Cougar

Hi Dan, Good talking with you also. Since you got no reading on either of the two sensor wires to ground that means that the wires are isolated from ground, which is good. There shouldn't be any connection that way. The 1.5 ohms reading you got between the ground points seems ok but I was hoping for less than 1 ohm. One other reading I would like you to make is measuring the engine ground to the dash ground. Place one probe to a good metal spot on the engine and the other probe to a place under the dash. You may need a jumper wire to do this to extend the probe wire. You should see near zero ohms if things are good. One last test is to check the resistance between the sensor wires with the ends isolated still. You should have infinate resistance. This means the wires are not tied together somewhere.

Tim, If I remember correctly that relay is also fed from the ignition relay. I would also check the fusible links for a loose connection. You have a bad connection somewhere in the circuit.

The relay supplies power to the pump. It sounds to me you either have an intermitttant power problem to the pump or possibly a problem with the injector or the circuit to it.

Welcome to the forum. There are a lot of dedicated Soob owners here. The car you talk about may work out ok but it has a lot of miles on it for the age of the vehicle. Since you plan on adding a lot more I would try to find something with around 40k miles if you can find one. I assume you can't afford the payments for a new vehicle. There are some good buying incentives going on if you can swing the payments. If you do get the '03 Forrester I would suspect the engine would be the first area of trouble. Head gaskets have been an issue for some. I would try to save what you can in a repair fund from each of your paychecks to try and be ready for any future problems. If you can afford to save $50-75 dollars a month it will add up pretty quick.

Dan, Sent you an email.

Dan, You asked what the meter should read when measuring a short. Almost anytime you have a meter reading less than 1 ohm you can consider that a short. A piece of wire is basically zero ohms. When you see very low resistance between whatever you are measuring across it means that the connection is tied together. You stated that you took a second measurement of each of the sensor leads to the ECU ground lead and got readings of .2 and .3 ohms. This means that both sensor leads are tied to ground somehow. Not good. Your leads from the sensor to the ECU are good because you got a low reading on each wire but it appears both wires are also connected to ground somehow. Did you isolate both ends of the harness when you took the measurement? Take another reading of pins 2 and 70 while the harness is isolated. Put one of the probes on either of the pins and put the other meter probe to a good chassis ground point. Do you still see a short there? If so then you will need to track the problem down. Even if that is ok, you need to see why the leads are tied to the ECU common. I'm not sure why the RPM's are showing a reading. It may not mean anything or it could mean some other trouble. You should get the sensor wires fixed first.

It sounds like they are trying to measure the resistance of the harness going through the ECU. I assume that the connector on that end was still connected. This is not a real good test to me if that is what is really going on. You stated that the sensor pins tie to pins 2 and 70 of the ECU if I remember correctly. If that is right then lets test each wire from both ends of the harness. First remove the sensor connection and ECU connector so that the harness is isolated. Place the meter probe on one of the sensor connections on the harness and then check the ECU harness end. You should see close to a short on one of the wires if the harness is good. Do the same thing on the other wire. You also should not see any resistance to ground on either of the wire connections or any resistance between the two wires going to the sensor.

Your welcome.

You may have an intermittant connection at the coil pack. I doubt that one plug wire is bad and you just happened on it. The problem may be internal to the coil also.

Sticking with NGK will keep things going well.

Looking at a diagram for a '92 model the power is supplied to the pump via the fuel pump relay that is tied to fuses MB-3 and SFB-2. I assume that will be the same for your year.

I think you are correct about that. Swapping parts should be no problem.

I would do the normal tuneup things and check for any vacuum or intake manifold leaks.

I kind of doubt that the timing would do this (unless it is way off). Even though the sensor had a presumably good resistance reading it is possible that it may not work still. The proof will be seeing those pulses while cranking the engine. An oscilliscope would be the best thing to look at the pulses with but not many people have one of those handy. An analog meter would be the next best choice or a dwell meter should work also. A digital meter that has a frequency mode would also work.

Replacing the knock sensor did help with this same type of problem for one of the memebers here.

We can probably assume that the resistance you read is normal. The real test would be to see if there are pulse signals coming from the sensor while cranking the engine. The other pins you mention and took readings on may have no bearing at all on this problem. I'm not sure what is tied to those pins. The main thing is, you want those pulse signals from the sensor getting to the ECU. You may be able to see the pulses by setting a digital meter to AC volts and take a reading across the pins. An analog meter would see the pulses also. The 12 volts you are seeing is normal. That voltage will pulse as the sensor sees the tabs of the pulley go by.

The TPS may have dirty contacts. You may want to place an ohmmeter across the the wiper contact and the ground contact and then see how smooth the reading is as you move the throttle. Using an analog meter will be better for this if you have one available. If the contacts are bad, the reading will be erratic. You could try cleaning the contacts if you can get to them. You may have to replace the TPS also.

Your welcome for the help Dan. Keep us posted. You can send me the document by PM if you want.

Yes, that is correct. We are trying to determine the resistance of the sensor. I assume there is two leads from the sensor that tie to pins 2 and 7 of the ECU and that will allow the meter to read the resistance of the sensor. The wire leads to the sensor will be included in the measurement but that resistance will be insignificant as we saw previously. By disconnecting the harness from the ECU that will isolate any error to the reading if the connector was tied to the ECU. Your meter will see the resistance of the sensor that the ECU sees when the connector is normally in place.

Dan, In answer to your question about the wire resistance, yes the resistance should be low. When you measured the wire resistance you were measuring a short really. The wire will have some measureable resistance but it is so low that it really is a short. From your question about the resistance you saw in the sensor wire it sounds like you just measured the wire resistance from the sensor to the ECU, is that correct? If so, you need to measure the resistance between (across) the two sensor wires with the ECU end disconnected. You should see a real low resistance in each of the wires to the ECU but it is the resistance across the sensor that we need to know. Hope it warms up there.

Perhaps it may just be a bad connection to the sensor causing this. I would check the wiring for a problem. The sensor could have had a premature failure though. Does it have a warranty?

While following Rallyruss's good advice be sure that everything is off in the car, like it would be parked overnight. I would be concerned if the load amperage draw is more than 80 milliamps. Usually these kind of problems can be traced to added accessories to the car or devices like a relay that get power supplied to them somehow when they shouldn't be on.

It sounds like you did things correctly. Let us know what you find out, and your welcome for the help.

Another test you might want to try is checking the resistance of the sensor leads to ground. You may have a short to there also. Place one of the meter leads to either of the sensor leads and the other meter lead to engine ground. You should see a high resistance on the meter. Be sure to remove the plug that connects the sensor to the ECU while doing this so you don't get a bad reading on the meter. I assume you did this also for the reading you took already. Using you meter as a guide it should be pretty easy to find the short and repair it.

I think you can program the system to sound the horn or not sound it. It appears you will need to reprogram the system. Check your owners manual for details on doing this.