Electric Fan Not Turning On

Recently I needed to replace the thermostat in my '91. The top bolt in the thermostat housing was seized and I was afraid it would break off in the head, so I took it to a shop. They were able to heat the bolt and get it out without breaking it, but apparently they didn't take the coolant temp sensor out of the thermostat housing before applying the heat. When I drove the Jeep home I got a check engine light (code 22 - coolant temp sensor out of range) and the electric fan was running constantly. It turns out they fried the coolant temp sensor and it was leaking water into the connection. I replaced the sensor.

I assumed I had fixed the problem until I took it for a drive today and noticed that the electric fan is not kicking on when it should. The fan turns on when I turn the A/C on or if I unplug the coolant temp sensor, but if I let it idle the engine temp will come up and the fan will not come on.

This seems to point to the sensor being bad, but it is brand new. Anything else that could be the culprit? Is there a way I can more definitively test the sensor? Thanks.

Update: I found the following in the FSM

Coolant Temperature Sensor
1) Disconnect coolant temperature sensor. Using a DVOM, check
sensor resistance. Resistance should be as specified in the
COOLANT TEMP SENSOR & MANIFOLD AIR TEMP SENSOR RESISTANCE table.
Replace sensor if necessary.
2) Check continuity between PCM wiring harness terminal No. 2
and sensor connector terminal. Also check continuity between PCM
wiring harness terminal No. 4 and sensor connector terminal. Repair
wiring if an open circuit is indicated.
NOTE: Also see appropriate coolant temperature sensor tests in
G - 2.5L & 4.0L TESTS W/ CODES article in this section.
COOLANT TEMP SENSOR & MANIFOLD AIR TEMP SENSOR RESISTANCE TABLE
􀀀
􀀀
􀀀
􀀀
􀀀
􀀀
􀀀
􀀀
􀀀
􀀀
􀀀
􀀀
􀀀
􀀀
􀀀
􀀀
􀀀
􀀀
􀀀
􀀀
􀀀
􀀀
􀀀
􀀀
􀀀
􀀀
􀀀
􀀀
􀀀
􀀀
􀀀
􀀀
􀀀
􀀀
􀀀
􀀀
􀀀
􀀀
􀀀
􀀀
􀀀
􀀀
􀀀
􀀀
􀀀
􀀀
􀀀
􀀀
􀀀
􀀀
􀀀
􀀀
􀀀
􀀀
􀀀
Temperature  F (  C) Minimum Ohms Maximum Ohms
-40 (-40) .............. 291,490 .......... 381,710
-4 (-20) ............... 85,850 ........... 108,390
14 (-10) ............... 49,250 ............ 61,430
32 (0) ................. 29,330 ............ 35,990
50 (10) ................ 17,990 ............ 21,810
68 (20) ................ 11,370 ............ 13,610
77 (25) ................. 9120 ............. 10,880
86 (30) ................. 7370 ............... 8750
104 (40) ................ 4900 ............... 5750
122 (50) ................ 3330 ............... 3880
140 (60) ................ 2310 ............... 2670
158 (70) ................ 1630 ............... 1870
176 (80) ................ 1170 ............... 1340
194 (90) ................. 860 ................ 970
212 (100) ................ 640 ................ 720
230 (110) ................ 480 ................ 540
248 (120) ................ 370 ................ 410

I tested the resistance across the pins of the coolant temp sensor as the engine came up to temperature. The resistance seems low across the temperature range. At 210-ish I measured about 980 to 990 ohms. I'm not sure if this indicates a problem. I am now going to try to figure out step 2, the continuity test. Any input is appreciated.

sensors can be off by +/- 15 degrees... I changed out several different sensors and gauge senders until I found two that agreed with each other. I had to use a friends scanner to read the computer temperature.

Well, it's dark and I wasn't able to perform the second test. Can someone explain what is meant by 'PCM wiring harness terminal'? I figured there would be a plug on the outside of the PCM that I could disconnect, but nothing looked obvious. A picture might be especially helpful. Thanks.

ehall: my dash gauge will get to the 3/4 mark without the fan ever kicking on. Are you saying everything may be okay?

That is the harness that carries the wires to the Powertrain Control Module (PCM). The terminal part is the pin number on the connector at the PCM. Looking at the FSM for my '00 shows the wire from the sensor going to C1/P16 which is the black connector on the PCM. Not sure what color or pin it will be on your '91. There should be a connector though and it should be in the FSM.

It is possible that everything is okay, but the sensor may be off or that it doesn't want to play nicely with the gauge. Most use a temp gun to see exactly what the temp is before chasing things. You could return the sensor and try another sensor as a quick, cheap check.

Sounds like step is is a basic continuity check to be sure the wire isn't damaged/broken.



DMS #525
AMSA #1102
OBPA

Thanks for the suggestions guys. I think I can rule out the possibility that everything is okay because before I took it to the shop the electric fan would kick on about where I would expect it to on the gauge. Also, if I'm moving fast enough to get airflow, the gauge sits right at 210. So I think the gauge is reasonably accurate. So if the fan isn't kicking on when the gauge has reached 3/4 (that's as high as I have let it go), something is definitely wrong.

With a little daylight I was able to do the continuity test very easily. I got 0.6 ohms on one wire and 0.7 ohms on the other. For the heck of it I also tested resistance between the two wires and got an open circuit, as I would expect. So I seem to have every indication that the wiring is good.

Is there anything else I can check before concluding that the new sensor is bad? Thanks.

Edit: I also just rechecked the resistance of the coolant temp sensor. Even with the gauge about half way between 210 and the 3/4 mark I measured 750 ohms. At this point I'm feeling fairly confident that the sensor is out of spec. Unless anyone has another suggestion I'm going to exchange the sensor as soon as the jeep cools down. Thanks.

Still working on this. I exchanged the sensor, and the problem persists. In boiling tap water the second new sensor reads 750 ohms, while the original (melted one) reads 650. I'm beginning to think the NAPA sensors may not be made to the right spec. Is that possible?

On the bright side the I called the shop that melted the original sensor and they seem willing to help me diagnose this. They said they will do some research and to call them back in the morning.

I'm getting desperate. Any suggestions?

having the same problems......when it happened before i replaced the rear coolant temp sensor that was bad out of the box. now my fan will only sometimes kick on when I throw the defroster on, got a new fan though and going to try that.

Yeah when I swapped a 99 into my 91, I also took the opportunity to replace all the sensors and senders with new parts from Standard. Engine was always showing warm temperatures, like 210 on the highway and 220 at red light (the old engine showed 195 on highway and 210 at red light), while the computer sensor was also reporting oddball temps. I bought another Standard gauge sender, same thing. I picked up some senders and sensors from the junkyard and they were all over the place, gauge was lower but sensor was higher. Finally I bought a $7 sender from Autozone and it showed exactly the right temps, so I mated it with one of the sensors that was closest and called it done.

Remember:
the temp sensor does not control the fan. the PCM does, using data from the sensor, and also from the a/c control head.
When engine temp reaches a pre-determined temp or, when the a/c is requested (via a/c or defrost mode) the PCM turns on the fan relay.
a hundred or so ohms difference in sensors is not uncommon and not usually a problem.

Your best bet to troubleshoot is to use a scan tool to see what the PCM is interpreting the value of the sensor and see if it matches the gauge on the dash.
There could quite possibly be an air pocket in the head or a bad gauge temp sensor causing the gauge to misread the temp of the engine and its not getting as hot as it shows on the dash gauge. This is quite plausible as the PCM is obviously not wanting to turn on the fan.

Digger87xj said:

Remember:
the temp sensor does not control the fan. the PCM does, using data from the sensor, and also from the a/c control head.
When engine temp reaches a pre-determined temp or, when the a/c is requested (via a/c or defrost mode) the PCM turns on the fan relay.
a hundred or so ohms difference in sensors is not uncommon and not usually a problem.

Your best bet to troubleshoot is to use a scan tool to see what the PCM is interpreting the value of the sensor and see if it matches the gauge on the dash.
There could quite possibly be an air pocket in the head or a bad gauge temp sensor causing the gauge to misread the temp of the engine and its not getting as hot as it shows on the dash gauge. This is quite plausible as the PCM is obviously not wanting to turn on the fan.

Click to expand...

Keeping in mind that the PCM controls the fan is a good point. Is it possible that the PCM was somehow damaged? Could I test that by putting a 400 or 500 ohm resister in place of the sensor and seeing if the fan kicks on? As far as a scan tool, I don't have one. I do think the dash gauge is accurate though because it almost always sits at 210. The only time it goes above that is extended idling in traffic when the electric fan should be working.

And although I'm only measuring a 100 ohm difference at 212 degrees, the NAPA one is out of the range given in the FSM. Maybe I'm being stubborn and stupid, but I'm thinking about trying a sensor from Kragen.

Im having the same problem on my 96, I hope you dont mind if I just pull up a chair.

check the gauge with an infrared gun pointed at the aluminum thermostat housing--you want to measure coolant temp, not engine temp

220 is about halfway between the 210 and 3/4 tick marks (the gauge readings are accelerated at higher temps)

The relay is the window into the circuit and the point where you will really tell what is going on. I'm looking at Alldata and have wire colors, which do no good looking into a relay socket. I do not have the specific pinouts for pins 85 ,86, 87 and 30 but if there's an 87A it is not involved in this circuit. First things first jump 87 to 30, the fan should come on, if not the problem is in that area. If it does look at 85 and 86. One of the two is supplied power by fuse 18 (in a 96) and the pcm will ground the other side when desired, this pulls the contact over and effectively shorts 87 to 30 firing the fan up.

JJacobs said:

The relay is the window into the circuit and the point where you will really tell what is going on. I'm looking at Alldata and have wire colors, which do no good looking into a relay socket. I do not have the specific pinouts for pins 85 ,86, 87 and 30 but if there's an 87A it is not involved in this circuit. First things first jump 87 to 30, the fan should come on, if not the problem is in that area. If it does look at 85 and 86. One of the two is supplied power by fuse 18 (in a 96) and the pcm will ground the other side when desired, this pulls the contact over and effectively shorts 87 to 30 firing the fan up.

Click to expand...

Okay, I am willing to try this, but help me understand how the relay works. Here is my understanding: Given at least one of two conditions, A/C on or engine temp high enough, the PCM decides that the fan needs to be on. The PCM then sends a signal to the relay telling the relay to turn the fan on. The relay does not know or care which condition was met in the PCM. So if my fan turns on when I turn the A/C on, doesn't that indicate that the relay is good? Have I gone wrong in my logic or understanding here? Thanks.

JJacobs said:

The relay is the window into the circuit and the point where you will really tell what is going on. I'm looking at Alldata and have wire colors, which do no good looking into a relay socket. I do not have the specific pinouts for pins 85 ,86, 87 and 30 but if there's an 87A it is not involved in this circuit. First things first jump 87 to 30, the fan should come on, if not the problem is in that area. If it does look at 85 and 86. One of the two is supplied power by fuse 18 (in a 96) and the pcm will ground the other side when desired, this pulls the contact over and effectively shorts 87 to 30 firing the fan up.

Click to expand...

I shorted 87 and 30 and the fan came on. 86 is +12V. 85 is at ground. I'm not really sure what this indicates though.

It means the power supply to the output side of the relay, the wiring and the fan itself are all fine. The relay itself should measure about 70 ohms across pins 85 and 86, and click and cause continuity between 87 and 30 when 85 and 86 get power and ground.

All the pcm is doing is outputting a ground to either 85 or 86 to turn the fan on, so you're left with either a logic fault, burned out driver or intermittent circuit between the pcm and relay output (again, 85 or 86) burned out drivers aren't really common since it is a low current output but anything is possible.

I have just under 75 ohms between 85 and 86 on the relay. I also switched the relay with the on for the started and the jeep still cranked. By 'logic fault' do you mean the PCM is not operating properly? And what is the 'driver' that could be burned out? For the moment I think I am going to focus on the sensor and the connector because that is where the heat was applied (un-seizing thermostat housing bolt) that caused this problem. Thank you for all the help so far.

banderso said:

The relay does not know or care which condition was met in the PCM. So if my fan turns on when I turn the A/C on, doesn't that indicate that the relay is good? Have I gone wrong in my logic or understanding here? Thanks.

Click to expand...

Correct. You have successfully proofed the electrical portion of the fan/relay system. You proved the PCM has the capability to turn on the fan with an A/C request.

Thanks for sticking with me, guys.

I put a 500 ohm resistor in place of the sensor to see if it would kick the fan on, but for some reason it made the jeep very hard to start. I did get it to start once and fan went on. There may have been a check engine light too, I didn't think to look. So it could have been doing the same thing as when you unplug the sensor from the connector (check engine light on, fan on), but it at least gives me hope that the NAPA sensors may be out of spec. I have a Kragen sensor sitting on the bench next to the old sensor. When they come to the same temperature I will test the resistances and see if that gives me any insight. I also have a new connector coming in at NAPA in two and a half hours, so that can be wired in if I need to.