SVX can't climb hills

[NiftySVX]

Ah I was confused.

I was patting myself on the back and congratulating myself on still being useful after all these years despite being told otherwise by a few who don't like my ideas about transmissions.

Really, really, really like the exhaust diagnosis on the car. Rattling is not good. As I am sure you know, there should be three catalysts on the car. Most don't realize that these are a system and they work in tandem. Aftermarket cats are notorious for causing problems, they are often worse than the ones they are replacing. Can you show a picture of your center exhaust section with the catalyst system? If the "main" cats are aftermarket (which I am assuming is the case) then all I can say is that you need to rip that um….poo off your car and take it to a recycler.

My official recommendation is, at this point, all aftermarket components of the catalyst system be immediately replaced with factory parts. If your car was in my bay, this would have happened as soon as I saw they were aftermarket. Admittedly, I spent my "real" years at a new car dealership because this is how I like to fix cars. If your car comes to me, I will assume that you want it to leave as it was originally engineered, and I have taken a lot of flack in the past ten years (EDIT: Ok 15 years. I'm getting old) in this community because I am of that mindset. So it looks like "my" estimate for your repair is already at 1200 in parts(end disclaimer/recommendation )

I would not go any further with diagnosis work until the exhaust is replaced. Since I am currently distracting myself from my studies, I will now go into a lengthy discussion of the catalyst system themselves. Please feel free to roll your eyes and skip this paragraph. :clears throat: A catalyst is a device that is made of precious metals and some other expensive stuff bonded together into a sort of honeycomb/waffle type of deal that is sort of like a filter. Conceptually, think of it as an oxygen bank. As the engine runs normally, the ECU switches short term fuel trim from rich to lean so that there is excess oxygen flowing past our precious metal waffle. These precious oxygen molecules are captured and put in our waffle bank. Assuming our engine is running correctly, there is no problem getting this oxygen in that manner. We'd love to have simple oxygen to take (well, it's O2 because oxygen is homonucleaerdiatomic but I digress) but that isn't to be found in our SVX OBD I or II system really. So where do we get that oxygen you ask? Well the good news is that we have some oxides of nitrogen coming past, and we really, really hate oxides of nitrogen because they are so toxic. Gosh we hate NOx! So what we are going to do is reduce (separate) NOx into N2 and 02 (nitrogen is also homonucleardiatomic). Gosh, I am glad we got rid of those NOx emissions. But wait! We also don't like carbon monoxide. So when we see some CO coming on past us, we oxidize it by taking our oxygen we have been holding onto and give it to that CO molecule so that it leaves CO2. Okay, good. But we aren't done yet. We also don't like hydrocarbons, because those are bad. So we are going to make sure that our waffle is a specific temperature which will cause them to burn on it. So we have some H2 (again, H is a homo) and some C now.
We hopefully have been given enough component parts in H2, C, O2 in those gasses to make H20, CO2, N2, so that we don't have any nasty gasses coming out the tailpipe. So, what goes wrong? As the catalyst ages, it loses its ability to store oxygen, and therefore cannot make the "good" gasses. So you fail inspection, and me, as your technician, says to you "sorry, your oxygen storage capacity is degraded to below acceptable levels and thus your car is spitting out NOX, and or CO and or HC in too high of levels. You are impressed by my fancy words and are feeling fortunate that your vehicle is in such good hands. Yes I have been drinking.

Real catalysts made in the past 25-30 years use the "waffle". This is often called a Three Way Catalyst (TWC) referring to three REDOX reactions. They are typically made of a ceramic honeycomb and are about as restrictive as a window screen. repeat: they are not restrictive in the slightest. Removing a properly functioning TWC will give no performance gains, despite what bill down the street said about how much better his Silverado ran after he had them cut out It is possible for the honeycomb to break loose and jam up sideways in the pipe causing varying degrees of exhaust restriction. I would say that the only way this happens is from some sort of shock to the unit that breaks it. As I mentioned earlier in the thread, I have seen this happen on a TWC a few times over the years, most notably in a 1998 toyota sienna that caused almost your exact symptoms. Due to the arrangement of the catalyst in that vehicle it was impossible to see the element without cutting the pipe.

Early catalysts, such as those used on the DeLorean, used pellets. The catalyzing agent in those was on the outside of these pellets. Conceptually, think of a ball pit from Mcdonalds in your exhaust pipe. Obviously, these were outrageously restrictive and choked engines. Many vehicles easily suffered a 20% power decrease by using them. Additionally, these failed regularly. Aside from the actual catalytic process working poorly and not for very long, they frequently became clogged, the balls fell out, and/or generated excessive heat. A shocking number of vehicle fires were due to these units. These units are cheaper to manufacture and many aftermarket suppliers will use this inferior design. I would not be surprised if your catalysts are of this design.

OBD I cars just had the catalysts. The cars left the factory and everyone hoped all was well. If it wasn't you would fail your tailpipe emissions test. Unfortunately, as you might imagine, the intricate understanding of the catalyst system and the different problems in the combustion process that could cause too much of these bad gases for the catalyst to handle proved to be beyond 99.99 of the technicians in the field who were largely uneducated stereotypical grease monkeys. Understanding the results of a five gas analyzer is still quite uncommon outside of places that regularly use it, and even those that know the causes are often ignorant of the actual root cause beyond what part usually causes an abundance of a certain gas. This ignorance led to the development of the ASE L1 test (advanced engine performance) which is much more difficult than the basic 8. I believe a passing score is 70%, which has proved borderline impossible for many in the field. An ASE advanced level specialist who has passed this test is usually very well paid.

One of the most obvious part of the OBD II requirements were the addition of an oxygen sensor after the catalyst to measure performance and illuminate the MIL if excess oxygen was detected. A properly working 1996/7 SVX will show rich/lean rich/lean cycles on the primary sensors (remember we are doing this on purpose so we have some oxygen to put in the bank). Before the catalyst is at operating temperature, the post cat sensor will mirror the front ones. Once the cat is heated up, the rear sensor will flatline indicating that the oxygen that the front sensors saw is being consumed in the oxidation reaction of the catalyst. Often, what you will see, is a rear sensor will oscillate rich/lean like the front ones but slower. This is a weak catalyst. Eventually, the front and rear sensors will do the same thing indicating a completely useless cat. Modern systems use very fast air fuel ratio sensors up front to rapidly change short term fuel trim.

Modern systems use an auxiliary air pump, which has made a glorious return in electric form as opposed to the old belt driven units like grandpa's 74 AMC matador had, which he removed the belt from when it seized in 1975. This allows the bank to get full faster and the catalyst to heat faster, making it work sooner. You will also notice that the catalysts are as close as possible to the engine to aid in warmup

So in summary, be very cautious of aftermarket catalysts, and don't cut them out thinking you just gained 20 HP, and sorry you just wasted 10 minutes listening to me babble.

Gosh I am glad we got rid of those oxides of nitrogen. They are nasty.

[dannmarr]

Originally Posted by Blacky

Please let it be over. Please make it stop!

LOL! 2x

[WhatIsThisIDont]

Originally Posted by NiftySVX

I would not go any further with diagnosis work until the exhaust is replaced

It's definitely worked it's way towards the top of the list, but the engine should at least run with a flowing "exhaust". If I'm wrong, somebody tell me.

My plan at this point is to get her back on her feet, drive home for spring break, and put the wagon together for a few months since I'm honestly tired of fiddling in the cold with this.


But back to the driving home problem.

I'll go out probably Saturday and put more time into it.

Planned Diagnostic process in order:

1: Disconnect O2 sensors and see if problem goes away. If there is no change, I reconnect and move on. I want to try this first because of any O2 sensor damage which may have been caused by the back pressure.

2: Disconnect/ Test temperature sensor leading to ECU. I'm thinking of this one because one of the biggest variables between it running with issues and not running at all is the temperature. When it was 15 degrees and below, it ran well until it warmed up. Let it sit for a bit and fire it up, and it still ran rough for a few seconds. In that time, the coolant in the rad would be much cooler than whatever was near the sensor. Also considering the hardwired "always on" fans would keep the coolant cooler for longer.

Is there any particular way to test this sensor and/or pigtail?

3: Test/inspect cam and crank sensor. I've been reading through countless old threads from the last decade and found cam sensors to be a bit of a popular failure, with symptoms being similar to a few of my own. I plan on taking out the cam sensor by the battery and checking it for any bits of debris or damage on the end, and then doing the same for the crank and other cam sensor once I find them.

My big question for this one is how do I test these sensors? I saw some people say look for voltage fluctuation when a magnet is passed by, and others said to check resistance, but with wildly different resistances stated by different people.

4: Take off left and right side timing covers and check the timing. I've never done this on the EG33, but shouldn't there be timing marks on both cam sprockets? And shouldn't both marks line up with the marks behind them at the same time? I haven't even begun research on this yet, but given the issues I've come across so far, it wouldn't surprise me.


So that's my battle plan. Any information, tips, corrections, or thoughts you guys have would be great. It'll be a balmy 32 degrees F on Saturday, so I want to go in and bust through this like clockwork.

Thanks again guys! We're (hopefully) bringing this painful thread to a close soon.

[NiftySVX]

Yes it should. Your EGR could be messing up too, that is a possibility.

[WhatIsThisIDont]

Originally Posted by NiftySVX

Yes it should. Your EGR could be messing up too, that is a possibility.

I'm not going to lie, I have only a very vague idea of how most of the emission system works. How would an EGR failure cause these symptoms?

Considering the EGR system depends upon vacuum lines IIRC, then it wouldn't surprise me if that is an issue if not the issue at hand given the state of the hoses.

[Tapani]

You do have the manuals, yes?

If not, you can download them from here:

http://phantomotaku.com/SVX/

Apologies for repeating myself, but why not get access to the data at the diagnostics port? It is by far the easiest path forward.

Also, are you absolutely sure the second MAF you tried is "known to be good" - you have used it yourself or tried it on another vehicle? You may want to check MAF harness too and engine ground points, if you haven't so far. These may cause weird issues.

The engine should run rather well even the O2 sensors disconnected, you will just be "open loop" all the time and may run a bit rich and get a CEL.

I would be surprised it the EGR would cause the engine to bog down totally.

Have you checked any of your plugs for color?

Good luck,

Tapani

[NiftySVX]

Tapani is always stealing my thunder!

I am hesitant to discuss the EGR system because I don't feel like it would help to keep throwing out possible failures. A trained eye could monitor the scanner data and see if the EGR is working based on the changes in the temp sensor. I agree that the EGR would be unlikely to bog the engine down totally but it will make it die at idle if stuck open.

This can be checked by a trained eye watching the EGR temperature.

Like I keep saying, it is fun to discuss all the possibilities but I am unsure of the usefulness of it. It is probably time to take the car somewhere to have it evaluated. I would take it to a subaru dealer. They may charge you 100, but I bet they could give you a definite answer

[Chucksta]

Damn.. Do we all think the same thing at the same time? Just got done looking that up in my 92 FSM. The pages you want are in the "Fuel injector " section. (2-7) My FSM shows them as page (s) 55 to 60 and 77

The OHM readings are to trace through the wiring for continuity and excessive resistance..

The volt reading ( when cranking ) should be with the multi meter set to AC volts, as far as I can tell, so an accurate reading would be more likely. Set the range to a low setting , as the signal should only come back at around 0.1 volts.

The cam and crank sensors appear to be interchangeable.

Cam and crank sensor #2 share a common ground to the ECU.

Here's a thought I had, and would someone who actually knows, please either confirm or scuttle the following.....


There's a way to get codes out of the ECU that has to do with popping off the driver's side kick panel, the one to the left of the dash. Inside there's a black connector and a couple of blue wires. Hooking up one wire (to the proper place ) and turning the ignition on is supposed to make it spit out any stored codes by flashing the power light. My thought was that a CEL may go off after a few re-starts, if the cause doesn't come back, or doesn't come back to a point that the CEL triggers again. But wouldn't the code still be stored? I know you can pull both current AND previous fault codes from the transmission... Does the ECU operate in a similar manner?

Re: Tapani's post about the EGR.

I recently had a CEL for the EGR.. cleared it and it returned, so off it came for a thorough cleaning out ( gummed up and clogged). The code never returned. If it wasn't for the CEL, I wouldn't have known that the EGR had an issue, there was no difference in the way it ran, before or after cleaning.

The same section of the manual as the crank and cam sensor (2-7) pg 65, gives the specs for the temp sensor.. If you want to "lie" to the ECU, a 400 ohm resistor, across the harness connector, would tell the ECU that the coolant temp is 176 F . I think.. That one is definitely " at your own risk.". If you want to try that one, you should probably ask someone who knows more about it than I do... which would be pretty much anybody.. Cause I HAVE been known to blow things up.. We won't even talk about what happens when you put mothballs in the gas tank..

[WhatIsThisIDont]

Originally Posted by Tapani

You do have the manuals, yes?

If not, you can download them from here:

http://phantomotaku.com/SVX/

Apologies for repeating myself, but why not get access to the data at the diagnostics port? It is by far the easiest path forward.

Also, are you absolutely sure the second MAF you tried is "known to be good" - you have used it yourself or tried it on another vehicle? You may want to check MAF harness too and engine ground points, if you haven't so far. These may cause weird issues.

The engine should run rather well even the O2 sensors disconnected, you will just be "open loop" all the time and may run a bit rich and get a CEL.

I would be surprised it the EGR would cause the engine to bog down totally.

Have you checked any of your plugs for color?

Good luck,

Tapani

I would love to access the diagnostics port. Do you mean with something like Evoscan or the likes? I don't know much about these systems, but if they work on the early 90's pre-OBD2 Subies, then I'll have to pick some up just because it looks like my odd fascination with this vintage isn't going away anytime soon. Is there a "go to" bundle of cable and software to get?

That MAF has been used in at least two vehicles in the last year and was working well, so I'm still assuming it is good for now.

Originally Posted by NiftySVX

Like I keep saying, it is fun to discuss all the possibilities but I am unsure of the usefulness of it. It is probably time to take the car somewhere to have it evaluated. I would take it to a subaru dealer. They may charge you 100, but I bet they could give you a definite answer

That may be what I end up doing, but I have held back for 2 reasons. First, I called the local parts dept. about making a key a while ago and was asked if I was sure the SVX was a Subaru, so I was a little hesitant. And secondly (and mostly), my stupid and ridiculous bullish pride gets in the way

Originally Posted by Chucksta

Damn.. Do we all think the same thing at the same time? Just got done looking that up in my 92 FSM. The pages you want are in the "Fuel injector " section. (2-7) My FSM shows them as page (s) 55 to 60 and 77

The OHM readings are to trace through the wiring for continuity and excessive resistance..

The volt reading ( when cranking ) should be with the multi meter set to AC volts, as far as I can tell, so an accurate reading would be more likely. Set the range to a low setting , as the signal should only come back at around 0.1 volts.

The cam and crank sensors appear to be interchangeable.

Cam and crank sensor #2 share a common ground to the ECU.

Here's a thought I had, and would someone who actually knows, please either confirm or scuttle the following.....


There's a way to get codes out of the ECU that has to do with popping off the driver's side kick panel, the one to the left of the dash. Inside there's a black connector and a couple of blue wires. Hooking up one wire (to the proper place ) and turning the ignition on is supposed to make it spit out any stored codes by flashing the power light. My thought was that a CEL may go off after a few re-starts, if the cause doesn't come back, or doesn't come back to a point that the CEL triggers again. But wouldn't the code still be stored? I know you can pull both current AND previous fault codes from the transmission... Does the ECU operate in a similar manner?

Re: Tapani's post about the EGR.

I recently had a CEL for the EGR.. cleared it and it returned, so off it came for a thorough cleaning out ( gummed up and clogged). The code never returned. If it wasn't for the CEL, I wouldn't have known that the EGR had an issue, there was no difference in the way it ran, before or after cleaning.

The same section of the manual as the crank and cam sensor (2-7) pg 65, gives the specs for the temp sensor.. If you want to "lie" to the ECU, a 400 ohm resistor, across the harness connector, would tell the ECU that the coolant temp is 176 F . I think.. That one is definitely " at your own risk.". If you want to try that one, you should probably ask someone who knows more about it than I do... which would be pretty much anybody .. Cause I HAVE been known to blow things up.. We won't even talk about what happens when you put mothballs in the gas tank..

Excellent. You got me the section to look in and Tapani got me the FSM. You guys are excellent!

And the only time I ever heard of someone putting mothballs in a gas tank was with a 1944ish Ford, and the end result was hilarious speed followed by some angry engine components

[Tapani]

Download the Evoscan from here:

http://www.evoscan.com/

It does work on the SVX too, OBDI included.

The buy the interface from Mike: http://www.ecutune.com/ECUtuner.htm

And the cable for example here: http://www.ftdichip.com/Products/Cab...BTTLSerial.htm

Naturally, you need a laptop (with admin rights) to install it all.

You are ready to go. Instant diagnostics, data logging and trending any and all engine + tranny parameters available.

Works real good, trust me.

/Tapani

[WhatIsThisIDont]

I did some work today. I took the driver side cam sensor out and found a few interesting tidbits. First, there were some metal shavings on the magnet end. More importantly, though, I found the connector to be full of oil (the oil leaks are quite terrible). I disconnected it, cleaned both ends, put it together, and it fired up like a champ.

I let it idle for about half an hour just to make sure everything was good, which was a good idea, because it started surging while idling. When I gave it a little gas, it fell on it's face and died. From then on, it wouldn't restart for another half hour, at which point it would only run for 30 seconds and went back to being a pain.

Judging by how it goes back to normal after some time and seems to do better in colder weather, I'm looking at the temperature sensor. I tried the multimeter today and it gave me some odd, but not outlandish readings. I unplugged it, and it still wouldn't restart.

If I can find the part, I'll go out and swap it in tomorrow until I can get it home and get an actual Subaru one in it.


And I'll be getting that cable and software setup, it just won't be here before I leave for spring break. Thanks for the info!

[SVXRide]

Have you disconnected the O2 sensors yet to see if the problem goes away. Your SVX seems to have a lot of secondary issues that need to be addressed (what SVX doesn't? ), but I'd suggest you get the "dies on me" issue resolved first.

FYI - finding metal shavings anywhere in an engine is not a good thing. You might want to pull off the oil filter and drain its oil through a coffee filter to see if you've got any metal bits in the oil.

Cheers,
Bill

[WhatIsThisIDont]

Originally Posted by SVXRide

Have you disconnected the O2 sensors yet to see if the problem goes away. Your SVX seems to have a lot of secondary issues that need to be addressed (what SVX doesn't? ), but I'd suggest you get the "dies on me" issue resolved first.

FYI - finding metal shavings anywhere in an engine is not a good thing. You might want to pull off the oil filter and drain its oil through a coffee filter to see if you've got any metal bits in the oil.

Cheers,
Bill

There are more secondary issues than I'd like to admit lol. It's a work in progress that I'll be spending my summer on. I did the oil change over Christmas and didn't see any shavings then, but I'll screen it next time and put a magnetic drain plug in for future use.

I haven't disconnected those sensors yet because it started and ran well for a while, so I figured all was fine and dandy there, but also because, honestly, I can't find the fargin' clips where they connect from under the hood. AAAAaaaaannnnddd that's why I didn't become a mechanic

From what I've seen with the temperature sensor, it's just a variable resistor based upon temp. So IF I can't find another one before break, I'l run it until it dies again, and then put a resistor across the connector with the resistance the FSM says the sensor should have when at operating temp. If it easily restarts, then we will know that the problem is there.

[Chucksta]

Re: resistor to give ECU user selectable temperature reading.

Of course, once again, as I know Jack about electricity.. I would offer the following "McGuyver",,,

A volume knob is actually a potentiometer, that is, it offers it's highest resistance at the lowest volume setting, and it's lowest resistance (0 OHM) at the highest volume setting... Unless it's from Nigel Tufnel's amp... his went to eleven.

A trip to the local second hand shop would probably get you a used desktop computer speaker to make the ultimate sacrifice for less than ten bucks.. Open the back and nip the wires off the volume knob ( leaving as much as you can to play with ). Hook up your multi meter. find where the volume knob gives you 400 ohms and mark it, change the settings on the multi meter and see if you can get it to 4000 ohms ( almost zero volume ) and mark it. It'd be a crap shoot whether it'd go to 4000, but 400 should definitely be in range. If you have a decent electronics supply, or electronics surplus place nearby, a rated one wouldn't be much more than five bucks. Radio shack (in Canada) lists a 5K one for $3.49. 4,000 ohms is dead cold, 400 ohms is normal operating temp.. So, with a twist of the wrist, you can send the ECU on a trip from Alaska to Death Valley.

Of course, at your own risk.

Best of luck.

[THE NEW GUY]

have you check the roll over valve?