'83 300D Running Warm

[jdh122]

My '83 300D has been running warm lately. For example, on a drive today from Tulsa to Houston the gauge reached ~100C when it was probably 95F out and more or less stayed there for around 6 hours (I maintained the posted speeds of 65-70 the entire time). The thing is, it ran warm even before this ridiculous heat settled over us and I have already replaced almost the entire cooling system. Here is what I've done so far:

- Citrus flush (twice)
- New Behr radiator
- New Behr thermostat (checked prior to installation)
- New upper and lower radiator hoses
- New thermostat bypass hose (the old one leaked from the top)
- New hose from expansion tank to radiator
- New expansion tank (old one looked old and dirty on the inside)
- New expansion tank cap
- Replaced old metal engine fan with new 9-blade plastic fan (checked the fan clutch and it appears to work correctly)
- Rinsed the condensor fins

The only major part I haven't replaced is the water pump, but I'm reluctant to do that when it shows no real signs of failure (i.e. no leaks or odd noises).

Also of importance is that I seem to be losing coolant, although not as quickly as I was before I replaced the leaky bypass hose. When I disconnected the thermostat housing from the block in order to install that hose I noticed that the old gasket between the housing and the block was basically completely gone and I'm wondering if coolant is seeping out of that area and evaporating before it is able to pool on the ground. I had also thought that the head gasket might be bad, but the car makes no white smoke and there are no signs of coolant in the oil or oil in the coolant.

One other symptom is that sometimes the car will maintain an acceptable temperature - say, 90C - but then after I stop for 5 minutes to refuel and restart the temperature will read close to 100C and won't come back down at all even after I start driving.

Oh, one last thing: the PO installed one of those "Euro-style" grills on the car and I'm wondering if it is allowing proper airflow over the radiator. Could this make much difference in the radiator's effectiveness?

My conclusion is that I need to have the system pressure checked and the block temperature measured with one of those IR thermometers to ensure that the gauge is correct, but does anyone else have any ideas?

Also, would a small leak in the system account for the increase in operating temperature, assuming there was still adequate coolant in the system?

[jt20]

It sounds like you know what you're doing. You've covered all the major failures except the pump, which should weep when it starts to go. There have been reports of eroded vanes on the pump, however. Its a long shot, but just food for thought.


I agree that you should verify the accuracy of your sender and the gauge before jumping to any more conclusions or throwing more money at it.

A small leak will not directly cause overheating.

[jdh122]

I purchased an IR thermometer at Napa and took a few readings. Here are the numbers (other factors: the ambient temperature is 33.3C; I took readings in the shade with the car idling; the car had been idling for about 5 minutes by the time I took all of the readings):

Gauge: ~95C
Block immediately to the left of temp sensor: 90.4C
Upper rad hose at thermostat housing: 86.0C
Upper rad hose at radiator interconnect: 74.4C
Lower rad hose at thermostat housing: 71.2C
Lower rad hose at radiator interconnect: 65C

Clearly the hoses were warmer near the t-stat housing due to the heat radiating from the engine. Do you think that the ~5C difference between the gauge reading and the block reading is significant? It seems like the coolant coursing through the block could easily be a little warmer than the outside of the block itself.

[jt20]

keep it in the car and check those numbers after a good run.

I agree that the coolant (as you tested it) is hotter than the head surface.

[winmutt]

Search around there is an OHM chart with readings for the sensor. Also, double check ALL of your hose clamps for tightness. With the heat on do you smell coolant?

[jdh122]

I found the ohm chart:

The temperature sender would measure ohm resistance @ temp.Celsius as follows (roughly,+/-5ohm):
60C-110 ohm
80C-67 ohm
90C-51 ohm
100C-38 ohm

I'm sort of new at this - how do I go about conducting this test?

[jt20]

make a double boiler, put the temp sender in the main cup along with a mercury type thermometer, attach an ohmmeter / multimeter to the sender, watch resistance values as the temp rises.

[kerry]

Temp senders are cheap enough that I'd just swap in a new one to see what it reads.

[jdh122]

I rented a pressure tester (for free, incidentally) from Autozone. First I checked the expansion tank cap, which - unsurprisingly since it is brand new - tested fine up to the 1.4 bar it's designed to withstand.

After tightening all of the clamps, I then tested the system itself. I couldn't find any instructions on how much pressure to use so I pumped it up to 19 psi. Not sure if this was enough pressure to uncover any leaks, but the only thing I saw was the VERY slight seepage out of the top of where the thermostat housing meets the head, which I already knew about and to repair which I've purchased a new gasket.

The system held more or less steady at 19 psi for the 10 or so minutes I left the device attached to the expansion tank. Again, not sure if I was supposed to leave it on there longer.

Also, I ran the heater and it is odorless (i.e. no coolant smell).

Any other suggestions re my coolant loss problem?

Tomorrow I'll check the temperature sender as suggested to see if it is functioning correctly.

Oh, I recall now that I've seen the temperature and fuel gauges "flicker" up and down a couple of times in the past, and based on what I've read this could indicate a bad ground. I suppose I should try to fix that too, although I'm not sure if that would result in a fault reading on the temp gauge?

[carnut]

FIRST THING I NOTICED ON YOUR 1ST POST IS A FAULTY GASKET THAT WASNT REPLACED. 1ST MISTAKE. CANNT TALK ABOUT TEMP BEING STABLE OR TOO HIGH WHEN AIR IS GETTING IN FROM THIS SEEPING GASKET. A PRESSURE TESTER SHOULD NEVER BE PUMPED UP HIGHER THAN THE CAP PRESSURE, IN THIS CASE 14 LB. ALWAYS A GOOD IDEA TO LEAVE THE TESTER ON OVERNIGHT. SOME LEAKS ONLY SHOW WHEN THINGS ARE COLD. QUITE FRANKLY 100 DEGREES C ON A 95 F DEGREE DAY IS NOT ABOVE NORMAL. CHECKING THE MULTIWIRE GROUND BEHIND THE CLUSTER IS A GOOD IDEA TO SIMPLY STOP THE ERRATIC NEEDLE MOVEMENT AND A POOR GROUND WOULD ALSO HAVE HI RESISTANCE WHICH COULD CAUSE FAULTY TEMP GAUGE READINGS. IF THIS CAR IS NOT BOILING OVER WHEN SHUT OFF OR AT IDLE IT IS NOT OVERHEATING, JUST RUNNING 5 TO 8 DEGREES ABOVE IDEAL. EXCUSE THE CAPS, BUDDY'S FAULTY COMPUTER!!

[jdh122]

Originally Posted by carnut

FIRST THING I NOTICED ON YOUR 1ST POST IS A FAULTY GASKET THAT WASNT REPLACED. 1ST MISTAKE. CANNT TALK ABOUT TEMP BEING STABLE OR TOO HIGH WHEN AIR IS GETTING IN FROM THIS SEEPING GASKET. A PRESSURE TESTER SHOULD NEVER BE PUMPED UP HIGHER THAN THE CAP PRESSURE, IN THIS CASE 14 LB. ALWAYS A GOOD IDEA TO LEAVE THE TESTER ON OVERNIGHT. SOME LEAKS ONLY SHOW WHEN THINGS ARE COLD. QUITE FRANKLY 100 DEGREES C ON A 95 F DEGREE DAY IS NOT ABOVE NORMAL. CHECKING THE MULTIWIRE GROUND BEHIND THE CLUSTER IS A GOOD IDEA TO SIMPLY STOP THE ERRATIC NEEDLE MOVEMENT AND A POOR GROUND WOULD ALSO HAVE HI RESISTANCE WHICH COULD CAUSE FAULTY TEMP GAUGE READINGS. IF THIS CAR IS NOT BOILING OVER WHEN SHUT OFF OR AT IDLE IT IS NOT OVERHEATING, JUST RUNNING 5 TO 8 DEGREES ABOVE IDEAL. EXCUSE THE CAPS, BUDDY'S FAULTY COMPUTER!!

Thanks. The max cap pressure is 1.4 bar, which is about 20 psi. I'll go ahead and pump the pressure tester up to 15 psi or so and leave it on overnight. I left it on for 10 minutes only earlier because the instructions that came with the device said something about leaving it on for 2 minutes only. That sounded kind of short to me and frankly didn't give me enough time to do a close inspection, so I left it on for 10.

Surely when you take a car to the shop to have this done they don't leave it on overnight though??

[carnut]

Many, many times the tester was left on overnight especially when the customer said no evidence of coolant on the driveway. Some leaks stop when hoses swell up warm. The system is designed to be under this pressure continously so leaving it on presents no harm to the system.

[jdh122]

Last night around 10 p.m. I applied the pressure tester and pumped it up to 17 psi. This morning at 7:15 I checked things out and found the following:

1. Pressure had fallen to 14 psi
2. Slight seepage from the thermostat housing
3. Slight seepage from the top of the thermostat bypass hose

#2 was already known to me. #3 was not, as I had had this problem before and replaced that section of hose about a month ago, which apparently did not fix the problem. When it had leaked before it was bad enough that it left a puddle under the car. This time it was only just seeping out the top, and hadn't accumulated enough to drip or anything like that. The garage floor under the car was completely dry.

Are the small leaks I describe enough to account for the 3 psi drop in pressure over a ~9 hour period?

If I had a head gasket issue would the pressure have likely fallen further?

[tangofox007]

Originally Posted by jdh122

Are the small leaks I describe enough to account for the 3 psi drop in pressure over a ~9 hour period?

A temperature drop overnight would contribute to the pressure reduction.

[RAYMOND485]

1984 300D TURBO, 147K
YOU MAY HAVE AIR IN THE ENGINE BLOCK, PUT AUTO FRONT UP AND REMOVE TOP RADIATOR HOSE AT RADIATOR AND FILL WITH WATER AND FILL
OVERFLOW TANK.