: Radiator leaks and engine runs rough, no codes



kjhansen
09-13-09, 09:46 PM
Well, 189,000 seems to be the magic mile marker for my 1995 Seville SLS. Suddenly I had an overheating problem which culminated in rough running but no codes--but they may not be related.
I was driving to town a few days ago and suddenly the engine temperature started rising (I always keep the temperature on the information display). It zoomed up to 252 degrees, then cooled down to about 235 -- which is still too warm, but shouldn't damage anything... . In any case, I only drove about 5 miles total, then stopped and let it cool down for about a half hour then drove home. On the drive home it stayed pretty much in the 230-235 range. When I got home I popped the coolant tank lid and tried to add some coolant. It pumped it out under considerable pressure and made quite a mess. Not bubbles--a pulsing stream. I've seen this on other cars and it's always been the thermostat frozen shut. It forces the coolant back against the direction of the water pump pressure, over-pressurizes the system, prevents coolant from circulating through the radiator thus overheating the engine and blowing coolant out the tank lid. So I replaced the thermostat and that seemed to fix the cooling problem--in fact the engine now runs cooler than it has since I bought the car about 2 years ago. I guess the thermostat was failing even then. Only now the radiator leaks and the engine has developed a bad miss. I can see why the radiator might have developed a leak--it was over-pressurized by the water pump and overstressed by the heat. But why would the engine start running rough? I thought it might be a head gasket, but a) I don't seem to be getting coolant in the oil (no anti-freeze smell in the oil and the oil level isn't going up for no apparent reason) and b) I can't see or smell anti-freeze steam coming from the exhaust. I've had both conditions before... in GM products too.
So, what are the other options? A sudden need for a tune-up (new plugs and wires)? A failed ignition module or coil pack? These seem like unlikely coincidences. What are the recommended checks? Compression? Sniffing spark plug holes seems difficult for the rear bank... Has anybody had this sequence of cascading events? Any help appreciated!

Keith Hansen
1995 Cadillac SLS (up to now one of my most favorite cars)

Submariner409
09-13-09, 10:05 PM
Before you start throwing parts into a tune-up or more cooling system parts, either check or have a shop check the airspace over the coolant in the surge tank - you're looking for the presence of exhaust gas. Head gasket. Sudden overpressures and boilovers are the telltale. The rough running is coolant leaking into one or more cylinders. Very seldom will a Northstar put oil into the coolant or vice versa with a bad head gasket. The blown gasket bleeds gas into the coolant passages, not the crankcase, and there's no way for oil to get into the coolant.

Skiller.
09-13-09, 10:06 PM
Quick question...12 o'clock on the temperature gauge is 210 degrees?

Ranger
09-13-09, 10:07 PM
First off, you cannot over pressurize the cooling system. The cap will vent at 16-18 psi.

The Northstar rarely ever puts coolant in the oil when a head gasket goes, so do not use that as a benchmark. Are you using coolant? A leaky head gasket will run rough (coolant ain't flammable), but that is usually only at start up. Might want to get a block test kit from Napa and test the coolant for exhaust gases to either condemn or rule out the gaskets.

http://www.cadillacforums.com/forums/attachments/northstar-performance-technical-discussion/52467d1252411936-super-noob-needs-help-overheating-smallertempgaugewithnumbersupdated.jpg

Submariner409
09-13-09, 10:11 PM
Once again, with feeling:

This pic is posted in here, Deville, and Seville 14 times in the past 3 months. Snit over...........

Ranger
09-13-09, 10:13 PM
Oops. I went back and edited it to add the gauge, but I should have known you'd be all over it Sub.

BTW, mine's bigger than yours. :)

Submariner409
09-13-09, 10:16 PM
Overpressure.................:sneaky:

Skiller.
09-13-09, 10:23 PM
Once again, with feeling:

This pic is posted in here, Deville, and Seville 14 times in the past 3 months. Snit over...........

Thanks, and sorry! I just recently joined :D

Submariner409
09-14-09, 09:14 AM
Not jumping in your particular case - just a gentle barb to get you to pop a brew and go back in time and research old threads and use the search button up in the blue bar ^^^. Informative rainy day noodling.

Did you ever go up and read the entire Cadillac Technical Archive in the black bar ^^^ ? How about all the sticky threads at the top of Seville, Deville, and here ? How about down in "Site news, feedback......" the thread "Forum rules and guidelines" ?

kjhansen
09-14-09, 05:39 PM
Back to the original thread: The car DOES NOT push back through the cap anymore since I replaced the thermostat, and it runs cooler than it did before I replaced the thermostat (it was running 212-217, now it runs 199 or less to 208), and it doesn't smooth out at higher rpm under load (i.e., driving down the road)--still stutters and runs rough uphill or accelerating at all. Not saying that it's NOT a head gasket though and I will test for that if I can get the test kit at O'Reilly or Autozone (the only parts stores in town) or find an honest mechanic in this place. It seems they'll tell you anything to get the car in their shop and get their wrenches on it. I had one guy tell me the upper ball joint was out in a car with McPherson strut front suspension.. no upper ball joints. I had another mechanic steal the broken transmission out of my Turbo Regal and replace it with a cheap Oldsmobile transmission... Long story. Anyway there's a goop-style, radiator additive headgasket sealant at O'Reilly that a friend of mine swears by--I'll try that, but I am definitely not pulling the engine to replace a head gasket, so if it's the head gasket there'll be a 1995 SLS for sale cheap for parts pretty soon... Hey I only paid $2500 for the car a couple of years ago and I've put lots of miles on it, so that's the way it goes sometimes. I've got my eye on a 2002 Deville for $4000 or maybe less...
Keith

Ranger
09-14-09, 06:21 PM
Did you replace the plugs & wires?

Forget the head gasket repair in a bottle. It won't work.

kjhansen
09-15-09, 11:51 AM
Haven't replaced plugs and wires yet. I was going to test for the combustion leak first, either with the radiator test (tests for combustion gas in the cooling system), or do a compression test on each cylinder. While my normal method of maintenance is to throw parts at a problem until it's fixed, I don't want to put a lot of money and effort into a car that I won't keep. Best to diagnose first.
I've never had permanent luck with any kind of leak sealant either, but this kid tried this new product and it worked for him. Granted he has an older, low compression Chevy truck... I did try a head gasket sealant that worked on my 1970 Buick Riviera... but only for about a week.
Keith

kjhansen
09-16-09, 08:31 AM
Latest update: I was able to rent a combustion leak test kit from Autozone for $25. I also had to buy the fluid ($8), but I'll get the $25 back when (and if) I return the kit. Anyway, the fluid is blue and supposed to turn yellow when contaminated with combustion gas. I tried it, but I don't think I'm getting any fluid flow through the radiator overflow tank. The reason is that after pumping the bulb a number of times and getting bubbles, the tank runs out of air. No more bubbles. This is with the engine fully warmed and running at about 203-208 degrees. Coolant should be flowing, and as it flows, it should be picking up air (contaminated or uncontaminated) from the system, but that's apparently not happening.
Question: Am I going to have to remove the overflow tank and make sure I've got fluid flow or is there some way to force any blockage from the two lines that come into it without removal? One line appears to be a heater hose, which comes in low on the side, and the other a small hose from the block(?) that comes in high in the front.
Keith

Submariner409
09-16-09, 10:21 AM
That coolant surge tank should be half full, no more, cold. That assures the proper expansion and airspace over the coolant in the tank.

Bring the coolant to the proper level and drive the car for a while - you cannot do the test until the coolant and airspace is (possibly) saturated with combustion gas. It takes only a few seconds for the tank gas, bubbling through the test fluid, to change the fluid color if a problem exists.

I believe the coolant should be warm, with the engine OFF.

The small hose is a purge line which should constantly piss a small stream of coolant into the tank. That action passes air and gas from the coolant to the surge tank airspace, thus lowering water pump cavitation from bubbles in the main system. The lower line is the system-to-tank connector and it will only "flow" that liquid which comes from the purge line. It is designed to only allow for coolant expansion and contraction, not flow, with temperature. You have a 16# surge tank cap, I believe.

kjhansen
09-16-09, 12:49 PM
OK, well I took the system apart and cleaned it anyway--couldn't hurt. You're supposed to do the test with the engine running and warmed up. Once you take the cap off you've allowed ambient air into the tank, so you have to seal the testing tube against the hole of the tank and pump the bulb for two minutes. This purges the ambient air and then you start drawing the air from the coolant--if there is any--and it bubbles through the testing fluid. After cleaning the hoses out and refilling the radiator, I was able to do that. I got air flow (bubbles through the test fluid) more slowly once ambient air had been purged, but air flow nevertheless. The fluid stayed blue. After running the test I put the testing tube up to a tailpipe and pumped it for a while. The fluid turned yellow as it was supposed to as it drew in combustion gases. I guess I don't have a blown head gasket. On to other things. Plugs and wires, coil packs, ignition module or computer.... Sigh. Cheaper anyway.
Keith

Ranger
09-16-09, 01:02 PM
That's good news. All the other stuff is MUCH cheaper and easier.

kjhansen
09-17-09, 12:51 PM
I replaced plugs and wires, and even though the previous combustion test came up negative, I'm still not sure I don't have a blown head gasket. The plugs on cylinders 3 and 5 were clean (more or less) while the ones on the other cylinders were coated with gray residue, which is normal.

http://i223.photobucket.com/albums/dd82/kjhansen57/Car%20Pics/twoplugsdifferentbanks.jpg

Left plug is from cylinder 3 or 5
Right plug from an even numbered cylinder in the front bank.

The engine did smooth out a bit with new plugs and wires, but still runs rough and stumbles under load. Several of the coil pack contacts were corroded, and I cleaned those up. It really feels like an ignition problem. Still not using coolant or running hot or bubbling in the radiator overflow tank...

Keith

Submariner409
09-17-09, 04:14 PM
Are the plug wires hooked up according to these diagrams? It's important.

Google "waste spark ignition" to find out why.

Is that left plug wet with coolant? Not good.

kjhansen
09-17-09, 07:21 PM
Coil pack contacts are numbered and the front bank of plugs is numbered. The only ones you could mix up are the rear bank, and they're 1, 3, 5, 7 on the valve cover (although not numbered) and numbered on the coil packs. In any case, I wrote the plug number on each wire with a sharpie so I wouldn't cross-wire them. I've been working on cars (backyard wrench) for over 40 years... and my last car was an '87 Buick Turbo Regal so I became intimately familiar with the waste-spark ignition system.
The plug wasn't wet--just looks like it in the picture. Didn't smell of coolant either. But it is CLEAN which is a bad thing too, because it might be steam-cleaned. Sigh.
What is the reading supposed to be when you ohm the coil packs, and do you have to ohm top and bottom? With the Buick it was easy: you just read the ohms across the towers on the individual coil pack. On the Bonneville SSEi I used to have you had to ohm the resistance tower-to-tower and between contacts on the bottom of the pack too.
Next thing is to run a compression check. Saturday probably. Too tired today.
Thanks for the input.
Keith

Ranger
09-17-09, 10:43 PM
Aren't whitish plugs a sign of coolant in the combustion chamber?

kjhansen
09-18-09, 06:16 PM
They're more gray than white, and, in my experience, that indicates normal wear.

kjhansen
09-19-09, 12:00 PM
Resolved! The overheating and running rough were two apparently unrelated issues. As previously noted, the first was the thermostat, which fixed the overheating problem. The second turned out to be one bad coil pack--the last one on the right as you look at the engine with the hood open. I ohmed all of them and they all read 6.3 something ohms, except that one which read 1 (no reading). Bad sign. I took all of them off and cleaned the contacts thoroughly, tried again and got a 6.17 from the bad one. Not bad. When I drove the car it started out smoothly, then stumbled again after it warmed up. So I figured that one coil pack was breaking down under load. So I bought one coil pack and tried it in each position just to make sure. Needless to say, it was the one that ohmed bad in the first place. Now the car runs smooooooth, has power back, and the gas mileage is going back up as I drive.
Oh, there is still that coolant smell and what looks like a small leak in the radiator. But that can wait... .
Keith