I have a 1971 280SL daily driver with AC that overheated badly with the AC on in our South Georgia heat. After confirming that all the basics were okay (timing, fuel mixture, viscous fan coupling, fan belt, water pump, oil cooler) I moved up one level to "aftermarket" stuff. Water Wetter (by Redline) combined with a 20% antifreeze mixture seemed to help a little, but not much. I installed a small electric booster fan in front of the radiator. I made a special bracket that let me leave the AC condenser in place. The fan helped a little, but not much. Increasing the idle rpm and shifting to neutral at stops helped most, but my wife (who drives it to work and in her job travels) balked at the hassle. If you increase the idle and do not shift into neutral the overheating is made worse because some of the additional energy is turned into heat in the transmission and ultimately ends up back in the radiator.
Several things suggested that the problem was poor coolant flow. One, I pulled the head to learn why #4 cylinder compression was so low (105# vs 160 for the other five). I knew it was that way when I bought it with about 100,000 mi. The cylinder walls of #4 were scored badly in a way that indicated that the cylinder walls had overheated so severely that the rings wiped molten metal off them. Yet the other cylinders were undamaged. This suggested that the area around #4 got much hotter than the others. That area just happens to be near the temperature sending bulb, indicating that the guage was reading the hottest part of the engine.
Two, as others have commented, when you rev the engine after sitting at idle you get a quick drop in guage temperature. That suggests that cooler water is near the guage but is not getting there at lower rpms. If cooler water hits the guage that quickly, it must already be cool since the few seconds delay can't be enough time for increased air flow from the fan to cool the radiator water very much. Therefore, the radiator water must be cooling properly, even at idle. That's why an electric fan helps only a bit. That's also why Water Wetter and low antifreeze ratios are only marginally helpful.
Three, a hand held pyrometer showed that the area around #4 was 10-15 degrees hotter than at the front of the engine. Yet one would expect the front to be hotter because the water in that area has (to some extent,at least) already traveled past the rearward cylinders.
Then, while cruising the web I found a discussion on another site (no longer around) that referenced changes by MB on the late 280 cylinder heads to reduce overheating, and a reference to a "kit" from MB that could be installed.
I found one of the kits on line, but they wanted $400 for it and I could not learn enough about it to be sure it worked.
So, I experimented with my own. I removed the thermo-time valve located in the rearmost threaded boss on the cylinder head. This exposed an opening connecting directly to the water jacket. I installed a 3/8 hose barb fitting in the opening and then ran the hose to a "T" fitting which I made and installed in the upper radiator hose. This allowed water to travel directly from the rear of the cylinder head to the radiator without travelling through the cylinder head.
The effect was dramatic. The car would idle in 100 degree weather, in traffic, with the AC on, and the temperature would get near, but never reach, the red. In more normal driving, it stayed around 190 degrees.
I later saw a picture of the MB kit and it was a much more professionally made version of mine. Furthermore, it routed the water from the rear of the cylinder head to the heater return line which in turn directed the water into the block just below the thermostat. I could not understand why MB didn't send the water directly to the radiator, like I did. It seemed to me that you would get higher flow from my system because of the shorter route and fewer twists and turns.
That winter, I learned that (of course) MB knew a whole lot more than I did. With my bypass line in place, the engine would never warm up. Even with the thermostat fully closed (because the engine was just started), on a cool day (say 50 degrees), enough water flowed through my 3/8" hose to keep the engine temperature below 100 degrees even on the highway. That showed just how much heat a small stream of water can carry away. And it explained why MB routed the water through the heater lines and ultimately the thermostat. In their system, the thermostat controls all the water leaving the block, including that from the bypass hose. My jackleg solution was to put a globe valve in the hose. In summer I open the valve and let the water flow freely from the rear of the head to the radiator. In winter, I close the valve and there is no flow at all.
This has worked well enough that I have not gone further, except to make a fixture to hold the thermo time valve which can no longer plug into the head because the bypass hose take up its spot. The thermo time valve is no longe immersed in coolant inside the head, but it is touching the head and seems to activate the cold start valve as well as it did before, so I assume it must be reaching about the same temperatures outside the head as inside.
The real test came last Monday when my wife had to make a long trip on short notice. She came home Sunday complaining of noise from the AC. I checked and found that the AC compressor clutch bearing was going bad, and threated to lock up at any minute. All the shops were closed, and she had to leave early Monday morning. In this system, the AC belts also drive the fan and indirectly the alternator. So I could not just remove the belts and let the compressor sit idle. Furthermore, it was supposed to be 98 degrees for both days of her trip and wanted her to stay cool. Since I plan to replace the original York compressor with a new Sanden anyway, I just welded the clutch together so it would always be engaged. Didn't even have to take it off the car, since the front of the clutch is accessible just below and behind the radiator.
Since the clutch is permanently engaged, the AC compressor runs all the time. You can't turn it off. My wife was delighted. She said it stopped rattling (I knew it would do that), cooled better than ever before (a surprise because I assumed the clutch was engaged all the time anyway), and she didn't have to remember to turn the AC off when cranking (which I made her do to reduce the starter load).
If I have time I would like to install a coolant recirculating pump (like those in the later MB heater systems) in the bypass hose just to see what happens and a more effective radiator. A supplemental transmission fluid cooler should also reduce the load on the radiator. But the results with the simple bypass hose and fan have gotten us by for several years, and there are more pressing problems (like the welded up AC clutch! and the door locks with minds of their own).
If anyone is interested, McMaster Carr has a metric to NPT adapter that screws right into the threaded boss (14x1.5mm, I think) at the rear of the head and makes attaching the hose very easy. The "T" fitting for the upper radiator hose had to be made, but it is easy with a welder, and I might be able to figure a way to make it without welding if anyone is interested.
Charles H. Ford, Jr.
