Honda / Acura K20a K24a Engine Forum banner

1 - 17 of 17 Posts

·
Registered
Joined
·
357 Posts
Discussion Starter · #1 ·
Posting here to hopefully get some new suggestions on what to try to get my recent swap reading proper temperature..

K20a (cl7 swap) in an eg hatch
Hybrid racing drivers side rad
oem thermostat
fan switch in bottom of rad
temp sensor by the head in the corner of the RBC
heater core in tact. Heat works great in the cabin.
temp gauge does not fluctuate. Gets to around 1/3rd and never goes higher.

I completed the swap a while ago, tried to bleed the coolant several times, still won't get up to operating temp. It gets to around 120F in kpro then starts to cool down.

Things I've done..
  • Idle for 15-30 mins with the cap on. Verified bottom hose is getting hot after about 5-10 mins of idling. Tstat is opening. Top hose is also hot as expected.
  • Tried to idle with the rad cap off, it just pushes coolant out of the rad. I shut it off before it made a huge mess.
  • Idle for 15-30 mins with a coolant bleeder funnel attached, coolant level at the highest point. I got a good amount of bubbles out of it doing this. Made the issue a bit better. (previously it was only hitting 80-90F max)
  • with bleeder funnel attached, held RPMs at 3k for a few minutes. Temp climbs a bit but settles down after a while. I hit 137 at my highest this way. Did this 5 or so times until it stopped going higher.
  • During all of this, the fan would come on every few minutes as expected.
Anybody have any tricks for bleeding in an eg? I've done several swaps in DC5's and 8th gen civics and never had a single problem with coolant. This is probably the last thing I have to work out before it's solid. Any suggestions are greatly appreciated.
 

·
Registered
Joined
·
192 Posts
What's the fault? What temp does your fan switch cycle at?

With the heater on full noise the coolant temp may not come up any higher.
 

·
Registered
Joined
·
133 Posts
It’s weird that your temp gauge shows 1/3, KPro shows the highest temp at 137, but your fan is turning on, heat blows hot, and lower hose is getting hot. I’m asking myself “how is the lower hose getting hot after just 5-10 minutes of idling?” Then what is the ECT reading in KPro when you turn the key on and the car has sat overnight?

Why not try disconnecting the fan switch and having KPro control the fan operation? Since you can change when the fan turns on, I believe it’s set at 180 and shouldn’t turn on any sooner than that.

The symptoms you have listed sound very unusual. An OEM thermostat wouldn’t open at such a low temp unless it was stuck open. One good pull on a cool day and you can see a double digit drop in coolant temperature, and that’s without the fan on. Even the aftermarket “low temperature” thermostats don’t open at such a low temperature.

Is the thermostat the one that came with the motor or is it new? Did you run any diagnostic test on the ECT sensor if it’s reading incorrectly before you turn the car on (before starting it in the morning is it reading 50 degrees on a 30 degree day)? Does the motor have the OEM upper coolant neck or one with the radiator cap on it? Did you raise the front end high up with the coolant funnel on it and bleed it that way? Did you fill up the heater core hoses with coolant first (with the heater control valve open)? No coolant leaks or loose hose clamps that let coolant seep out.

Just throwing ideas out there. Nothing in my post is meant to criticize or insult your intelligence.
 

·
Registered
Joined
·
357 Posts
Discussion Starter · #4 ·
thanks for the replies guys!

There are no error codes. Fan switch is OEM so I'm guessing it turns on when the coolant hits 190. This is my friends car that I swapped for him, I no longer have the car at my place but he did try bleeding the coolant again and got it up to 161 before it started dropping again, so we have some progress but still not where it needs to be.

The lower hose did not get hot for a while the first time. It took around 15-20 mins of idling and holding RPMs at 3k for the thermo to open initially. After it finally opened, it cycles as expected after 5-10 mins of idling on a warm day. Given the way the lower hose acts, the thermostat definitely closes and stays closed until it opens the first time. I held the hose and it was cold for a long time, then got warm pretty quickly. Thermo is the OEM unit that came with the swap, it wasn't changed. I did remove it before putting the motor in and verified it opens as it should when submerged in hot water. Upper coolant neck is OEM as well, no rad cap present there. We did jack the car up as high as we could get it while bleeding with the funnel. It didn't make any difference really. A few bubbles came out.

We did not pour any coolant directly into the heater core. The heater blows consistently hot though so I'm guessing there aren't any bubbles there.

I'll try disconnecting the fan switch and see if it gets hotter to help push some more air out. Also noticed the fan switch has a tiny leak that drips a few drops of coolant so I'll get that re-taped and then vac the system down and try letting the coolant suck in instead of pour it in at atmospheric pressure.

I have not done any diagnostics on the ECT sensor. I'm not sure how to test that actually. It would be a good thing to try. Any resources on that test?

On a cold start before firing, the ECT reads around 60-70F. Outside temps are around there, slightly lower, so that part seemed right..

I'm starting to think maybe the lower rad hose still has some air bubbles trapped. I'm using the Ktuned hose kit and the lower hose comes out of the thermo, goes down, back up, back down, then into the bottom of the rad. The up part looks like something that would hold air pretty easily. Hoping the vacuum injection helps that out..
 

·
Arouse the DAMPFHAMMER!
Joined
·
7,611 Posts
...there are known issues with the aftermarket t-stats
I confirm this too. About 2 guys here and two of my customers suffered issues with aftermarket thermostats. All happened in a 5 months time range back in time. They all show a inconsistent and too low big circuit ECT from 60-75 °C during cruising after the coolant system was vented several times, so that we can assume a proper cooling condition. After, at least my customers, turned back to OEM ECT was where it should be, around 78-86 °C.

One should keep in mind all ECT based fuel corrections tables are based on a "normal" ECT of 81+ °C. Below that fuel is added, tuning does not make sense in that area because it is correlated to ECT below 81 °C, you get fuel added. E.g. at a ETC of 65 °C about 6.5 % fuel will be added at WOT condition. If one would tune at 65 °C and run it in an other situation on 81+ °C, it would lean out by 6.5 % less fuel. It is an extrem example, but I just want to make clear, a proper engine has a proper ECT control. A not proper working thermostat can cause engine damage or at least reduce it's lifetime and increase cost, beside the bigger emission pollution.
 

·
Arouse the DAMPFHAMMER!
Joined
·
7,611 Posts
I'm starting to think maybe the lower rad hose still has some air bubbles trapped. I'm using the Ktuned hose kit and the lower hose comes out of the thermo, goes down, back up, back down, then into the bottom of the rad. The up part looks like something that would hold air pretty easily. Hoping the vacuum injection helps that out.
Did you confirm a proper leak down rate of the coolant system before you filled it up? Coolant leak down tests are done at around 0.4 bar + atmospheric pressure. Tools for that are easy to get on online shops or your local automotive tool shop, prices start at around 30 USD. The precious time one invest in sealing issues without validate it before is much more worth then the tool invest. One hour of failure investigation cost more then the tool itself, therefore I highly recommend doing this before the fill up. The test can be done also after the fill up, but it is less efficient, water doesn't leak that easy like air through the same gap size, only steam 😉.

BTW, changed the thread topic from "woes" to "hose". Please confirm if this is correct, I am no native speaker, but dictionary said "wose" word does not exist.
 

·
Registered
Joined
·
357 Posts
Discussion Starter · #8 ·
Are you using an oem t-stat or a Ktuned?

OEM

there are known issues with the aftermarket t-stats
I confirm this too. About 2 guys here and two of my customers suffered issues with aftermarket thermostats. All happened in a 5 months time range back in time. They all show a inconsistent and too low big circuit ECT from 60-75 °C during cruising after the coolant system was vented several times, so that we can assume a proper cooling condition. After, at least my customers, turned back to OEM ECT was where it should be, around 78-86 °C.

One should keep in mind all ECT based fuel corrections tables are based on a "normal" ECT of 81+ °C. Below that fuel is added, tuning does not make sense in that area because it is correlated to ECT below 81 °C, you get fuel added. E.g. at a ETC of 65 °C about 6.5 % fuel will be added at WOT condition. If one would tune at 65 °C and run it in an other situation on 81+ °C, it would lean out by 6.5 % less fuel. It is an extrem example, but I just want to make clear, a proper engine has a proper ECT control. A not proper working thermostat can cause engine damage or at least reduce it's lifetime and increase cost, beside the bigger emission pollution.
That's a good point. He's starting an e-tune at the moment, I'll have him hold off until the ECT is squared away.

Did you confirm a proper leak down rate of the coolant system before you filled it up? Coolant leak down tests are done at around 0.4 bar + atmospheric pressure. Tools for that are easy to get on online shops or your local automotive tool shop, prices start at around 30 USD. The precious time one invest in sealing issues without validate it before is much more worth then the tool invest. One hour of failure investigation cost more then the tool itself, therefore I highly recommend doing this before the fill up. The test can be done also after the fill up, but it is less efficient, water doesn't leak that easy like air through the same gap size, only steam 😉.

BTW, changed the thread topic from "woes" to "hose". Please confirm if this is correct, I am no native speaker, but dictionary said "wose" word does not exist.
Not before filling it up. It was drained yesterday, leak fixed on the fan switch, then it was placed under vacuum for 5 minutes and held vac the entire time. Still isn't getting up to the right temp.

How can one test the ECT sensor for proper operation?

I did mean to put "woes" in the title. define woes - Google Search

No big deal though :)
 

·
Registered
Joined
·
133 Posts
There is are 3 diagnostic flow charts in the RSX service manual for the ECT sensor. One is for range/performance, and the others cover low or high circuit voltage. Sometimes, a sensor or wiring fault (high, low, no resistance) can be bad enough to cause a problem but good enough not to cause a trouble code.

In KPro, you are pretty much able to check test range/performance. In the service manual it asks you to check the temperature reading with a scan tool after the car is fully warmed up (until the fan comes on then it is idling). If the reading is between 176-212F or sensor voltage is between (0.5-0.8v), check the thermostat and cooling system. If it isn’t in that range and the thermostat/cooling system checks out, replace the ECT sensor. You are able to check the coolant temperature in KPro so there is no need for a scan tool. It is a thermosistor, so if it is on the verge of failing it will have the wrong resistance values and begin to send the wrong day to the ecu despite everything else working as they should. But why would the cluster read a low temp as well, if it is using a single wire sending unit that isn’t schematically tied to the coolant temperature sensor?

Maybe the circuit is reading low voltage but this can be due to a failing coolant temp sensor rather than a fault in the wiring. When the voltage reads low, that means the temperature is high (car warmed up) and the resistance is low within the sensor. When it reads high, it is the complete opposite. So before you start the car it should be high voltage (5v if it’s coldish outside) and then drop as it’s warming up (0.5-0.8v) I can post the actual pages to all 3 flow charts.

You could also wait until the temperature drops enough (radiator cap isn’t hot or warm) and note the temp in KPro. Then remove the cap and use a thermometer to test the actual coolant itself. If there is a difference, I’m looking at bubbles in the coolant or a bad ECT. Bubbles in the coolant cause all kinds of issues andhaving hot air coming through the HVAC doesn’t necessarily mean that all the bubbles in a coolant system have been removed. Air bubbles literally act like plugs so wherever they are stuck they will cause issues for coolant-dependent sensors and components.

There has to be a reason why you have a fan turning on well below the operating temperature of an engine and the thermostat opening well before it should, temps dropping while idling. It is understood that temps will drop when the fan turns on but not by a dramatic amount. Once the fan turns off it takes a minute or two before temps get high enough to trigger the fan, assuming ambient temps aren’t very high or humidity is high). Maybe everything with the coolant is fine, no bubbles whatsoever. But the ECT sure isn’t showing information to reflect that.

When you tested the thermostat opening, I’d be more concerned with how and if it is closing properly, as in it isn’t getting stuck. Your issue is with temps not reading high enough rather than being too high.

With any leak, it is bound to introduce air back into the system. I would bleed the coolant again to get any air out, nose of the car as high as you can get it, with a coolant funnel on it. Remove the rad cap, open the heater control valve all the way (should be parallel to the heater hoses it is attached to), and install the funnel before you even turn it on (cold motor). Put some coolant into the funnel and once you are done getting the air out use a turkey baster or anything that can get the coolant out without making a mess. Massage the hoses while it is idling, particularly the heater hoses to force out any air bubbles in that part of the system. If your upper and lower hoses are soft enough, massage those as well, especially the lower when the thermostat opens.

This is just a reference to help and not meant to be the end-all solution.
 

·
Arouse the DAMPFHAMMER!
Joined
·
7,611 Posts
That's a good point. He's starting an e-tune at the moment, I'll have him hold off until the ECT is squared away.
The tuner should say this to him. If he did not the tuner has some greaaaaaaaaaaaaaaater potential to learn basic tuner stuff 😜.

How can one test the ECT sensor for proper operation?
Pull it, take a voltage source and test it on ice water and bubbling water. It's a thermistor technology, easy to measure the ohmic resistance. The back calculation to °C or °F I can do afterwards.


did mean to put "woes" in the title. define woes - Google Search
Nice word to know. Thanks for linking it, I really appreciate it.
 

·
Registered
Joined
·
357 Posts
Discussion Starter · #11 ·
There is are 3 diagnostic flow charts in the RSX service manual for the ECT sensor. One is for range/performance, and the others cover low or high circuit voltage. Sometimes, a sensor or wiring fault (high, low, no resistance) can be bad enough to cause a problem but good enough not to cause a trouble code.

In KPro, you are pretty much able to check test range/performance. In the service manual it asks you to check the temperature reading with a scan tool after the car is fully warmed up (until the fan comes on then it is idling). If the reading is between 176-212F or sensor voltage is between (0.5-0.8v), check the thermostat and cooling system. If it isn’t in that range and the thermostat/cooling system check out, replace the ECT sensor. Again, you’d be able to check the ECT voltage and the temp in KPro so there is no need for a scan tool. It is a thermosistor, so if it is on the verge of failing it will have the wrong resistance values and begin to send the wrong day to the ecu despite everything else working as they should. But why would the cluster read a low temp as well, if it is using a single wire sending unit that isn’t schematically tied to the coolant temperature sensor?

Maybe the circuit is reading low voltage but this can be due to a failing coolant temp sensor rather than a fault in the wiring. When the voltage reads low, that means the temperature is high (car warmed up) and the resistance is low within the sensor. When it reads high, it is the complete opposite. So before you start the car it should be high voltage (5v if it’s coldish outside) and then drop as it’s warming up (0.5-0.8v) I can post the actual pages to all 3 flow charts.

You could also wait until the temperature drops enough (radiator cap isn’t hot or warm) and note the temp in KPro. Then remove the cap and use a thermometer to test the actual coolant itself. If there is a difference, I’m looking at bubbles in the coolant or a bad ECT. Bubbles in the coolant cause all kinds of issues andhaving hot air coming through the HVAC doesn’t necessarily mean that all the bubbles in a coolant system have been removed. Air bubbles literally act like plugs so wherever they are stuck they will cause issues for coolant-dependent sensors and components.

There has to be a reason why you have a fan turning on well below the operating temperature of an engine and the thermostat opening well before it should, temps dropping while idling. It is understood that temps will drop when the fan turns on but not by a dramatic amount. Once the fan turns off it takes a minute or two before temps get high enough to trigger the fan, assuming ambient temps aren’t very high or humidity is high). Maybe everything with the coolant is fine, no bubbles whatsoever. But the ECT sure isn’t showing information to reflect that.

When you tested the thermostat opening, I’d be more concerned with how and if it is closing properly, as in it isn’t getting stuck. Your issue is with temps not reading high enough rather than being too high.

With any leak, it is bound to introduce air back into the system. I would bleed the coolant again to get any air out, nose of the car as high as you can get it, with a coolant funnel on it. Remove the rad cap, open the heater control valve all the way (should be parallel to the heater hoses it is attached to), and install the funnel before you even turn it on (cold motor). Put some coolant into the funnel and once you are done getting the air out use a turkey baster or anything that can get the coolant out without making a mess. Massage the hoses while it is idling, particularly the heater hoses to force out any air bubbles in that part of the system. If your upper and lower hoses are soft enough, massage those as well, especially the lower when the thermostat opens.

This is just a reference to help and not meant to be the end-all solution.
I appreciate that very thorough and detailed response. That gives me several things to check out before reporting back.

Since the system just got refilled with the last leak fixed, I'll take a look at the voltage output of the sensor. I wasn't aware kpro had the voltage, so that's good news. After that I'll try bleeding it some more with the funnel and see if anything changes. If not, I'm going to pull the thermostat again and do a more thorough test on it to see if that could be the problem.. When I first tested it, I only cycled it open once. I'll do it at least 5 times this time and take note of the temp each time to see if it's sticking in any way.
 

·
Registered
Joined
·
357 Posts
Discussion Starter · #12 ·
The tuner should say this to him. If he did not the tuner has some greaaaaaaaaaaaaaaater potential to learn basic tuner stuff 😜.

Pull it, take a voltage source and test it on ice water and bubbling water. It's a thermistor technology, easy to measure the ohmic resistance. The back calculation to °C or °F I can do afterwards.


Nice word to know. Thanks for linking it, I really appreciate it.
nice! I'll give that a try as well to test the sensor with known temperatures and report back my results.

The tuner hasn't received the first datalog yet, so I'm sure he would mention it. I had the same guy tune my k24a2 when i swapped that like 5 or so years ago.

Really appreciate the feedback everyone!
 

·
Registered
Joined
·
133 Posts
Let me correct myself.

KPro can measure the coolant temperature, but does not show the ECT voltage. I thought of another sensor that shows voltage on the fly, but ECT unfortunately is not one of them. Maybe I noticed it in the Hondata app, which I use at times when I’m driving if I don’t have a laptop hooked up. I only use it to check the ethanol content at gas stations after I pump. Either way, I apologize for that one.

I have to go for a ride in the car tomorrow. I’ll connect to the KPro and see if I find the ECT voltage there.
 

·
Registered
Joined
·
357 Posts
Discussion Starter · #14 ·
Let me correct myself.

KPro can measure the coolant temperature, but does not show the ECT voltage. I thought of another sensor that shows voltage on the fly, but ECT unfortunately is not one of them. Maybe I noticed it in the Hondata app, which I use at times when I’m driving if I don’t have a laptop hooked up. I only use it to check the ethanol content at gas stations after I pump. Either way, I apologize for that one.

I have to go for a ride in the car tomorrow. I’ll connect to the KPro and see if I find the ECT voltage there.
no worries. I've gotten really good at using my multimeter with this swap :D
 

·
Registered
Joined
·
357 Posts
Discussion Starter · #15 ·
Update! Got the issue resolved..

It's always the dumbest things that stump you the longest.. Kind of embarrassed it took so long to figure this out but it is what it is lol. Hopefully this helps someone in the future.

I had the IAT and ECT connectors swapped.. My ECT connector was connected to the IAT sensor, the IAT connector was connected to the ECT sensor. I didn't put 2 and 2 together until the tuner pointed out that the values looked to be switched at the end of a datalog.. Then it clicked.. 180F IAT and 120F ECT.. 180 is good operating temp, so he must be onto something.

With the engine running, I disconnected the IAT sensor. Immediately got a ECT high voltage code which confirmed my issue.

The IAT and ECT connectors are identical, so it can happen pretty easily. Once I swapped these the idle went down to 750 and everything is running as it should.
 
  • Like
Reactions: Bjorn

·
Arouse the DAMPFHAMMER!
Joined
·
7,611 Posts
With the engine running, I disconnected the IAT sensor. Immediately got a ECT high voltage code which confirmed my issue.
My nose wasn't that bad 😉. Anyway the wose is gone and the knows is back :D. Great :)! Thanks for replying the root cause.
 
1 - 17 of 17 Posts
Top