Radiator Fan not coming on

V

VBNSX

Experienced Member
Joined
2 July 2003
Messages
1,436
Location
Virginia Beach
i was in the process of changing all the hoses (which was successful) although i ran into a small snag.

I used that mighty vac thing which excavates the air from the system and then auto fills the cooling system in the car when u open the valve and the following happened:

1. I forgot to put the car on an incline to the nose... i thought this would've been no issue but when I went for the bleed, there was still air (this time car slanted)

2. I used just under 4 gallons of Honda coolant... personally i thought it would've required more (5).

3. The big issue: The fan won't come on.

Now to note, i had to use the vac thing a few times... i left a drain valve open so the gauge dropped.. so i found that then did it again, etc.

My main concern is the fan not coming on.

Things I know:
1. I did pull off the ECT gauge sending unit... ie, gauge cluster one no longer works in favor of one that has readings.

2. I pulled the ECT temp unit plug and put a jumper to it, the fan came on. (verifying relay and fan functionality)

3. At this point i'm lead to believe that it's either the ECT unit is bad or by using the vac so many times I may have screwed up the thermostat.

All that comes down to the questions:

1. Do you think that I may have locked up the thermostat?

2. How often to ECT temp units really fail?

3. From the wiring diagrams, the ECT gauge sending unit has no impact on the relative 'fan switch'... at least from what I see.. maybe i'm missing something??

thanks for any help anyone may add.
 
How do you know how hot the temp sensor thinks the coolant is? Don't know how often they fail, but mine was out when I bought the car. I quickly noticed that the rev limiter was kicking in too early, and verified the sensor was bad.
 
Daedalus said:
How do you know how hot the temp sensor thinks the coolant is? Don't know how often they fail, but mine was out when I bought the car. I quickly noticed that the rev limiter was kicking in too early, and verified the sensor was bad.
Click to expand...


I replaced the oem ect gauge sending unit with the one that was supplied with the stewart warner ECT gauge.

From there I can still get a reading on the temperature
 
The amount of coolant you refilled appears correct according to the manual. It is possible that you have an air pocket in the radiator near the sensor which is throwing off the sensor's ability to accurately read the temperature. I would recommend rebleeding the system , using the procedure outlined in the manual. I experienced the identical problem when replacing the coolant in an old Ferrari. Once it was bleed properly, the fans kicked on at the right time.
 
Doc308 said:
The amount of coolant you refilled appears correct according to the manual. It is possible that you have an air pocket in the radiator near the sensor which is throwing off the sensor's ability to accurately read the temperature. I would recommend rebleeding the system , using the procedure outlined in the manual. I experienced the identical problem when replacing the coolant in an old Ferrari. Once it was bleed properly, the fans kicked on at the right time.
Click to expand...

The procedure in the manual suggests an empty system w/ refill. Should I just empty it all out or can I just open the bleeders in the suggested fashion?

Oh and as of right now the car is slanted, nose down and heater on at full blast.

ty much,
x
 
VBNSX said:
I replaced the oem ect gauge sending unit with the one that was supplied with the stewart warner ECT gauge.

From there I can still get a reading on the temperature
Click to expand...

Are you sure you mean the "gauge" sending unit? The gauge sending unit isn't the one that controls the fan. And if you did replace the one that controls the fan, my next question is whether it has the same temp-resistance characteristic of the OEM gauge.
 
Here's a dumb question. How long did you let the car idle for while waiting for the fan to turn on? You'd be surprised, sometimes it'll take up to 20min for the water temp to get high enough (220F i believe) to turn the fan on. Those radiators are pretty darn efficient :wink:

Also, a bad thermostat will not stop the fan from working. The fan switch is located on the engine side of the thermostat (as opposed to the radiator side).

Also, I've never seen a fan switch go bad before, so aside from blowing a fuse (which you already said you tested by jumping the fan switch connector), I'd say, you just simply didn't wait long enough for the fan to turn on :biggrin:
 
Daedalus said:
Are you sure you mean the "gauge" sending unit? The gauge sending unit isn't the one that controls the fan. And if you did replace the one that controls the fan, my next question is whether it has the same temp-resistance characteristic of the OEM gauge.
Click to expand...

yeah... it's the 'gauge' sending unit. the gauge sending unit is the 'single' nipple one w/ a sleeve type of boot for the connector.

The one that controls the fan is the ECT sensor that is a 2 pin, round connector.

Lastly, the gauge sending unit does 'not' carry the same characteristics (0-5V) as the oem one.

x
 
C-speed said:
Here's a dumb question. How long did you let the car idle for while waiting for the fan to turn on? You'd be surprised, sometimes it'll take up to 20min for the water temp to get high enough (220F i believe) to turn the fan on. Those radiators are pretty darn efficient :wink:

Also, a bad thermostat will not stop the fan from working. The fan switch is located on the engine side of the thermostat (as opposed to the radiator side).

Also, I've never seen a fan switch go bad before, so aside from blowing a fuse (which you already said you tested by jumping the fan switch connector), I'd say, you just simply didn't wait long enough for the fan to turn on :biggrin:
Click to expand...

At max i waited till 210F.

I wish i knew how to read the chart but:

Thermoswitch "on" temp (low) = 183 +/- 2 (84 +/- 1.2)
Thermoswitch "off" temp (low) = subtract 11 +/- 3 (6 +/-2) from the actual "on" temperature (low)

Thermoswitch "on" temp (high) = 194 +/- 2 (90 +/- .12)
Thermoswitch "off" temp (high) = subtract 11 +/- 3 (6 +/-2) from the actual "on" temperature (high)

page 3-4 in the manual

well technically i can 'read' it.. i just don't know what the hell it means. :tongue:

however I waited till 210 thinking that would be sufficient. Should I go higher?
 
update

Well I'm thinking it's the ECT switch. I tested the resistance with the car pushing 200F and it still shows 5k ohms.

According to the manual it should be within the range of 200-400. This means the bastard is stuck.

so either:

1. I have an air pocket 'at' the sensor
2. it's bad

It's a $30 part so I figure it's worth checking out. I have to wait a bit though since I'll be picking it up along w/ the Acura ball joint separator (yes I forked over the $$$ for it), and boot kits.

Although in the interim I'm more than happy to try out anything else people may offer.

thanks again,
x
 
Briank said:
Disregard the above it is completely wrong. C-speed shame on you, you should know better.

The sensor is a thermistor, that along with the fan control unit controls the cooling fan operation, when and what speed the fan runs. If your car is not over heating then its unlikely you have a problem and you more you mess around the better chance you with create a problem.

Have you opened the bleader at the thermostat housing ?
Click to expand...

:P At the risk of sounding like a complete idiot, I'm not sure I agree Briank.

I believe the fan temperature thermo switch, is simply a switch, and not of variable resistance type (like the one used for the temperature gauge and ecm).

I believe the fan switch circuit is completely independant. I'll have to verify this in the manual though, i'm about to run out the door.

You can smack me if I'm wrong :biggrin:
 
Briank,
Ok I will eat my words... in the service manual on page 22-122, the NSX thermo switch is indeed a variable resistance type. The test procedure is also listed there using coolant in a pot and heating it up.

In my own defense (hehe), I assumed the NSX system was the same as an Integra/Civic, etc they have on/off thermo switches for the cooling fan. But I guessed wrong. :tongue: (ie: 94-01 Integra Helms page 22-12 Thermo switch test procedure)

* My original post is edited


PS: You cuff me on the side of the head, or a punch in the arm will suffice :P
 
Last edited:
Briank said:
Disregard the above it is completely wrong. C-speed shame on you, you should know better.

The sensor is a thermistor, that along with the fan control unit controls the cooling fan operation, when and what speed the fan runs. If your car is not over heating then its unlikely you have a problem and you more you mess around the better chance you with create a problem.

Have you opened the bleader at the thermostat housing ?
Click to expand...

I have opened the bleeders at the pipe and the housing. I've actually done the bleed sequence 3 times (garage floor is a tad sticky now).

I hear ya on the overheating thing.... however at what temp is the fan 'supposed' to come on? As I mentioned earlier, I let the temp get to about 210F... at that point I didn't want to risk letting it get any hotter...

I followed the procedure on page 11-66 regarding the ECT sensor and my ohm reading is still sitting at 5k@almost 200F

thanks for the continuous input.. I like it when feedback keeps comin.
 
C-speed said:
Briank,
Ok I will eat my words... in the service manual on page 22-122, the NSX thermo switch is indeed a variable resistance type. The test procedure is also listed there using coolant in a pot and heating it up.

In my own defense (hehe), I assumed the NSX system was the same as an Integra/Civic, etc they have on/off thermo switches for the cooling fan. But I guessed wrong. :tongue: (ie: 94-01 Integra Helms page 22-12 Thermo switch test procedure)

* My original post is edited


PS: You cuff me on the side of the head, or a punch in the arm will suffice :P
Click to expand...

mine only goes to 22-91 :(

however on 23-122 there is the ECT sending unit for the gauge cluster.. that's the one I removed and put the Stewart Warner sending unit in place of.
 
VBNSX said:
mine only goes to 22-91 :(

however on 23-122 there is the ECT sending unit for the gauge cluster.. that's the one I removed and put the Stewart Warner sending unit in place of.
Click to expand...

Download another copy of the nsx pdf :P

The ECT sending unit for the gauge cluster is completely different than then thermo switch for the fan.
 
can anyone verify what this means?

Thermoswitch "on" temp (low) = 183 +/- 2 (84 +/- 1.2)
Thermoswitch "off" temp (low) = subtract 11 +/- 3 (6 +/-2) from the actual "on" temperature (low)

Thermoswitch "on" temp (high) = 194 +/- 2 (90 +/- .12)
Thermoswitch "off" temp (high) = subtract 11 +/- 3 (6 +/-2) from the actual "on" temperature (high)

page 3-4 in the manual

really would like to know the temperature at which I should say: "ok it's not comin on"
 
VBNSX said:
really would like to know the temperature at which I should say: "ok it's not comin on"
Click to expand...
226 degrees




It should come on low speed at 183 degrees, there is a great chart in the manual sec 8-page 22

page 896 of http://www.nsxprime.com/FAQ/Reference/1991_svcman/1991servman.pdf

Also see 23 -116, Page 1102

I would check the resistance of the sensor at 180 - 200 degrees and you should see a little more then 1K ohm. If not replace it.

Also see the note about bleeding on 23 - 122 page 1108.

This is not that complex of a system so what's the big deal VBNSX. Figure out whats wrong and fix it.
 
Last edited:
Briank said:
This is not that complex of a system so what's the big deal VBNSX. Figure out whats wrong and fix it.
Click to expand...

I concur sir. I guess I was just being overly thorough and perhaps asking too many questions.

However here is where I am at currently.

referring to page 174 (10-9) of the manual:
DSCN3887.JPG


I pulled:
DSCN3885.JPG


DSCN3886.JPG


and got this as a result:
DSCN3888.JPG


Obviously these are not the same product. The item on the right used a 23 to pull, where as the left a 19... and of course thread size differ.


product on the right is what I pulled, product on the right is what I received from Acura. The part number listed is

37870-PJ7-003 SENSOR ASSY., WATER TEMPERATURE (TW5) (MATSUSHITA) 1 1991 NSX 36.80 28.58 Qty


pulled and ordered from this picture:
(item 23)
untitled2.JPG


other item close in direction and size is 22 which is:
37760-PH7-003 SENSOR ASSY., RADIATOR FAN CONTROL

Although I guessed that since 21 was closer in bearing/position to 23 I figured that was it.

I am going to order the other part to see if that is what it is.

imo it seems odd if that is it but... if anyone else has issues like this I hope you don't run into the same 'dumbass' attacks I did.

Thank you for your patience and answering of these questions.
 
That part is over $100 if I remember right. Have you already verified it is bad using an ohmmeter, a thermometer and a pot of water on the stove?

Also, is your car a '91? The radiator thermistor was changed somewhere along the way. Same temp-resistance relationship, but different threads.

EDIT: Nevermind, the thermistors are the same for all years. I was looking at the wrong part #.
 
Last edited:

Similar threads

TANTO
COOLANT BLEED (my experience)
Replies
10
Views
369
Honcho
Honcho
TANTO
AC CHARGE PROCEDURE
Replies
1
Views
195
drew
drew
asylum
Oil pressure sensor / sending unit
Replies
8
Views
192
Old Guy
Old Guy
Chris_Lum
DIY: Koyo Radiator Installation | 1991-2005
Replies
0
Views
95
Chris_Lum
Chris_Lum
C
Mothra - JDM K20 Turbo/Quaife Dogbox build - Help please
Replies
2
Views
293
CatchinVibes
C
Back
Top