Drive Accord Honda Forums banner

1999 Accord V6 3.0 EXL Overheating Issues

12K views 38 replies 13 participants last post by  CDsDontBurn 
#1 ·
Hello Drive Accord!,

I have a 1999 Accord V6 3.0 EXL that Ive owned since new, and it just started overheating recently.

I immediately noticed the fans were not coming on. I turned the AC on, and the fans came on. I removed the connector from the fan switch, jumped it with a paperclip, and with KOEO, the fans came on.

Ive done a ton of web searching, and looked to be the fan switch, so I purchased a new OEM fan switch and an OEM thermostat, and replaced both. Still overheating, and the fans are still not coming on.

Here is a summary of the troubleshooting things I have done so far in attempt to figure this out:

  • Fans work when fan switch is jumped.
  • Replaced the fan switch with a brand new Honda OEM switch., but fans still dont come on.
  • Fans DO come on when AC is turned on, but dont seem to reduce the temp.
  • The heater DOES blow hot air.
  • Ive already replaced the thermostat with a new OEM Honda thermostat
  • Ive tested the fan switch in boiling water. There is no continuity when cold, but once its in the boiling water for a few minutes, there is continuity.
  • Ive tested the Thermostat, and it opens in boiling water, closes slowly when I remove it.
  • I removed the radiator, and when I put a garden hose on the thermostat housing and run water through it, water comes out of the top hose port with what seems like a good flow.
  • Removed the thermostat completely and looks like there is some flow when looking in through the radiator cap opening. Cant say it looks like great flow, but the water seems to be moving.
  • I see no visible collapse or kinks in the two hoses.

Tomorrow I will be pressure testing the system after I reinstall the thermostat with a coolant pressure tester I got from the autoparts store loan a tool program.

Here are a few things I am suspect of.

It seems as soon as I start the car when cold, the temp gauge is immediately at 50%. Seems like this was only about 1/3rd before, and that was after it warmed up for a few minutes. Seems like it stayed at cold for the first few minutes prior to the problem, but I cant say for sure.

Could this be the Coolant Temp Sensor that is bad? Could that be 1) giving me a false reading of an overheated condition, and 2), could that prevent the fans from coming on? I am not seeing any boiling in the overflow tank, but the hoses are quite hot.

Thanks for the help in diagnosing this. Ive been suffering with this on-and-off for almost two weeks now.
 
See less See more
#2 ·
When you say overheating, do mean just the temp gauge, or is the car actually spilling coolant? When was the water pump replaced?...was it OEM?...I've heard the Gates ones had issues...

It sounds like you are doing a good diagnosis already and a pressure test would be my next step. Maybe it's a clogged/defective radiator...

I'm still kinda new to Hondas, but I've heard of these plastic radiators splitting at the seams causing overheating...

Good luck getting it sorted...
 
#8 ·
When you say overheating, do mean just the temp gauge, or is the car actually spilling coolant?
When I say overheating, Im only talking about the temp gauge. I dont see any real sign overheating in the overflow, etc.


When was the water pump replaced?...was it OEM?...I've heard the Gates ones had issues...
I almost certain the waterpump was changed around 110K miles (180K now), and also almost certain it was an OEM, but Im not completely sure.


It sounds like you are doing a good diagnosis already and a pressure test would be my next step. Maybe it's a clogged/defective radiator...

I'm still kinda new to Hondas, but I've heard of these plastic radiators splitting at the seams causing overheating...

Good luck getting it sorted...
Thanks!
 
#3 ·
Sounds like you may have a leak, but a pressure test will show if you have a leak. The other issue you probably have is an air bubble in the cooling system. If both fans come on when the A/C is on then you know the fuses, fan, and relays are good. Be sure the front of the condenser is clean and free of bugs, also check between the condenser and the radiator and make sure there is nothing blocking airflow.

Make sure the overflow tank is at the full mark when you shut off the engine, normally the car will pull in coolant from the overflow bottle when it cools.

It is possible you have a bad radiator, especially if it is original. When was the water pump replaced?
 
  • Like
Reactions: Zero1Niner
#5 ·
Good tips ITT. A bad coolant temp sensor should throw a CEL. It can also cause intermittent no-starts.

If the needle jumps to 50% shortly after a cold start, it sounds like a bad WP or rad. I'd think the fans would be blasting at that point, though, because the PCM would command them on in a [futile] attempt to compensate for a bad WP/rad.
 
#7 ·
Water pump
Radiator

I agree the pressure tester will reveal/confirm/deny your diagnosis.

Rare, but I have seen one or three "clogged" radiators over the years. Very old coolant, tap water with lots of minerals- they can start clogging passageways. You can't really clean it out despite what the bottle says - you have to replace the radiator.

When you did the garden hose thing, you said water came out of the top. Do you mean you shot water in through the radiator cap and water came out of the radiator's "inlet" hose opening (top radiator hose)? What about the bottom hose opening? What about the radiator's drain valve opening?
 
#6 ·
When my van had a head gasket leak (leading up to blown ones), the temp gauge would jump like that too. I know you already said the thermostat is fine but our Explorer's thermo got stuck closed and it did the same thing; replaced it and the radiator cap, and it was fine. Just for reference. I guess test the radiator cap too?
 
#11 ·
So thinking a little more about this, I have a few questions. These two engine codes

P0117 - Engine Coolant Temperature (ECT) Circuit Low Input
P0118 - Engine Coolant Temperature Circuit High Input

could be pointing toward a faulty ECT, right?

Could the faulty ECT be giving a false temp reading?

Could that be why Im not seeing the fans kick on?

I am smelling a more rich fuel burn that I previously remember when the car is running.

Could that also be related to the ECT?
 
#14 ·
Couple things to note about the temp gauge.. In the '99 there is a single wire temp sender going to the gauge and then there's the two-wire thermistor that feeds the PCM. Then you also have two fan switches and a fan control unit. 3 separate systems. I think in the 2000+ they deleted the single wire gauge feed and instead use the PCM input to show on the gauge.

Do you have a multimeter or a scanner? If you have a scanner it should be easy to read what the ECT is when cold and hot. After sitting for a long time, it should be close to ambient temp and also compare to the Intake temp sensor reading. If you have a multimeter ECT should be somewhere around 10k ohms. A warmed up engine should show a low 330 ohms and correspond to the 180f+- operating temp. The fan switches are hotter and click at something like 207f and 220f.
 
#16 ·
Thanks DonutLimo!,

I dont have a scan tool...just a simple code reader. I do have a multimeter. Seems like the scan tool option would be quite a bit more accurate. Wonder how I can get my hands on one to test this.

A few questions on the '3 systems'.
Which is the ECT? The single wire connector or the 2 wire thermistor?

Since I already have replaced the fan switch, seem reasonable that this one can be eliminated, right?

If the single wire is just sending a temp signal to the gauge only, is it reasonable to assume the sensor reporting to the PCM is the likely problem?

Im assuming that the ECT is part number 15 in the diagram (that is the sensor that has 2 wires), and the other is the temp sender(just a single wire)?

Do I have this right?
 

Attachments

#18 ·
Yeah....that should be lower; that's operating temp!

Order an OBD Bluetooth reader and Torque Pro for your phone... It's kinda fun to datalog and then graph it in Excel! LOL
 
  • Like
Reactions: Zero1Niner
#19 ·
Sorry for double posting, but...

Would these help at all?
 

Attachments

#20 ·
You have them right. My needle doesn't hit C until my scanner shows 130f so yeah there's definitely a short in that or the gauge is broken. The good thing is that system is totally separate and has no effect on performance or engine health. If you unplug the single wire the gauge should read cold. A VERY brief jumper to ground on the sender end of the wire should make the needle swing to H but don't let it peg or it could damage the gauge (if it's not already).

A multimeter resistance value from the ECT will give you an answer same as a scanner will but having the ability to see live info or graph/store while you're driving could be invaluable. The ECT directly feeds the PCM and if it is out of range (converts to voltage for PCM and has range of .5v to 4.5v) at any time it will set those codes and possibly cause running or starting problems. In fact the PCM may have already rejected it's values and is running without it since you have those codes stored. That could cause it to run a little less stoich before it starts using the O2 sensor for fuel control.
 
#23 ·
A multimeter resistance value from the ECT will give you an answer same as a scanner will but having the ability to see live info or graph/store while you're driving could be invaluable. The ECT directly feeds the PCM and if it is out of range (converts to voltage for PCM and has range of .5v to 4.5v) at any time it will set those codes and possibly cause running or starting problems. In fact the PCM may have already rejected it's values and is running without it since you have those codes stored. That could cause it to run a little less stoich before it starts using the O2 sensor for fuel control.
Hey DonutLimo,

I just took the cold test with the multimeter, and Im not very familiar with how to use a meter, so thought Id ask if this looks right before I start it up and see where it lands when its warm. Want to make sure I have it on the right setting, and if that looks like a good cold indication.

I just took two alligator clips and clipped each one on to one pole of the switch. The insulation keeps the clips from touching each other as far as I can tell.

Does this look right so far?
 

Attachments

#22 ·
I definitely found one problem today. Following DonutLimo's advice, I pulled the temp sensor connector off to see if the gauge read 0 (which it did), but the sensor probe broke off inside of the connector. I got it out, but clearly that part was toast.

Tonight Ill test the ECT.

Is it necessary (or advisable) to use Honda OEM sensor for this temp sender, or can that be a local autopart purchase?

One other question, the plastic connector has a shroud around it that is plastic, and the plastic has become brittle. It was breaking apart a bit when I pulled it off and when digging out that probe that was lodged inside. Seems that only the lip around the edge was brittle. The actual connector seemed ok. Should that connector be replaced? Its a very short wire to begin with, so changing that wont be too much fun, but if it will affect the reading, I guess Ill have to do it.

Perhaps I could cover it with heat shrink tubing?
 

Attachments

#24 ·
Overkill.. Just replace the switch and see if your problem magically disappears.. You have to replace it anyhow.

After market parts can be a hit / miss, but I have had good luck with most of them on my Hondas.
 
#27 ·
Of course that temp sending sensor will be replaced. Not messing with that at all. Just want to know if I need to replace the ECT. Its a $40+ part that Id rather not change if its working OK already. Ive already spent a chunk on stuff I probably didnt need to change after all.

Turns out that autozone doesnt have the sending sensor anyway, so I guess it will be Honda after all.
 
#25 ·
Well i guess you found your original problem. I don't know what the oem sending unit costs vs aftermarket but it could be a potential savings area since it wont affect anything other than the gauge reading. I wouldn't bother with the plastic parts, like you said you can always put heat tape around it as a shroud to keep anything from grounding it.

You're at 1670 ohms which is reasonable- all depends on garage temp. You could sketch an exponential graph based on that table I posted and get an idea where you fall according to your garage temp. You can also check the ECT ohms against the IAT ohms (intake manifold- an almost identical thermistor) before you run the car which should give you about the same resistance. The only thing that you cant really check is for an intermittent problem while driving. If it passes the garage tests, clean the connections and see if it sets the code again next time you're on the road.
 
#26 ·
You're at 1670 ohms which is reasonable- all depends on garage temp. You could sketch an exponential graph based on that table I posted and get an idea where you fall according to your garage temp. You can also check the ECT ohms against the IAT ohms (intake manifold- an almost identical thermistor) before you run the car which should give you about the same resistance. The only thing that you cant really check is for an intermittent problem while driving. If it passes the garage tests, clean the connections and see if it sets the code again next time you're on the road.
Based on the chart, im roughly extrapolating that the garage was around 86*, which is right in the middle of the 68 & 104, so that would mean the number at that temp would be right around 2,000ohms, which seems like im pretty much in that ballpark.

Ill warm it up now and see how much it drops.
 
#29 ·
That makes no sense, you can't have negative resistance. Bring the range back to 2k on the meter and try it again. Before you connect the leads to the ECT check that the meter shows 0 when leads touched together so you know your connections are good.
 
#30 ·
OK. Something was wacky on the 20K setting. When I touched the leads together, it started jumping all around and showing negative numbers again.

So I just tried it again at 2K, and when I touched them together, it was a solid .002

The cold engine was reading 1.129. after starting the car about 1 minute later, it was reading .958. When I could start to feel the bottom hose warming, it was .285, and then it dropped to .250, then as low as .185 then I turned the car off.
 
#31 ·
1. When the car is warm-hot and the temp gauge is going up do this, turn the heater on to high and see if the gauge drops to normal.

2. when engine is cold start the car with radiator cap off, wait and see looking at the coolant for air bubbles. If you see a lot of air bubbles and they dont go away after a while then you have pressure-air in the system. I had seen another honda with overheating issue and it was a head gasket leak causing a lot of pressure in the system.
 
#32 ·
Hey guys,

The car is fixed!! Today I replaced the temp sender, and what I realized was the car was never getting up to temp, and thats why the fans were not kicking on. With that replaced, the fans came on periodically like they always did, and the temp gauge stays just over the 1/3rd position like it always did.

Only thing left to do now is drain the system one more time, and refill with honda coolant. Think I might do one quart of the flush just for good measure since Ive gone through all of this.

In a way, Im kinda glad to have gone through this deep of a dive. Everything else proves the car is in pretty good shape.

One other thing is I really hope this ends up being useful to the next guy that goes through this.

This is a really great website, with some super smart and helpful people. Special thanks to everyone that chimed in on all of this, especially DonutLimo. You guys are awesome!!
 
#35 ·
In thinking about this, early in the diagnosis, I mistook the ECT for the fan switch. I jumped that with the paperclip, and the fans did not come on obviously. Perhaps that was what tripped those 2 codes.

I did clean the connectors and reset the CEL with my code reader. I drove it for 35 miles yesterday, and everything was perfect. No codes, temp needle stayed pegged at 1/3 ish, and fans were kicking on an off as they normally did.

Amazing how that one issue lead to so many hours of figuring things out.
 
#37 ·
Hello Drive Accord!, I have a 1999 Accord V6 3.0 EXL that Ive owned since new, and it just started overheating recently. I immediately noticed the fans were not coming on. I turned the AC on, and the fans came on. I removed the connector from the fan switch, jumped it with a paperclip, and with KOEO, the fans came on. Ive done a ton of web searching, and looked to be the fan switch, so I purchased a new OEM fan switch and an OEM thermostat, and replaced both. Still overheating, and the fans are still not coming on. Here is a summary of the troubleshooting things I have done so far in attempt to figure this out:
  • Fans work when fan switch is jumped.
  • Replaced the fan switch with a brand new Honda OEM switch., but fans still dont come on.
  • Fans DO come on when AC is turned on, but dont seem to reduce the temp.
  • The heater DOES blow hot air.
  • Ive already replaced the thermostat with a new OEM Honda thermostat
  • Ive tested the fan switch in boiling water. There is no continuity when cold, but once its in the boiling water for a few minutes, there is continuity.
  • Ive tested the Thermostat, and it opens in boiling water, closes slowly when I remove it.
  • I removed the radiator, and when I put a garden hose on the thermostat housing and run water through it, water comes out of the top hose port with what seems like a good flow.
  • Removed the thermostat completely and looks like there is some flow when looking in through the radiator cap opening. Cant say it looks like great flow, but the water seems to be moving.
  • I see no visible collapse or kinks in the two hoses.
Tomorrow I will be pressure testing the system after I reinstall the thermostat with a coolant pressure tester I got from the autoparts store loan a tool program. Here are a few things I am suspect of. It seems as soon as I start the car when cold, the temp gauge is immediately at 50%. Seems like this was only about 1/3rd before, and that was after it warmed up for a few minutes. Seems like it stayed at cold for the first few minutes prior to the problem, but I cant say for sure. Could this be the Coolant Temp Sensor that is bad? Could that be 1) giving me a false reading of an overheated condition, and 2), could that prevent the fans from coming on? I am not seeing any boiling in the overflow tank, but the hoses are quite hot. Thanks for the help in diagnosing this. Ive been suffering with this on-and-off for almost two weeks now.
My problem is vividly identical to yours. I've replaced cooling fan switch along with the coolant temp sensor, not positive or rewarding results. So I'm waiting for the water pump to arrive, but stumbled across a question......there's no way the water is communicating to tell the fans "hey, engine hot!".So I'm thinking, the THERMOSTAT is the winning street. Thinking about my residential central air and heat unit, and it's thermostat along with its functions. So back to this honda, I believe one of the jobs of the thermostat is relay message to computer verifying temp, and request fans be turned on. Is this affirmative?
 
#39 ·
On most cars, a thermostat is nothing more than a spring loaded wax plug, where once the wax is hot enough, it melts allowing for the spring to open allowing coolant flow to the rest of the engine.

A bad thermostat is likely due to a failed spring not opening or not closing as needed.
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top