87 GL Wagon MPFI overheated while towing ... advice?

[jkhackney]

Hi,

 

My Subaru overheated while towing and while I read some other threads about overheating, there isn't much written here about towing, so I thought I'd start a new thread.

 

I live in Switzerland with an EA82 MPFI 3-speed Automatic GL Wagon (Leone II) with a hitch. No A/C, no turbo. About 130,000km on the original engine.

 

As scheduled maintenance in February I had replaced the HG's, intake manifold gaskets, sealed the cam towers, valve seals, Mickey Mouse seal, etc.  I also replaced the plugs, plug wires, dizzy rotor and dizzy cap at that time, and topped off the driveline lubricants. It seemed to be running perfectly, quietly, and consuming little oil and less fuel than before.

 

On paper it has more torque and more Horsepower than my old Land Rover, so I tried towing our 800kg/1800lb camping trailer last week on vacation over some big mountains. The car also had 4 bikes on the roof and 2 adults/2 kids inside.

 

Wife at the wheel, doing 4500-5000 rpm up a 10-12% grade in 1st gear and 35C/90F outside temperature, it boiled over.

 

After it cooled, I recovered the green cap from the overflow canister (it was down in the bodywork), put in about 2L of coolant (!) and we continued, only to overheat and blow the green cap off again and need another 2L of coolant. The electric fan was working. There's nothing blocking air through the grille.

 

Returned home to swap cars and now I'm back from vacation figuring this out.

 

1) the car has been normal when not towing anything.

 

2) It has always pinged like mad under load (steep uphill, 2 adults/2 kids). I had the timing checked by a Subaru garage and they said it's fine.

 

3) Before leaving with the trailer, I filled the coolant overflow tank. It was empty. I had not been aware of coolant loss as it's my wife's car and I'm not on top of the day-to-day situation. I didn't know at the time that an empty overflow tank meant I should also fill the radiator. Thus it COULD have been low on coolant from the outset!

 

5) After it overheated, we noticed a weep of coolant from the thermostat housing gasket. I had never replaced that gasket.

 

So what am I looking at here?

 

Was something wrong with the Subaru to have caused overheating, or would you have expected the overheating under this much load?

 

It overheated even when I re-filled the radiator and overflow tank. So it being low on coolant wasn't the cause of overheating.

 

Does the repeated overheating indicate that I have perhaps warped the heads? Might they have cracks?

 

Could a malfunctioning thermostat cause this?

 

Thanks,

Jeremy

[MR_Loyale]

Did you check the vehicle weight limit on the door jam?  Sounds like you exceeded the weight limit.

[jkhackney]

The hitch is a Subaru part and its limit is 1500kg trailer/70kg tongue weight.

 

The vehicle's "circulation document" (title) says maximum towing load 1130kg, same as the vehicle weight. Maximum vehicle load is 550kg.

 

My trailer and the load I had onboard were well below these limits. I would think that the black-on-white has as much to do with lawyers as with engineering. The allowances could be over- or underestimated. I thought maybe I might find some de facto information in the forum because of two things:

 

My friend and I towed this trailer with the Subaru 600 miles from Belgium where I bought it. That was winter (cold) and flat except for a final 10% gradient for about 6 miles shortly before home. It pulled up that hill like a train, much stronger than my Land Rover. So I was surprised it konked out this week.

 

Since Subarus are used by farmers here to haul loads, I expected it to handle the hills better.

 

Thus I wonder if I have a major mechanical defect or was kidding myself about it going so easily in winter.

 

BTW for those frightened by the prospect, this trailer has got an electric brake.

 

-Jeremy

Edited August 6, 2013 by jkhackney

[jkhackney]

OK I've decided my radiator needs replacing. The fins are brittle and green and I can push my finger through them. I ordered one, and a thermostat with a new gasket.

 

I let the engine idle yesterday until warm and it seems fine. Valves are quiet, oil color is fine, coolant color is fine. I left the radiator cap off while it warmed up and coolant spilled out. I assume that's a bit of air expanding (no bubbles came out though), but maybe it's also a good sign that the impeller on the water pump works ? I should bleed the air out, but if I'm just going to replace the radiator on monday it's not crucial at the moment ...

 

I doubt we'll get back up to the 80's F this year again but I'll give it a good test after I replace the radiator and report back.

 

And pinging? I'll check the timing, myself this time. Before, I wasn't sure how to hook up my RPM meter but now I have a clue where to find a suitable 12V source.

 

-Jeremy

[scoobiedubie]

Since your overflow bottle was empty and you had head gaskets doen recently, it looks like you need to retighten the cylinder head bolts as they will need tightening several thousand miles after a HG job.  You probably also have a gray or black deposit on the bottom side of your radiator cap and your coolant bottle is dirty on the inside. 

[NorthWet]

I am somewhat at a loss as to your comment regarding an rpm meter.  Such a meter should not be needed, and the battery should provide access to power and ground connections.

 

Are you aware of the need to connect the green diagnostic connectors in order to set the timing.  If not, or uncertain, please ask.

 

Regarding the overheating, I was going to ask if you had checked the radiator for hot/cold zones on the core, indicating tube plugging (very common), but since you are replacing it this has now (mostly) irrelevant.  (It would indicate if this was the major issue or something else is the major fault.)

 

Since overheating has occurred, you might want to check the condition of your automatic transmission fluid:  Subaru automatics have issues with overheating their ATF, and the 3-speed has issues when it doesn't have clean ATF.  The automatic's torque converter might have contributed to the overheating issue.

 

Regarding the pinging:  Checking the ignition timing is a good starting point.  Low oil level (causing high oil temperatures) and high coolant temperatures will contribute to pinging.  ANY oil in the intake charge will drastically increase the susceptibility to pinging:  Check to make sure that your PCV system is functioning correctly and there isn't too much crud buildup at the PCV hoses at the valve covers.

[scoobiedubie]

The extra drag of 4 bikes on the roof, is not figured into the design load recommendations.  Don't be surprised if you cracked your cylinder heads, instead of just suddenly loosened the bolts up.  Henceforth, don't be surprised that it keeps drinking coolant and you get frequent boilovers. 

[NorthWet]

Non-turbo EA82s do not have issues with head cracks. except the "normal" ones between the valve seats.  Headbolts don't loosen from overheating, though the results of excessive differential thermal expansion (aluminium vs steel) might make it seem that way.

 

I assume that your engine does not have an auxiliary cooling fan, either engine driven or electrical (common on A/C equipped models)?  You might want to consider getting one. 

[MR_Loyale]

Non-turbo EA82s do not have issues with head cracks. except the "normal" ones between the valve seats.  Headbolts don't loosen from overheating, though the results of excessive differential thermal expansion (aluminium vs steel) might make it seem that way.

 

I assume that your engine does not have an auxiliary cooling fan, either engine driven or electrical (common on A/C equipped models)?  You might want to consider getting one. 

 

Wait a minute. Excessive differential thermal expansion? Are you saying the heads and block are different metals? I thought they were both aluminum.

[Quidam]

Jeremy wrote: "2) It has always pinged like mad under load (steep uphill, 2 adults/2 kids). I had the timing checked by a Subaru garage and they said it's fine."

 

"fine" probably means set where it's supposed to be. If that's the case it needs a higher octane fuel. You could just loosen the distributor, then rotate it clockwise until the ping stopped. I don't recommend it though as that would be retarding the timing. Gasoline is a whole different formulation now compared to when these cars were built.

 

Pinging under a load like that produces a lot of heat, more than normal. Radiator and stat sounds like a good start.

 

HTH

 

Doug

[NorthWet]

Wait a minute. Excessive differential thermal expansion? Are you saying the heads and block are different metals? I thought they were both aluminum.

Sorry for being OT, but:  the differential expansion exists between the aluminum parts and the steel fasteners trying to clamp them together.  If anything exceeds its limit of elasticity, it will be permanently deformed.  Also, different alloys of the same "metal" can have very different rates of thermal expansion.

[jkhackney]

Thanks for your replies.

 

Yes it would be a good idea to first know what the cause was, before willy nilly replacing parts of the coolant system.

 

I'll try to address some of the mentioned issues.
 

1) bikes on the roof and aerodynamics ... certainly a drag but, well, maybe not at 20mph, our speed when we overheated!!

 

2) To clarify its pinging. Our local Subaru garage took it for a day, checked the timing, and took it for a test drive, reporting that the ignition timing was set up correctly and that they didn't hear pinging. $70. That frustrated me.    It was REALLY pinging a lot more than usual the day it boiled over. We alway use the fuel that is labelled "95" here and the garage says that's fine. "98" is available. my wife reported no difference when she used it, once. It seemed to ping less after I did the heads and decarbonized, but I really hardly drive the car so I can't say for sure. Our altitude is 580 meters by the way, roughly 2000ft.

 

3) Wasn't aware of green diagnostic plugs to set iginition timing. I was going to use a timing light?

 

4) Why I wanted to check the RPM at idle to set the ignition timing ...  in case the ignition timing had maybe been set relative to the wrong RPM (maybe in "D" instead of "P" or "N").

 

5) The old radiator fins are certainly degrading, which justifies replacement. But it's still in the car, so I can check it for cold/warm areas to troubleshoot it for clogs and to check the thermostat at least. But since the thermostat housing joint is leaking anyway, I'll take out the thermostat to test it in a pot of hot water before just swapping it out. Both these tests might rest my mind that our issue is just the cooling system.

 

6) Radiator cap is clean, will check the inside of the overflow bottle for deposits, thanks.

 

7) Retorquing the heads, hadn't read about that being necessary. What a pain! I have to jack the engine up a few inches to reach the bolts. The heads have the superficial cracks between the valves, but I tested them for leaks (inverted them and put in diesel overnight) and they were OK as far as my facilities allowed me to test them.

 

8) I have no auxilliary cooling fan and I was thinking afterward that this would be a good thing to add.

 

9) The PCV system is clean of slag and clogs but the pipes are really brittle. I replaced one with a garden hose because not even my local dealer could get an original hose.

 

10) I'll check the ATF as I change the radiator (some will spill out, then). I assume I'm looking for scorched color, burnt smell?

 

11) On the other hand I was thinking, we tried to tow an 1800lb trailer with a 25 year old Subaru over a pass that averages 10% for about 11 miles. Even having grown up in Colorado, I have to admit that's a huge pass, and a bit of a high expectation.

 

I expect the radiator to be delivered tomorrow, Monday,

 

Jeremy

[NorthWet]

RE:

 

1) Agreed.

 

2) I run our "87" octane-rating here in all may cars, spfi, mpfi, turbo, and usually in my SVX (10.5-ish CR).  Pinging is only been very minor, and I have a short-but steep grade in my commute.  Engine/intake-charge temperature and presence of oil much greater sources of pinging in my experience.

 

3)  You still need to use a timing light, but, since the timing is totally ECU controlled based off the static setting, you need to connect the green connectors to tell the ECU to not alter timing while you are adjusting it.  Since you say yours is an MPFI, the connectors should be inside just below the steering column and above your shins.  I have only dealt with left-hand drive vehicles, so if yours happens to be right-hand drive things MIGHT be different.   If your car were an SPFI model, the connectors would be in the engine compartment next to the wiper motor.  XT/Alcyone is in the trunk/boot.

 

Anyway, these connectors need to be connected, the timing light used to check/set the timing, and the connectors disconnected.  Many people forget to disconnect these connectors afterwards.

 

4) RPM is irrelevant.  The distributor has no advance mechanism, just a shaft rotating an optical interrupter disc.  The ECU makes all relevant adjustments to this base timing.

 

5)  I would be tempted to just put a new Subaru/equivalent-designed T-stat back in, so this is eliminated as a possibility.  I am not impressed with my own success rate checking t-stats in boiling water.

 

 

7) Retorqueing the heads is recommended by Subaru for their OEM HGs.  FelPRo's "PermaTorque" says that a retorque is not necessary.  Other brands may have varied instructions.

 

 

9) Some knowledgeable members state that the PCV hose sizes are important, but I think that relates mostly to the "tree"-fitting where the hoses come together.  Did you check the PCV valve itself?  (My money would be on pinging from overheating, but any oil really does increase the pinging/detonation.)

 

10) Yeah, the ATF will turn ugly brownish and smell bad.  How bad it is sometimes does not show up until you drain/flush it.  (I personally like to disconnect the cooler hoses from the radiator and place them in a catch container, then start the engine and let it pump out the fluid; stop, refill, start and pump out.  Others might say that this could be damaging.)  I have had ATF that looks like mud flush out of both 3ATs and 4EATs.

 

11) Not the kindest thing to do, but sometimes you just need to get things done.

 

You could also use some auxiliary cooling methods.  Since this is an exceptional situation, you might do things like use a water spray on either the radiator core or on the engine block itself.  The water can carry off a tremendous amount of heat (around 10 times more than air, and that is without considering phase-change: water to steam).  Something like a spare windshield washer setup might work nicely. 

 

Others here have recommended using methanol-water "injection" to control pinging.  Some fairly simple methods have been written about (like using a windshield washer setup...)

[Quidam]

Is this car an 1988 with the optical distributor?

 

I swapped an '88 ECU into a Loyale SPFI 3AT 4WD. I put a vac gage on it and advanced the distributor to the highest vac reading. Ran it for a while with a mix of between 87 and 93 octane in it until it ran clean with the distributor set like that.

 

I could turn it to 7500 rpm in first gear where it would just shut down, 6800 RPM in second was all it would do, no ping at all. The '87 EA 82 had 170,000 miles, IIRC.

 

The ECU did nothing to alter the advanced timing.

 

Go on, try it if you don't believe it.

 

Got to go.

 

Later

[jkhackney]

Thanks,

 

I was able to observe the rad after a long drive. It gets hot all round, well maybe two spots in the bottom just left and right of center are a bit less hot. Both rad hoses are hot. Re-starting a warm engine, the radiator gets hot immediately, telling me the fluid circulation is good.

 

The parts supplier failed to deliver and I've had to re-order with a Subaru garage.

 

This garage owner (same one who tested my ignition timing last year) says the 1800lbs should have been easy to tow, and that there's "definitely" cracks in my heads. He bases this on many years' experience. It's actually the same thing my local non-Subaru garage said, too (heads are leaky).

 

Well, there's no question the rad is falling apart so I'll change it and see. Also I'll ask my wife to systematically test the car with 98 octane fuel.

 

Yes I tested the PCV valve when I did the heads and it's good. It does burn oil. I don't know if it's coming in the injectors or the valves.

 

My head gaskets were from HB Jakoparts with no mention of retorquing in the instructions.

 

It's a Swiss Edition Station, first reg 10/1987. Yes, left hand drive. It has a slightly raised roof (maybe you've seen this) and slightly different tail lights, and higher wattage headlights than the U.S. version. I have the U.S. factory manuals for 1988, and a Haynes. But I can't say for sure if mine has the exact features of the U.S. models marketed at the same time. It is MPFI with no turbo, but to get the correct intake manifold gaskets I had to order "turbo" ones.

 

I didn't read mention of those green plugs, so thanks for that tip. The manual does say to set the timing with the advance disabled. Yes it's an optical distributor for sure.

 

OK I won't be ablel to start on this until next Tuesday evening at the earliest.

-Jeremy

[NorthWet]

Did you have the heads re-surfaced at the HG replacement?  If not, this may be part of the issue.

 

Your mechanics experience is arrayed against a broader experience from long-time USMB members.  Exhaust divider cracks are the only common head crack, and those only in MPFI heads and those (to my questionable-knowledge) exclusively in turbo heads that have been seriously overheated.  These cracks dump coolant into the exhaust more than directly cause overheating.

 

The slightly-raised roof is probably what is referred to as a "Touring Wagon" in the US.

[Gloyale]

Pull the Y-pipe and inspect the inside of the exhaust ports for cracks.

 

Look at were the wall or web extends across the middle of the port.  Where that part attaches to the walls they develope cracks ussually to the outside end of this "bridge" that divides the 2 cyls.

 

If there are no cracks, then just replace the radiator and maybe add a trans cooler and call it good.

[WoodsWagon]

If you're replacing the radiator, I would make sure the new one is a double core one from a turbo model. You can also add the belt driven clutch fan from an a/c equipped car for a big increase in cooling capacity. It bolts to the waterpump. You may need to use longer studs on the waterpump for the fan to go on. That fan and its shroud will move a ton of air through the radiator, much more than the electric fan will.

 

With the 3spd auto, you have no lockup clutch in the torque converer so it is always slipping and making some heat. Towing a trailer and climing a grade is working the transmission hard as well as the engine, and it is cooled through the same radiator. You can add an auxilary transmission cooler in line with the heat exchanger in the radiator and that will take some of the transmission cooling load off of the engine's cooling system.

 

I wouldn't go too far down the headgasket/cracked head/failed anything path on the engine. The overheating could easily be just the cooling capacity of the single electic fan on an old radiator beeing exceeded by the load you were putting on the car.

 

One other thing is to pressure test the radiator cap. If it isn't holding pressure, the boiling point of the coolant is much lower and it will boil over through the reservoir.

 

Changing the transmission fluid would be a good idea. It will look darker and smell like burnt celery (the vegetable) if its been overheated.

[jkhackney]

Hello again,

sorry I haven't updated the thread in over a month.

 

I got the only replacement radiator I could find in Europe. Identical to the original except silver in color.

 

I put in the radiator and thermostat (Subaru thermostat, 72 deg) and checked the timing with the green wires plugged in.

 

Thanks for the tips on checking for cracks in the head (looks easy to visually check) and about the improved cooling alternatives.

 

No I did not have the heads planed when I did the HG's. Yes they are non-turbo but MPFI (might have mentioned it).

 

My preference would be to put in a manual 5-spd! But I'll look at what I can find as far as an aux electric or pulley fan and inline transmission cooler.

 

I should have just bought a radiator cap when getting the rest of the coolant parts.

 

The pinging seems to be lessened but it still does it. Must be some oil coming into the combustion chamber ?

 

The transmission fluid was OK.

 

I haven't tried towing uphill again. Next week is school vacation again so we may give it a try.

 

Someone snapped the rear wiper off last week so I'm going to the junkyard anyway. Maybe they have some of these cooling parts but these cars are very rare here nowadays so I'm not optimistic.

 

-Jeremy