Once again chasing a cooling fan problem on my friend's 1992 740.

The relay that controls the radiator fan is new.

It would appear that the temperature sensor is defective. With the two wires connected to the sensor the fan runs continuously. Once one of the leads is removed from the sensor the fan stops. Bypassing the sensor and connecting the two wires the fan once again engages. So it would appear the sensor is defective as it is constantly "closed" even with cold coolant.

When he sent me a photo of the radiator showing the sensor I was puzzled as I have never seen a set up like this before. It must be an aftermarket radiator as all the ones I have dealt with had the sensor on the back side with the grommet and sensor placed in an opening in the radiator.

I hesitate to have him try to remove that large hex (plug?) for fear of it damaging the radiator shell. I don't know if the sensor will simply pull out with a little effort.

Does anyone have any experience in dealing with radiators of this type or the sensor in particular? Thanks for any help.

EDIT: Yes, we need to deal with that coolant seepage.




--
Any twenty minute job is just a broken bolt away from a three day ordeal

hi randy,

that unit in the radiator is a thermal switch, open cold, closed hot.
if it is reading closed with your ohmmeter when cold it's junk.

bill

Yes, that is the problem it is closed at all temperatures.

Randy
--
Any twenty minute job is just a broken bolt away from a three day ordeal

can you tell the manufacturer of that radiator, they might have the replacement for that switch.

Looks like it might be the stock switch. Rad shops wouldn't want to worry about having a switch with the correct temperature and electrical connector. Would only need suitable rubber washers/o-rings to keep it sealed.
--
Dave -still with 940's, prev 740/240/140/120 You'd think I'd have learned by now

I recall that photo. I wondered if it was made by a manufacturer that sort of "adapts" their basic radiator to fit multiple cars. They may use hex plugs of varying inside diameters in order to make it work with whatever sensor the buyer's car is equipped with. I'm just guessing. Regardless, I'd be afraid to give that thing a twist also.

You know the sensor is bad so you don't have much to lose if you break it. You might be able to grab one of the spades with a pair of needle nose pliers and carefully give it a little tweak up and down to see if it's got a rubber O-ring between it and the inside of the hex plug. If it wiggles independently of the hex plug then maybe you can wiggle it around a bit while pulling it out. Can you grab each spade with a separate pair of pliers?

Thanks to both of you for your responses, and sorry for the lack of the images I had originally posted. I have added three images to this reply.

This is the type of radiator I am familiar with and it is on the 1991 RexRegina car I keep as a spare:



This is the radiator on my friends car that has the defective temp sensor:



The sensor seems to be "captured" and getting to it appears to be by unscrewing the crusty hex cap. He has not had the opportunity to dive into that as of yet. He is running the fan full time when driving the car in this hot weather.

After going through my spares I found a radiator with this setup.


I'm thinking this is similar to the radiator with the bad sensor. It would capture or enclose the sensor in a way to seal against leakage but expose it to the coolant.
--
Any twenty minute job is just a broken bolt away from a three day ordeal

Yes, almost certainly a failed rad thermal switch.

Interesting, I haven't seen a rad without the thermostat mounted in a rubber press-fit collar on the rear at the top right corner, but there are obviously variants out there from other manufacturers sized to fit the rad mounting and trans cooler lines. Doesn't really matter where the thermostat mounts as long as the wires reach (or are extended). Those switch mountings you show would actually be a slight improvement in design as the stock switches are known to blow out under excess pressure (actually had that happen once).

I wouldn't want to force the mounting collars on an aging plastic tank if they don't come off with only moderate force after using something like PB-blaster. Restraining the spade connectors as a counterhold would certainly help, doesn't matter now if they get damaged or break off. Unlike the trans cooler lines, which are solidly attached on the inside to the metal core structure and can take quite a bit of force, I rather suspect there is just a smallish metal (or nylon) flange as reinforcement inside the side tank to support the types of switch mountings you show.

One DIY thought I had as a hack for your friend to avoid buying an expensive new rad is to leave the old thermostat in place and instead use an external thermal switch surface mounted on the rad side tank. These can be had for as little as a couple of dollars on places like Amazon. You would need a normally open thermal button switch (thermostat) that has a suitable temperature, rated for at least 12 volts DC (many are rated for 240 volts AC) and can handle perhaps one amp (many can handle well over 10 amps). Being surface mounted it wouldn't be as sensitive to actual coolant temperature as the stock thermal switches which has direct contact. Something that kicks in not too far above maximum ambient air temp, say 50-60 deg C, would detect when the coolant thermostat is open and the fan might be needed. Initially it could be surface mounted on a flattish area of the side tank using a zip tie between the spade terminals wrapped around the side tank through the core (as many recommend for the stock sensor). Once you're satisfied that it's working you could later expoxy it in place to protect it from the elements (JB-Weld, add a zip tie for good measure)) with perhaps a little electronic thermal paste (used with power transistors and CPUs) between the button surface and the side tank to improve thermal conductivity. See my other reply in this thread to Bob Ritter suggesting that operation of the thermal switch may not be all that critical.

https://www.jaycar.us/normally-open-thermostat-switch-50-degrees/p/ST3831

--
Dave -still with 940's, prev 740/240/140/120 You'd think I'd have learned by now

For some reason, the pictures you posted are not showing up on my end. But it is neither here nor there. I have a 93 940 with Regina ignition and a similar set up on the radiator fan switch. The radiator in the car when I bought it had the switch on the side. I recently replaced the radiator and did not realize until it was already installed that there is no port at all for the switch.

My 89 780 has an unused port on the back side of the radiator, so the radiators do appear to come in both styles -- side mount or back mount switch.

My solution for now was to install a toggle switch on the dash to operate the fan. I just have to pay attention to the coolant temperature. The correct radiator was three times the cost of the one I purchased and believed was the correct one for Regina ignition.

I would concur with you that the switch itself is defective if indeed the fan is always on. One wire is constant hot 12 volts, so touching the two together will engage the fan.

If you are purchasing a new switch, be sure to purchase the correct one. They come in both slip fit and threaded styles.

I'm curious, is your dash toggle switch to manually operate the fan all that necessary? You say you need to pay attention to the coolant temperature using the switch. If you were to just leave the switch on at all times so it's always operating, how different would things be watching the temp needle on the dash? As long as your dash bypass switch is hooked up using the old leads on the rad thermal switch, the fan relay will only be energized when the ignition switch is ON (KP-II).

My guess is it will take slightly longer to reach normal operating temp, but otherwise you can just leave it on and let the cooling system thermostat do its job looking after maintaining normal operating temp. This wouldn't be much different than having a full time belt driven fan (as the 1989 240s had) other than the fan would run in KP-II whether or not the engine was turning over, so could run the battery down. Also, in the later 940s without the thermal rad switch, the fan was on full time whenever the A/C was turned on and running (using a pressostat on the condensor as the A/C sensing switch) in addition the ECU being able to operate the fan on demand whenever the block temp sensor said the engine was approaching hot. Electric fans were primarily introduced to slightly improve overall gas mileage, slightly lessening the load on the engine, and the rad switch was there to cycle the fan, lessening the demand on the battery and alternator as well as reaching normal operating temp sooner for slightly improved gas mileage.
--
Dave -still with 940's, prev 740/240/140/120 You'd think I'd have learned by now

I saw a through the radiator fan switch on an older BMW at a recent car show. I had actually forgotten about that until you mentioned it. It's probably the better way for me to go.

Relieving the stress on the battery/alternator as well as prolonging the life of the fan are the main reasons for the toggle switch. I have found that the fan is not all that necessary unless I am sitting in traffic for a long time or on hot days at the traffic light to keep the A/C cool. I only activate the toggle switch when I'm stopped. Otherwise, it's off.
--
94 940 n/a 330k+ and rolling, owned 20+ yrs; 93 940; 89 780 turbo; 05 XC90; 07 S60 T5; sold 92 745T; sold 88 764T