swiftt Posted July 13, 2007 Share Posted July 13, 2007 Hi all. I have a 2000 OBS that's misfiring like crazy and shows P0302 on the ODBII diagnostics (misfire on cylinder 2). It sometimes shows P0420 (inefficient CAT) but I think this code is due to the misfire. I've searched the board but nothing addresses the situation, specific to what I've done/seen to date. So here's what I know so far: -All cylinders have spark all the way to the plugs -The fuel rails are getting fuel -All injetors measured within the same range for internal resistance (can't remember what the range was off the top of my head but I remember not worrying about it 'cause they were all close) -Injectors 1, 3, and 4 are switched between .25 and .5 volts with the engine running. -Injector 2 is switched between 5.0 and 5.5 volts with the engine running. Ugh! So, not knowing for sure, I'm thinking the injector fires when the line voltage drops to 0.25Volts and then closes again at 0.5volts. With the outrageously high voltage on cylinder #2, I'm thinking it's simply holding the injector closed and starving the cylinder of fuel. Anyone have a critique of my thinking here? Does anyone have any ideas what may cause this high line voltage on cylinder #2? I can't help but think this is the root of my problem. My initial thought was one of the two O2 sensors, but I'd expect that to effect two, or more, cylinders and not just one. Am I right in thinking this? Anyone else have any thoughts? Thanks in advance. -Dave Quote Link to comment Share on other sites More sharing options...
nipper Posted July 13, 2007 Share Posted July 13, 2007 Hi all.I have a 2000 OBS that's misfiring like crazy and shows P0302 on the ODBII diagnostics (misfire on cylinder 2). It sometimes shows P0420 (inefficient CAT) but I think this code is due to the misfire. I've searched the board but nothing addresses the situation, specific to what I've done/seen to date. So here's what I know so far: -All cylinders have spark all the way to the plugs -The fuel rails are getting fuel -All injetors measured within the same range for internal resistance (can't remember what the range was off the top of my head but I remember not worrying about it 'cause they were all close) -Injectors 1, 3, and 4 are switched between .25 and .5 volts with the engine running. -Injector 2 is switched between 5.0 and 5.5 volts with the engine running. Ugh! So, not knowing for sure, I'm thinking the injector fires when the line voltage drops to 0.25Volts and then closes again at 0.5volts. With the outrageously high voltage on cylinder #2, I'm thinking it's simply holding the injector closed and starving the cylinder of fuel. Anyone have a critique of my thinking here? Does anyone have any ideas what may cause this high line voltage on cylinder #2? I can't help but think this is the root of my problem. My initial thought was one of the two O2 sensors, but I'd expect that to effect two, or more, cylinders and not just one. Am I right in thinking this? Anyone else have any thoughts? Thanks in advance. -Dave How are you measuring the injector voltage? A digital meter wont pick it up as the sampling time is not fast enough. You do know that one side of the injector is high, and the low side goes to the ECU. nipper Quote Link to comment Share on other sites More sharing options...
swiftt Posted July 13, 2007 Author Share Posted July 13, 2007 How are you measuring the injector voltage? A digital meter wont pick it up as the sampling time is not fast enough. You do know that one side of the injector is high, and the low side goes to the ECU. nipper Yeah, I used a digital multi-meter. I uplugged the injectors, one at a time, with the engine running and measured the voltage across the line that plugs into the injector (ie. one lead connected to each of the two incoming wires, respectiely). I did get consistent readings so I'm not sure that I follow you. I was able to watch the votlage switch up/down on it so I thought that a pretty clear indication of the ECU signal being sent to the injector. The large voltage discrepency between cylinder #2 and the rest was also pretty extreme. I went around and did this twice and the results were repeatable.All that said, what would be a better method of testing? Why would the ECU send a significantly higher voltage to one cylinder? I can switch a couple injectors around to see if the problem follows but that looks to be a PITA, so I'd rather pin the problem down prior to pulling fuel rails and injectors.Any thoughts? Thanks again.-Dave Quote Link to comment Share on other sites More sharing options...
Cougar Posted July 13, 2007 Share Posted July 13, 2007 Going by the good readings it sounds like you are measuring the voltage on the ECU side of the circuit and to ground. The ECU makes the ground for injector circuit so if you are seeing 5 volts there, then there is either a problem with the external connection to the ECU or the ECU has a internal problem for that injector. Edit: You posted while I was typing my reply. You have a grounding problem to the ECU like I stated above. Quote Link to comment Share on other sites More sharing options...
nipper Posted July 13, 2007 Share Posted July 13, 2007 You need a analog meter to read the voltage. Digital volt meter take a sample every X milliseconds, since it has to process the data. The more expensive the meter, the faster the sampling rate. Also every time a digital gets an input, it triggers the processor, so it removes the previous datapoint. It is possible that you can get data faster then it can process. Digital is not the end all of meters and never will be. There are times that you need a needle, or a scope. I always have a cheap analog and a expensive digital. Also the injector timing is affected by the cam position sensor. nipper Edit : slow typist Quote Link to comment Share on other sites More sharing options...
swiftt Posted July 13, 2007 Author Share Posted July 13, 2007 First, thank you both for your thoughtful replies. Just a couple questions it raises for me... -I had considered the cam position sensor when initially looking into this. I unplugged it while the engine was running and idle went to hell. Plugged it back in, the idle raced for a moment, then it settled back into a regular idle (albeit a 'regular' idle wiht one dead cylinder). I'm thinking if that sensor failed, timing would be effected globally and wouldn't be isolated to #2. Yeah? -With regard to the ECU ground, wouldn't that behave the same? As in, if the ECU weren't properly grounded, wouldn't I notice all kinds of odd behavior vs. a single cylinder misfire? ...Or... Is each injector circuit grounded independently? If it's this latter scenario, where would one physically look for these grounds? Incidentally, there was a ground I found that wasn't even connected. I found it hanging down, inside the transmission tunnel as viewed from under the hood. I connected it to the firewall right behind the airbox but alas, no changes to the misfire issue. Maybe my stereo will play louder w/o distortion now though...? -Dave Quote Link to comment Share on other sites More sharing options...
nipper Posted July 13, 2007 Share Posted July 13, 2007 Well we can say the cam position sensor is good. Also each individual injector grounds out through the ECU. Grounds can do really weird things. nipper Quote Link to comment Share on other sites More sharing options...
swiftt Posted July 13, 2007 Author Share Posted July 13, 2007 Well we can say the cam position sensor is good. Also each individual injector grounds out through the ECU. Grounds can do really weird things. nipper Yeah, like kill cylinders...:-\ So if each individual injector grounds through the ECU, I assume it will be enough to verify the ECU ground before moving on and looking elsewhere? Is there some sort of test I can run on the ECU to isolate the injector circuits in turn? I have a generic ODBII code reader but nothing Subaru specific that I can 'talk' to the ECU with. It's unfortunate, but Subaru seems to keep their workshop info and specialty tools pretty close to their vest (or prohibitively expensive). -Dave Quote Link to comment Share on other sites More sharing options...
nipper Posted July 13, 2007 Share Posted July 13, 2007 no. They ground through the ecu, thats how they operate. The ecu completes the circuit, and they energize. There is a tester you can buy at the autoparts store to test injectors while the car is running. It tells you that there is an electrical pulse. Its a noid light. Here is some tmi http://www.asashop.org/autoinc/dec2000/mech.htm noid light http://www.drivewerks.com/catalog/ShopCart/TOOL/POR_TOOL_CAT447_pg28.htm Quote Link to comment Share on other sites More sharing options...
swiftt Posted July 13, 2007 Author Share Posted July 13, 2007 no. They ground through the ecu, thats how they operate. The ecu completes the circuit, and they energize. There is a tester you can buy at the autoparts store to test injectors while the car is running. It tells you that there is an electrical pulse. Its a noid light. Here is some tmi http://www.asashop.org/autoinc/dec2000/mech.htm noid light http://www.drivewerks.com/catalog/ShopCart/TOOL/POR_TOOL_CAT447_pg28.htm I see. Very cool lead on the tool. I appreciate the insight. I'll see if I can track one down in the next day or so and give it a go. Thanks again. I'll let you know how it turns out. -Dave Quote Link to comment Share on other sites More sharing options...
Cougar Posted July 13, 2007 Share Posted July 13, 2007 You should also be able to see it with your meter, like you did previously. Check the voltage to ground on one of the good injector return wires and then compare it with the bad one. If the the voltage is higher on the bad one then the problem is internal to the ECU. Quote Link to comment Share on other sites More sharing options...
ron917 Posted July 13, 2007 Share Posted July 13, 2007 I wouldn't be so quick to condemn the ECU. There are wires and connectors between the injector and ECU that could cause the problem. If a noid light, scope, and/or analog meter confirm that the signal at the injector is bad, then check the signal for that injector at the ECU. You'll have to find the wiring diagram, and check it at the ECU connnector. Check for continuity from the ECU connector pin to the injector connector pin. When I had similar problems twice in the past (on Volvos), once it was the connector at the injector, and once it was the ECU itself. In that case, I was able to swap in an ECU from another car to confirm the diagnosis, but we don't always have that luxury. I don't like to see people swap expensive parts without absolutely confirming the diagnosis. Quote Link to comment Share on other sites More sharing options...
nipper Posted July 13, 2007 Share Posted July 13, 2007 Lets try something simple. Unplug and then plug back in the connector at the harness. If its a dirt issue, that is usually enough to clean up the connections. nipper Quote Link to comment Share on other sites More sharing options...
swiftt Posted July 13, 2007 Author Share Posted July 13, 2007 Lets try something simple. Unplug and then plug back in the connector at the harness. If its a dirt issue, that is usually enough to clean up the connections. nipper Would this be one of the two harness plugs located beneath the intake, on the passenger side? ...Right in front of the airbox? Thanks, -Dave Quote Link to comment Share on other sites More sharing options...
Cougar Posted July 13, 2007 Share Posted July 13, 2007 I wouldn't be so quick to condemn the ECU. There are wires and connectors between the injector and ECU that could cause the problem. If a noid light, scope, and/or analog meter confirm that the signal at the injector is bad, then check the signal for that injector at the ECU. You'll have to find the wiring diagram, and check it at the ECU connnector. Check for continuity from the ECU connector pin to the injector connector pin. When I had similar problems twice in the past (on Volvos), once it was the connector at the injector, and once it was the ECU itself. In that case, I was able to swap in an ECU from another car to confirm the diagnosis, but we don't always have that luxury. I don't like to see people swap expensive parts without absolutely confirming the diagnosis. I agree. The ECU is the last item you want to call bad and wasn't saying it was bad. In my previous posts I wanted the testing done at the pins on the ECU. This way you will know which direction to look for the trouble. I didn't explain that in my directions for testing very well. Quote Link to comment Share on other sites More sharing options...
swiftt Posted July 13, 2007 Author Share Posted July 13, 2007 So I picked up a set of NOID lights and tested. All indications are that the signal is there and behaving as one would expect with the NOID light cylcing off/on while engine is running. Now I'm thinking clogged injector, perhaps? Anyone willing to go out on a limb and reassure me before I drop some coin on new injectors and seals(I'm thinking it would be best to replace all at the same time. Thoughts?)? -Dave Quote Link to comment Share on other sites More sharing options...
nipper Posted July 13, 2007 Share Posted July 13, 2007 DO the seals first. nipper Quote Link to comment Share on other sites More sharing options...
Cougar Posted July 13, 2007 Share Posted July 13, 2007 I am still wondering about those voltage readings. My info for a Legacy shows the pin for #2 injector goes to pin 13 of the ECU. Number 1 injector goes to pin 4 of the ECU. If you don't mind, would you check the voltage on those pins to ground and see if that is the same? If only some of the voltage is dropping across the injector it could cause a problem in opening. The Noid light may not tell you the whole story here. Quote Link to comment Share on other sites More sharing options...
srs_49 Posted July 14, 2007 Share Posted July 14, 2007 I have a general question about the injectors. Someone said that one side of the injector is energized at some voltage (5V?), and the ECU grounds the other side to open (turn on) the injector, right? I assume that the frequency that the injector is energized is directly proportional to the speed of the engine. Is the duration that the injector is energized fixed, or does it vary? I would think that a simple low pass filter (an RC network) might be able to be used to filter (average) the pulses on the ECU side of the injector and create a DC signal that was proportional to the engine speed and anything else that affects injector "On" time or frequency thereof. An anlog voltmeter, as Nipper was alluding to, essentially does this just due to the mechanical inertia of the meter movement. Digital voltmeters, because of their sampling nature, can give very strange readings when monitoring a fast time varying signal (such as the injector signal). Quote Link to comment Share on other sites More sharing options...
swiftt Posted July 14, 2007 Author Share Posted July 14, 2007 Got it. Yay! It ended up being a fouled injector. My car only has 85K miles on it, so it seems to have failed a little early but such is the nature of the machine. ...Unpredictable. Thanks to everyone for sharing their expertise. Best Regards, -Dave Quote Link to comment Share on other sites More sharing options...
Cougar Posted July 14, 2007 Share Posted July 14, 2007 Thanks for the update Swiftt. Glad you fixed it. Quote Link to comment Share on other sites More sharing options...
porcupine73 Posted July 14, 2007 Share Posted July 14, 2007 Excellent fix! It's nice when it's fixed. I have a general question about the injectors. Someone said that one side of the injector is energized at some voltage (5V?), and the ECU grounds the other side to open (turn on) the injector, right?I believe so, but I believe it is 12V, not 5V. This is why the ECU grounds many circuits for control. It uses an NPN transistor so the ECU can use it's internal 5V on the base as a signal and ground the circuit being controlled through the collector. This 'open collector' method is common in lots of devices; then the device can use a low voltage, as low as 1V even, to control a circuit running a higher voltage such as 24VDC (common in industrial/control systems). When the transistor is not forward biased the collector is esentially unhooked from the emitter, at least for DC. I assume that the frequency that the injector is energized is directly proportional to the speed of the engine. Is the duration that the injector is energized fixed, or does it vary?It is proportional, except each injector fires once per two turns of the crank. The ECU controls the injector on-time. It does vary. This is how it controls A/F ratio for the amount of air being drawn in and other parameters. The longer the injector is open at a given pressure the more fuel that is injected. The fuel pressure regulator increases the fuel pressure as the vacuum decreases (and engine load increases), so the same injector on-time at lower vacuum would inject more fuel. The RC network idea could work, but you'd need a diode involved too. The inexpensive analog meter is easy. Quote Link to comment Share on other sites More sharing options...
srs_49 Posted July 14, 2007 Share Posted July 14, 2007 Thanks for the info, porcupine73. Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.