legacycentral bbs

These are off my 92 legacy (2.2L N/A). They have roughly about 13000 miles on them. I had pulled them because I was wanting to see if they were oil soaked from the mysterious subaru oil disappearing trick.

BKR6E11 Vpower plugs. Replacement are just the standard NGKs

I'm mainly concerned about the driver's front. Im not sure if it's from running too thin of oil in the car (5w20, recently went back to thicker), and an iffy valve guide, or my friend at work suggested they looked like coolant deposits? I've never seen any spark plug reading guides refer to coolant deposits, then again I don't have any indepth books. Cylinder when I tested it months ago was at 180psi, like the other two were, and it was the driver's rear that was lower at 163psi. I've pressure tested the coolant system and found no leaks, and don't ever have any white smoke or coolant loss.


It does run rich at medium heavy to WOT throttle. The igniter helped, but it's still pulling timing (possible Cam sensor), and is evident by black smoke.










Thanks in advance

They're actually not THAT bad. Normal plugs are a little white ... as aposed to black ... those look like your getting a little oil blow by. Either from piston rings or valve seals. IDK how many miles you have on your car but from what I've seen on EJ22's after around 200k or more that's normal. Rings DO wear out ... valve seals DO wear out. But it doesn't look serious. Unless your having engine issues, I wouldn't worry about it.

I will say it does look like you beat on your car a bit ... just a bit lol

engine has 166K on it. It runs great other than the running rich at medium heavy to wot throttle. It would be nice to get a cable built to see if I'm getting retard or if timing just isn't there. The car has run a ton better since the t-belt replacement, and hasn't had a no start since I put the new igniter in this weekend, but still don't like the black smoke. ECT is about a year old genuine subie. 02 sensor is genuine subie less than year old with additional ground wire. Main cat didn't look plugged when I had the exhaust done (could see thru it). Plug wires year old. Plugs are new now. Air filter clean. good fuel pressure. I've run 87,89,93 and still runs rich. Unless the Knock Sensor (Jecs branded but from Autozone) is "seeing" knock and pulling timing, not sure.

If you reckon you can see smoke at wide throttle, presumably in the rear vision mirror, then it would need to be seriously rich. Those plugs dont look rich to me. But...
If you had a cable you could check the IPW and then know if the ECU was telling the injectors to richen. The easiest way to "build" a cable these days is to buy a FTDI cable straight off the shelf. They are only about twenty US dollars. Here is some of a bit I wrote for a local forum...

Dontronics... http://www.dontronics-shop.com/ftdi-usb ... cable.html $44.00 N.Z. dollars delivered to your door.

Strip the black, yellow and orange wires. Connect the black wire to the ground pin in the subaru's yellow diagnostic connector and the other two go the three and two pins. Don't worry which way, they can be swapped and tried both ways.

http://www.vwrx.com/index.php?pg=selectmonitor Free software. I can post more ecu addresses to suit this software. Or,

http://www.vwrx.com/index.php?pg=selectmonitor Cheap software that is leading the world and made right here in N.Z. I also have extra addresses for this software

log1call wrote:If you reckon you can see smoke at wide throttle, presumably in the rear vision mirror, then it would need to be seriously rich. Those plugs dont look rich to me. But...
If you had a cable you could check the IPW and then know if the ECU was telling the injectors to richen. The easiest way to "build" a cable these days is to buy a FTDI cable straight off the shelf. They are only about twenty US dollars. Here is some of a bit I wrote for a local forum...

Dontronics... http://www.dontronics-shop.com/ftdi-usb ... cable.html $44.00 N.Z. dollars delivered to your door.

Strip the black, yellow and orange wires. Connect the black wire to the ground pin in the subaru's yellow diagnostic connector and the other two go the three and two pins. Don't worry which way, they can be swapped and tried both ways.

http://www.vwrx.com/index.php?pg=selectmonitor Free software. I can post more ecu addresses to suit this software. Or,

http://www.vwrx.com/index.php?pg=selectmonitor Cheap software that is leading the world and made right here in N.Z. I also have extra addresses for this software

this is visible black/dark gray smoke from the tailpipe visible in mirror, and is visible after some time on the rear hatch. Car still has main cat (secondary was removed). Exhaust and Intake gaskets leak free. Unless it has to do with the 304 stainless exhaust it has and it holding in heat allowing normal pipe soot to come off, but I would rather think it would bake on the pipe..

Also would note that in heavy throttle afterwards there is a distinct smell that radiates into the car from the front of the engine. I found another thread mentioning knock sensor, and a "hot exhaust" smell after WOT runs.

LOL after going out there I wonder if the exposed wire on the harness side of the knock sensor has anything to do with it. It's so close to the connector I can't fix it. Going to have to try and find a spare.

Yeah, I'd fix the knock sensor wire.

Other thing to note, the owner's manual states not to use oil less than 10w30 for "sustained" high speed driving.

The tolerances are such that, especially an older engine, I wouldn't run anything lighter than a 10w30.

I had been running 5w20 for awhile, but usually after loosing a quart every 1000 miles, it got 10w30 mixed in. This last batch was 3 quarts of 15w50 to 2 of 5w20, and that was a tad too thick. Maybe I'll just go to a nice 10w40, or a 15w40 diesel oil.

I got the knock sensor wire fixed per your pics josh. I made sure wire on KS was at the 10:30 ish position. I left the ground wire I added off it for now. It seems to be better. Smoke is not as heavy, but it's still there. Still get the smell but it doesn't last as long. It did it's no start on me again today, but instead of missing when it fired back up, it ran normal. the Code that kept popping up before was the Cam Sensor.

Ran it in D-check mode for 20 minutes all was well.

Can the Cam Sensor if somehow putting out a signal but perhaps the incorrect one, cause timing retard, or not? I'm wanting to rip into the dash this weekend, and make sure the ECU isn't all corroded up or something, and try and find a spare to see if maybe the ECu is doing all this.

well your def burning oil. From the piston rings or the valve seals ... or BOTH ... that's a different story. You can replace both in one shop really and the parts arn't much. But it's labor intensive and requires a moderate amount of mechanical knowledge and aptitude

yeah, I'm not to worried about it burning the oil. It's just my daily driver until I can get the $$ around for a newer subie, as the body cancer hit this one pretty hard from previous owners lack of care. I'd just like to nail this rich running crap, so I can stop using so much gas.

Those cakey deposits on your plugs look to me like they are most probably zinc. Zinc is an additive that supposedly extends the useful life of oil by taking part in a chemical reaction at the point of metal-metal contact. It is added in higher quantities to thinner oils because they stand a higher chance of creating a dangerous metal-to metal contact situation. What's probably happening is your thin 20w oil slips past your rings enough that it combusts and sticky zinc is left deposited on the plugs. High zinc content in oil is a well known fouler of plugs, and this is why they recommend you change your plugs more often if you run any of that silly "high mileage" oil (which has much more zinc).

I know it's been discussed, but mixing up your own mutli-viscosity oil is really not recommended. The chemistry becomes complicated because different long-chain polymers are used to make different viscous points. The issue is that there is no mixing on the molecular level.. 50% 40w + 50% 20w does not = 100% 30w. It may mean a colloidal mix of 40's and 20's that's otherwise heterogenous.

Thought of another way, corundum sand has a hardness of 9 on the Moh's scale, and talc a hardness of 1. Mixing these 1:1 results in a 'solution' with an average hardness of 5, but that "solution" would still scratch the surface of a quartz crystal which has a hardness of 7. This is because grains of corundum sand are still in the solution; corundum and talc will not mix on the molecular level at room temperature.

What this all means is that for your mixture of 20w+50w oil to actually turn into homogeneous solution you'd need to heat the two past the point of thermal breakdown, which would destroy the integrity of the oil. Because of the mixture, you have none of the benefits of either oil.

One other thing... a larger range in viscosities of a multi-viscosity oil means shorter oil life - or at least a shorter amount of time at which an oil performs at it's cited viscosities. This is because to create the range in say 5w40, more long-chain polymers are needed, which tend to suffer from physical breakdown due to shear. Long chains made short drastically lowers the viscosity of the oil.

None of these things make immediate differences... but will over time. Just thought I'd chime in, sorry if it's not necessarily pertinent.

thanks morgan. I really have only mixed the oils because of the 15w50 M1 I got on clearance ofr $1.74 at Target.

I'm not wanting to call you out, but many people sucessfully mix oils and can be done as long as you know their base viscosities at certain temps. I'm not disagreeing with you on a molecular level or anything, but simple UOAs/VOAs prove it works.

http://www.rohmax.com/rohmax/en/custome ... lmixtures/

You are right about the high Zinc even with MC 5w20 being SM API

Aluminum....0
Chromium....0
Iron........0
Copper......0
Lead........0
Tin.........0
Moly........30
Nickel......0
Manganese...0
Silver......0
Titanium....0
Potassium...0
Boron.......157
Silicon.....3
Sodium......0
Calcium.....1711
Magnesium...7
Phosphorus..551
Zinc........698
Barium......0

Yes, I agree that oil can be successfully mixed, but it's difficult to do correctly. There was an interesting article in the car bible on this a couple years back, and this is a debate that's smoldered for years. I really did not intend to sound condescending, apologies if I did. Essentially, the only oils that can truly be mixed to react at certain predictable viscous points are synthetics. Dino can be mixed, but it works on averages.

The real advantage to this is finding the temperature at which your motor prefers to run and grading/blending an oil to have a desirable viscosity at that temperature.

Those plugs don't look to be too oily to me, but...
If a motor is burning oil, the oil can make it pink.

The real issue at hand is I need to get a damn cable. Need to find out what's pulling timing, whether it's seeing knock and retarding, or it's just not adding timing.. Had the wife verify today and it's nice dark smoke.

log1call, do you think it may be possible to use a FTDI based USB to serial adapter I have here already? Mine has the board built into the serial connector? I used it for OBD1 stuff on my GM stuff.

It will depend how old it is.

I have used several setups to connect to subarus. First I had a max232 adaptor that went to the serial port. Then I used a ftdi serial to usb adaptor to connect that old setup to my new laptop. Now I have a later model ftdi lead as in that link, it's the simplest yet.

If you have a serial port lead, or a serial to usb adaptor just try it. You can't do any harm. If it's a 232 adaptor you will need twelve volts and earth off the ssm plug, they can't be swapped or the 232 will blow. The other two, the RX and TX wires you can swap around if you aren't sure which way they go and they won't hurt anything.

Plug the adaptor you have into the laptop, feed it power and earth if it needs it then have a look in control panel/system/hardware and check what port number the adaptor/lead is using. Start up your free software and set it's port to the port the adaptor is using. Connect the TX and RX leads to 2 and 3 of the ssm plug. Start the free software(with the key on) and there should be data. You may have to "discover" ecu addresses if your car isn't covered by the softeware but you should get a ecu identifying number at very least which wil help find the addresses you need.

You could deduce if it's having timing pulled by the knock sensor by disconnecting said sensor and going for a gentle drive, slowly working up the load while listening for knock/pink. You know what pink/knock sounds like right? Depending on your ecu it will either pull the timing straight away as a safety measure or it will not get the signal and there will be knock under load. Ignore the check light while you are doing it. If the timing gets pulled straight away then it will behave as your car has ben behaving. if it ignores the disconnected sensor it will go better but knock. Don't let it knock too much or too hard.

it's not that old.. Let me find a link. it's USB to serial with the FTDI chipset in the serial part. I'm not sure what the pinouts for the 9 pins are. Maybe that will help. I tried using pins 2 (receive) 3 (transmit) and 5(ground) on the serial side, but nothing.

this be the one..

ebay link

or with this type of cable should i be doing what Josh has posted on this link in referal to the 9 pin..

http://www.surrealmirage.com/vrg3/b10scan/

But on those instructions there is nothing that connects to the pin 2 so how does it receive..

If I can find a break away USB to TTL board would that work as long as following the wiring guide, or does FTDI stuff only work?

This is the one you want...
linky
The top one, it's made to work at five volts. The next one down is for three volts and they don't work.

The adaptor you have already is a serial port adaptor. To use a serial port, whether an adaptor like yours or an actual serial port on the back of the laptop, you have to invert it's signals and change their voltage using a max232 integrated circuit. It's easier to just buy the new lead from ftdi.

Get the lead, strip three wires, black goes to earth, orange and yellow are for RX and TX, try them both ways till you find which is which.

Use this software to check it goes and it will get you a ecu identifying number... http://www.vwrx.com/index.php?pg=selectmonitor.

Once you have the number it will either connect straight up to the car or you might have to rename one of the other ecu identities to your number in the configuration file that comes with the software.

If you like that software, you will really like the evoscan software, it's heaps faster, does more cars, more makes, more parameters. It's the best software availiable at the moment.

I have added a few more models to the evoscan setup, if your ecu number isn't in the vwrx config file we can put some of the ones I have in or make you a new one.

Those plugs are fine.

Worn.

But fine.