Overheating problem, help please...

[redsvx94]

Thanks for the comment, SVX-FF.

>If the engine temp rises during idle and then drops when you increase the engine rpms I would say it is time for a new water pump.

That makes sense, but actually what's happening is if I step on, the engine temp rises, and if I ease off the temperature decreases. But overall T slowly keep rising under idle.

>To check to see if you are circulating coolant, remove the rad cap with the engine cool. Start the engine and when she warms up you can see the coolant flow past the cap opening. You can also make a big mess doing this too. be careful.

I actually did that, its more like the coolants are bursting out of the radiator neck, I don't know whether this is the sign of coolant flow, or sign of coolant bubbling out...

[redsvx94]

I was reading posts regarding overheating, and wondering a few things.
1. The fact that top part of radiator is MUCH hotter than lower part suggest, there isn't enough water flow.
2. When rpm goes up a little the engine temp gauge goes up and when I lower the rpm the temp gauge goes down.

1 simply suggests that there is not enough water flowing. But 2 seems to suggest that the water pump is spitting water out, albeit hot water is routed temperature sensor since the pathway through radiator is blocked?

Do these sound right? If so, then only thing I can think of that is restricting the flow is clog in the radiator or somewhere in the line, while I don't know how such a clog suddenly develops. Is there any way to check whether there is air pocket in the system, except re-draining and refilling?

Thanks for your help!

[SSSVX]

Did you say that you have changed a new water pump and the thermostat?

If I were ya, I wud do a few more flush or take it to the shop to the power flush on the radiator/entire cooling system...and see how it works later...

[redsvx94]

No, I changed Thermostat but not the water pump, since it is a lot more work for me. I was hoping to rule out the possibility of headgasket leakage or other possible reasons before I start to do water pump and other things that are better to be done along with it. If it is head gasket leakage, I won't have money to fix it.

My car overheats pretty quickly, after running idle condition for 5- 10 min. Any body know whether these are what likely symptom of head gasket leakage or having some air pocket in the cooling system?

[Earthworm]

Take the car to a garage that can check for hydrocarbons in the antifreeze. This is evidence of a cracked head/headgasket allowing exhaust to enter the cooling system.

If the water pump is working properly the top radiator hose should be hard due to the pressure in the system. This is assuming the previous test shows no hydrocarbons.

[redsvx94]

Thanks for reply, Earthworm. I've had not so good experiences with garages near by, so I try to do myself as much as I can. I heard about the HC tester sold in Napa, but they were closed yesteday. Will get that tommorow if I don't start on other things today.

I saw a previous post by subafreak stating that if the top hose gets hard within 1-2 minutes from COLD start, suspect headgasket leak. I suppose that is due to hot exhaust leaking into coolant?

http://www.subaru-svx.net/forum/show...ght=headgasket

I guess what you mean is that its should get harder IMMEDIATELY after turn on the engine, right? I think mine did not get harder immediately nor after 1-2 minutes, but I think i got harder as the engine temp get closer to normal. Would you have any sense how hard it should be?

[SVX-FF]

To check for a head gasket leak is easy, but not so easy on the boxer 6.

With all the coil packs disconnected remove one spark plug at a time. Hook up a compression gauge and measure each cylinders pressure. About two to three cranks will give you a good average reading. This is a two person job. One to crank while the other one watches the gauge.

If there is a cylinder that has a bad head gasket then you will see a lower compression reading from the others.

Good luck

Steve

[redsvx94]

Thanks for the tip, SVX-FF. This seems to be sure a fire way of checking headgasket leaks. But unfortunately i don't have a compression guage. How expensive is it? Also is there any step by step instruction for this? Is it difficult to disengage coil packs? Also I have not yet rotated the engine that way.

[redsvx94]

O.K. David,

I just looked at the NAPA website. A compression gauge doesn't seems to be too expensive and a good thing to have. If I can't find the HC test kit, I would like to try the compression testing..

-John

[Beav]

Too many things can cause a low(er) compression reading - sticky piston rings, a small valve leakage, etc. A head gasket leak that causes overheating doesn't have to be very big and cylinder-to-cylinder compression tolerance is normally 20%. So, at roughly 200psi a 20% difference would be 40 psi. How much lower than that would you consider necessary to condemn the head gaskets?

If you were searching for a cause for a misfire and found adjacent cylinders with low compression then you could surmise a head gasket leak, otherwise a compression test is pretty useless for overheating diagnosis.

However.... pulling the spark plugs may reveal a head gasket leak. Look for exceptionally clean or plugs with a greenish-white coating.

[SVX-FF]

Originally Posted by Beav

However.... pulling the spark plugs may reveal a head gasket leak. Look for exceptionally clean or plugs with a greenish-white coating.

Spank me stupid while riding a big old donkey butt.

Now why did I not think of that.

Steve

[Beav]

Well, if the plugs don't reveal anything it doesn't necessarily mean anything. Compression pressures are a bit stronger than coolant pressure...

[redsvx94]

O.K. thanks for the clarification, Beav.
I got a Block tester kit today from NAPA. This is the type that fume from coolant are drawn into a transparent cylinder in which the testing liquid changes its color from blue to yellow if exhaust gas in present in the fume.

But I wonder if circulation of coolant is limited ( as I suspect it is ), the exhaust gas won't make it to the tester and so, I won't know whether there is leak, even if I don't see the color changes? Does any one have a experience in this thing with non-circulating coolant?

[svxcuseme]

Are you refilling the radiator or resevoir frequently? Do you have fumes from the exhaust, white smoke from the tailpipe that dissipates in the atmosphere quickly? IS your oil, milky looking (water and oil mix). If not, then It's more likely to be something much simpler like the Radiator or water pump, then a head gasket.

Assuming you go through with a compression test and they turn out O.K., I would replace the radiator before the water pump, especially if it's the original. My two cents. You can have coolant flow through the system, at a decreased rate with a good water pump and clogged raiator. I've seen this this more than once. Do it yourself flushing really does nothing and you can't have a shop soak it in an anti-sediment bath as we have the plastic tanks.

I realize the part is more expensive but it's much easier to get to and involves much less work. Besides, replacing a 10-14 year old radiator can almost be thought of as maintenance if you're unsure of the vehicle history and how the coolant has been maintained. On top of that you see the SVX radiators cracking all over the place. Replacing the radiator is nothing but good and could possibly save you a lot of headache!

One last thing, if you've replaced the thermostat and the car is still overheating, depending on the degree of overheating that is occurring, you could be damaging the new thermostat and causing it to slowly fail again. While you're trying to resolve your problem, I would remove the thermostat from the equation entirely. You can buy a generic thermostat housing gasket from Advance or Autozone for a couple bucks (ask for one for an 86 Subaru GL) . This way, you'll know the thermostat is not compounding the issue. It's simple to drop it back in once you find the issue. Removing it will not cause any problems to the engine or cooling system for the short term.

[redsvx94]

> Are you refilling the radiator or resevoir frequently? Do you have fumes from the exhaust, white smoke from the tailpipe that dissipates in the atmosphere quickly? IS your oil, milky looking (water and oil mix).

I don’t have any of that. I flushed the radiator, putting tap water where the hose connects to the top. While some dirt water was coming out initially, good amount was coming out, not drizzling.
On the other hand, I’m having overheating pretty immediately even under idling. So I also feel that it’s the water pump or clog somewhere in the system ( but may be not radiator due to above.)

Reason why I’m concerned about headgasket is that with stagnant water, I really don’t know how hot the engine has gotten. Also I saw some posts in the forum that describe some of the symptom like my car’s, and find there was headgasket problem ( although leak may have developed due to overheating problem, not the cause of the original overheating ), even though they didn’t have signs of oil in coolant and vice versa. It could just leak exhaust gas into the coolant.

If I know my block / headgasket is fine, I may be interetsed in replacing the radiator, even those aluminum ones. But if the headgasket is bad, I really don't have money to cover that, since I can't do it myself.

I couldn’t do compression test, since gauges I could find don’t seem to be able to fit in the spark plug hole on SVX. I need to screw in the 14mm adaptor to the hole and tightly screw in the tip from the gauge to the adaptor, but our spark plug hole is too deep and the part of the tip that I can put wrench on is way in the hole. Besides Beav seems to think that compression test will probably not reveal headgasket problem.

Thus my question about the block tester as the post below. May be I should start another thread since this is a specific topic that I couldn’t find info about in the Forum.

One question, bottom of my radiator is warm / cool. I guess the thermostat never got hot on that side. Would the water in the water pump side get hot enough to fail the thermostat?