. On one car, it was a pin in a blue 3 conductor connector under the dash on the passenger side. Another was the contacts in the starter, and dirt on the disk in the starter contactor. I had more than one used starter with worn contacts and dirty disks, so they need to be checked.

Make sure all 4 tires are the same brand, size, and miles on them. If you start with 4 new tires and never rotate them, about half way through their usefull life, the front ones will be smaller enough to cause noticeable drag also.

Man, you got it stuck good! How about a drill made for drilling bearing races? I forgot the exact name but I have seen them in the MSC catalog. http://www1.mscdirect.com/CGI/NNSRHM I think they were carbide. Maybe a diamond drill?

Ok, rebuild the body with 316 stainless steel. (or maybe another alloy, if there is abetter one) Rebuild all the steel suspension parts w/ stainless also. Do something about the @#%$^@# timing belts - wider or replace w/chain. Or push rods. My EA81 NEVER broke down and stranded me on the road. Or build a 6 cyl turbo Diesel boxer. Maybe make it about 150HP. And whatever changes to the transmission ratios to deal with the lower RPMs. Also maybe electric would be fun - not a typical wimpy one - one like this: http://www.wired.com/wired/archive/7.03/drag.html?pg=1&topic=&topic_set=

On the rod that goes from the top of the strut to the mount in the top of the wheel well there is a hard plastic / rubber bumper. I found one had broken loose and was rattling. It can be hard to see, there is a plastic shield over the piston rod, and all this is inside the spring.

Yes, the resistor is the means of varying the speed. There are 3 resistors connected end to end. One end connects to the full speed position of the switch and the +12V. The far end is the lowest speed, the 2 taps in the string are the middle speeds. To devide in between, you would have to add taps into the coils of the resistors. A switch will always have steps for each speed. If you get a switch with lots of positions, it can get to the point of being very little difference between such a switch and a rheostat / infinitly variable controller. Of course, such a switch would not be cheap, and you need more resistors.

Just adding to make sure we are all on the same page: The earlier bad cat code (I don't remeber the number any more) was like 2 years ago, and went away when I replaced the cat. The mechanic at the place I got the car did a bunch of checking and was in agreement to replace the cat. I guess what you are saying, if you are referring to this earlier problem, is that the rear O2 sensor could be out of calibration, making the ECU think the cat was bad?

Some more electronics / heater blower details: The electric motor that turns the blower is a 12Vold DC motor. What ever DC is applied to the 2 wires that connect to the motor will make it spin. 2 volts it spins real slow, 12 volts, it spins real fast. Anything in between, it will be in between. The resistor block is the lowest cost way to make selectable speeds. A rheostat (power variable resistor) could be used, but they are big and expensive at the power level required for the blower. I looked up a quick guess one. 3" diameter behind the panel. Cost $40.00 if you buy 100. Probably would run hot enough to melt plastic at some settings. But that one can't handle the 10Amps max current. For that, you need the $360.00 per 100 pcs unit. The heater core is fed water from the engine, bypassing the cooling system thermostat. The temperature lever on the dash moves a big flap that forces the air in the blower box through the heater core or around it. A 555 with a BIG transistor would work, but not be much better than a linear regulator. My initial variable speed controller was similar to that. A switched mode controller is much more efficent. (Smaller heat sink) Go here, look up buck mode / simple switcher: http://www.national.com/ specifically: http://www.national.com/pf/LM/LM2673.html If the link doesn't work, search for LM2673. This chip is too small for the blower motor. It can draw 10 or 20 Amps at full speed. But it is a place to start if you want to see how a switcher works. For anyone interested in electronics, from beginner to advanced: http://www.nutsvolts.com/ If there is more interest in this, I can look for the schematics for the controller I built. If I remeber correctly, many of the parts can be scavenged from PC power supplies.

Yes, it has the before and after. I learned about that a while (year or 2) back, when the CEL came on with the code that says the cat isn't working good enough. I got the new cat at cost since the car was just out of warranty (used, not new). I replaced the front sensor for this new trouble code, as the FSM states. So far, the CEL is still off, at least another 2 cycles on the car. I initially thought this OBDII stuff would be an improvemnet over the basic system in my loyales, but I am beginning to think that this is not so. I am going to keep my EA82's running as long as possible.

[answering several question] I didn't check the wiring - I figured that an open or short would have given me one of the other codes for those conditions. I replaced the original with a Bosch from NAPA. It was around $80.00. It looks the same, but that doesn't mean much. The engine has not been burning oil. Not sure if I'm being paranoid, but maybe sometimes not as smooth as I remember. My wife drives the car 90% of the time, so I may have to "take it over" for a few days. I reset the CEL the night I put up the original post, it isn't on yet today. At least 2 start / run /stop cycles.

The biggest one is 0.5 ohm. The second is 0.68 ohm. On the block I have handy, the third is broken. The one I fixed, I just found a similar length of similar sized nichrome. Probably from an EA81 blower, because I had several of them before moving to the EA82 cars. The mud condo is probably a mud wasp nest. A linear voltage regulator would need a big heat sink. The nichrome resistors are all black and crusty because they run hot. That means a lot of power is dissapated in them. A linear regulator would need to dissapte the same amount of power, but has to be kept much cooler. I did make an infinitly variable switched mode speed controller for one of these blowers I used for another application where high efficiency at lower speeds mattered.

96K I added a significant detail to the 1st post - I replaced the sensor. Then the CEL came back on today.

A while back, I got the CEL, read the codes. Only the P0133 showed up. Whent through the FSM procedure, ends with "replace O2 sensor". [edit] I replaced the sensor. A week or so later, the CEL comes back on, same trouble code. Same result from the FSM.. Any ideas? I forgot to mention: 2001 forrester

I have a spare one here, pulled the cover off. A piece of plastic has 4 brass tabs molded into it. the 3 nichrome reistors are wired in a string from 1 tab, to the next, next.... 2 tabs are at the extreme ends (highest & lowest speeds). 2 tabs are along the string. Each of them has 2 resistor ends connected to the tab. My original post was refering to the rsistors themselves breaking. Not likely to loose only 1 speed (except the lowest) by 1 resistor failing. And also, from the perspective of repairing the block, or wire. A wire that goes from the block to the connector could cause only 1 middle speed to fail. Probably best to remove it and take a look. You could test it with an ohm meter without removing the block, just pull the 4 pin plug. Expect the resistances to be pretty low.

Here is what you need: 1) T-1 3/4 super bright LED 1) 1200 ohm 1/2 Watt resistor A few feet of 22 AWG wire some heat shrink tubing Solder about 3' of wire onto the Cathode of the LED. This is the ground wire. Solder 4-6' of wire to the Anode of the LED. Solder the resistor to the end of the long wire. Leave the other lead of the resistor about an inch long. To use the tester: Tape the LED on the dashboard where you can see it. Run & tape the wires so they don't get in the way of operating the car. Connect the ground wire to any convienient grounded fastener. The resistor lead is a good size to sneak into most of the connetors on an EA82 & older Subaru. It works on the ECU connectors and the ones with 1/4" flat tabs. I have used this to figure out numerous electrical problems over the years. Intermitant no crank - Start by slipping the resistor into the signal connector on the back of the starter. If the LED lights when you turn the key, you know that everything from the connector back is good. If it doesn't move it to another of the 3 or so in the harnes back towords the key switch. I used it to monitor the EGR valve computer signal a while back. If I remeber correctly, the EGR solinoid is activated until the engine reaches normal operating temperature. This disables the EGR valve while the engine is cold. The LED is great for these intermittant troubles, because you can leave it connected, and check the function every time you use the car.