IAT Sensor - Where is yours?
Registered User
Joined: Mar 2011
Posts: 1,415
Likes: 2
From: Woodburn, OR
for example, with the stock iat sensor, i would make a pass at the drag strip and the temp would not start rising until half way into the run and would not start decreasing until i was on the return road.
Joined: Jun 2009
Posts: 180
Likes: 1
From: Dela-ware?
I haven't forgotten about this. Turns out that answer to this problem might have been staring us right in the face the whole time. If I have time tonight, I'll try to sit down and do a write up of my findings. But let me ask this question - what do you think would happen to the response time of the IAT if we were to strip away the plastic housing convering the actual thermistor of the sensor? 
Registered User
Joined: Feb 2011
Posts: 1,256
Likes: 2
From: Eastern NC
I can be cruising and see temps as high as 150...and doesn't matter what I do...they just stay there. I know they have to be changing somewhat...but once they get above 130 any movement is very slow.
Registered User
Joined: May 2011
Posts: 69
Likes: 0
After a 45min cruise I saw temps up to 172deg. It took a while to drop into the high 160's after I shut the car off and opened the hood.
I'm definitely going with the GM style sensor on the charge pipe just to prevent heat soak.
I'm definitely going with the GM style sensor on the charge pipe just to prevent heat soak.
Registered User
Joined: Feb 2011
Posts: 1,256
Likes: 2
From: Eastern NC
Originally Posted by Legion
It's one of those "you can't have your cake and eat it too" type of situations. Thermistor response times are rated by their time constant. The smaller the thermistor, the smaller the time constant. Also, any barrier between the thermistor and the air is going to slow the response time even further.
It's a simple concept really - you want the air as close to the center of the thermistor as possible to achieve fast readings. The toss up is that the thermistor is very sensitive, and exposing it leaves it vulnerable to handling damage.
So the Honda IATs not only use large epoxy filled thermistors, but they have an (insulating!) layer of plastic over them that further kills response.
It's a simple concept really - you want the air as close to the center of the thermistor as possible to achieve fast readings. The toss up is that the thermistor is very sensitive, and exposing it leaves it vulnerable to handling damage.
So the Honda IATs not only use large epoxy filled thermistors, but they have an (insulating!) layer of plastic over them that further kills response.
I haven't forgotten about this. Turns out that answer to this problem might have been staring us right in the face the whole time. If I have time tonight, I'll try to sit down and do a write up of my findings. But let me ask this question - what do you think would happen to the response time of the IAT if we were to strip away the plastic housing covering the actual thermistor of the sensor?
Registered User
Joined: May 2011
Posts: 69
Likes: 0
Does anyone have a link to a write or can someone post the part number(s) for the GM sensor and how to hard-wire it into the stock harness?
This would help anyone in the s2k community and possibly bring more traffic onto the site which alway$ help$!
Thanks!
This would help anyone in the s2k community and possibly bring more traffic onto the site which alway$ help$!
Thanks!
Joined: Jan 2006
Posts: 20,788
Likes: 8
From: location, location
Originally Posted by Legion
It's one of those "you can't have your cake and eat it too" type of situations. Thermistor response times are rated by their time constant. The smaller the thermistor, the smaller the time constant. Also, any barrier between the thermistor and the air is going to slow the response time even further.
It's a simple concept really - you want the air as close to the center of the thermistor as possible to achieve fast readings. The toss up is that the thermistor is very sensitive, and exposing it leaves it vulnerable to handling damage.
So the Honda IATs not only use large epoxy filled thermistors, but they have an (insulating!) layer of plastic over them that further kills response.
It's a simple concept really - you want the air as close to the center of the thermistor as possible to achieve fast readings. The toss up is that the thermistor is very sensitive, and exposing it leaves it vulnerable to handling damage.
So the Honda IATs not only use large epoxy filled thermistors, but they have an (insulating!) layer of plastic over them that further kills response.
I haven't forgotten about this. Turns out that answer to this problem might have been staring us right in the face the whole time. If I have time tonight, I'll try to sit down and do a write up of my findings. But let me ask this question - what do you think would happen to the response time of the IAT if we were to strip away the plastic housing covering the actual thermistor of the sensor?
As far as tuning its only the response rate that should be an issue as x degree F can be adjusted for as far as IAT corrections.
Thread Starter
Joined: Sep 2004
Posts: 4,894
Likes: 9
From: Tampa, FL
Plastic is a good insulator, so possibly the temps may drop and effect the response rate somewhat. But i imagine with the plastic stripped away it will just get dirty (oiled) and gradually give poor inaccurate readings.
As far as tuning its only the response rate that should be an issue as x degree F can be adjusted for as far as IAT corrections.
As far as tuning its only the response rate that should be an issue as x degree F can be adjusted for as far as IAT corrections.
Bingo.
Thread Starter
Joined: Sep 2004
Posts: 4,894
Likes: 9
From: Tampa, FL
Disclaimer: I am only referring to KPro for the terms of this discussion. Other EMS's (Haltech, ProEFI, AEM, etc) may have an IAT conversion/scalar table that you can use. KPro currently doesn't (or any piggyback based on the stock ECU).
Here's where we're at. Below are the three common options you can choose from if you're looking to improve the response time of the IAT sensor (in addition to moving it out of the intake manifold for AP1's)
The T1 and GM sensors splice into the factory IAT wiring and are faster than the stock IAT, out of the box. But there is one issue - KPro is stock ECU based and currently does not have a conversion table for a different IAT sensor. If you use the T1 or GM sensor, your IAT readings will be off. How off you ask?
That's pretty significant, but doesn't really tell the entire picture. Let's look at a full conversion table in * F:
That's about a 10-14* difference almost across the board that the GM sensor reads less than the actual temperature. I don't know about you, but I don't like that at all. There are a ton of temp compensation tables that depend on the IAT for fueling and ignition, and I don't like the idea of throwing them off. Now, I've been told that its all relative and as long as the car is 'tuned' correctly, the improper temperatures shouldn't matter. I don't know if that's true or not, but I don't like the idea of me or my tuner having to "guess" what the actual temps are. I'd rather the numbers to be correct from the get go and not have to worry about it.
Soooo that leaves us with the slow, stock IAT. If you've been paying attention, the first picture in the post shows an OEM sensor that's been modified. Let's take a closer look. Rather than regurgitate, I'll just quote it:
This is exactly how I plan to modify my IAT and install it into the charge piping. For AP1's, this is relatively easy to do. Just have a fabricator create a two bolt flange like the one on the IM and weld it on. Then, using that same flange, create a block off plate that can be installed over the stock IAT port location. For you AP2 guys, I believe the IAT's stock location moved to the intake and it does not use bolt type flange. Here's an option for you:
Installing the OEM sensor into the manifold:
As you know, the oem sensor is not threaded so I had to get a little creative in order to install it into my jacson racing manifold. I went to Menards and bought a brass 3/8 to 1/4 NPT bushing, and also a tube of epoxy.
Next, I simply roughed up the surface of the plastic a bit with my knife, or sandpaper would work well. This is to help the epoxy get a good grip - you don't want you boost pressure to blow the sensor out or anything (it only has to hold 1.1 lbs force at 10psi of boost, so nothing to worry about). Mix up the epoxy and coat the inside of the bushing and the outside of the sensor, put them together, and let cure.
The finished product. Note that the sensor element is 'rough' looking because I sanded it slightly - to increase the surface area in hopes that it might promote more heat transfer.
There you have it. All you need is a 1/4 female weld on bung like this: http://www.prostreetonline.com/pnsku/rhp-867-04-0.asp welded onto your charge pipe and you should be able to relocate the sensor. For a cleaner install, you may be able to use a 3/4 to 1/8 NTP fitting (should be tighter around the sensor).
All information pulled from ClubRSX via this thread: http://forums.clubrsx.com/showthread.php?t=697367. Lots of graphs, tables and logs and a ton of good tech discussion to digest in there. I pulled all the relevant bits out, but if you'd like to reference everything, there's the link.
When I complete this, I'll be sure to post up results. They will most likely be in my upcoming build thread.
Here's where we're at. Below are the three common options you can choose from if you're looking to improve the response time of the IAT sensor (in addition to moving it out of the intake manifold for AP1's)
Code:
GM Actual Diff 190.4 194 3.6 186.8 190.4 3.6 183.2 186.8 3.6 176 183.2 7.2 172.4 179.6 7.2 168.8 176 7.2 165.2 172.4 7.2 161.6 168.8 7.2 158 165.2 7.2 154.4 161.6 7.2 150.8 158 7.2 147.2 154.4 7.2 143.6 150.8 7.2 136.4 147.2 10.8 132.8 143.6 10.8 129.2 140 10.8 125.6 136.4 10.8 122 132.8 10.8 118.4 129.2 10.8 114.8 125.6 10.8 111.2 122 10.8 107.6 118.4 10.8 104 114.8 10.8 100.4 111.2 10.8 96.8 107.6 10.8 93.2 104 10.8 89.6 100.4 10.8 86 96.8 10.8 82.4 93.2 10.8 78.8 89.6 10.8 75.2 86 10.8 71.6 82.4 10.8 64.4 78.8 14.4 60.8 75.2 14.4 57.2 71.6 14.4 53.6 68 14.4 50 64.4 14.4 46.4 60.8 14.4 42.8 57.2 14.4 39.2 53.6 14.4 35.6 50 14.4 32 46.4 14.4
Soooo that leaves us with the slow, stock IAT. If you've been paying attention, the first picture in the post shows an OEM sensor that's been modified. Let's take a closer look. Rather than regurgitate, I'll just quote it:
So, someone a while back said they stripped the plastic off of their oem IAT sensor. I have done the same, tested it against both the GM sensor and oem sensor for response time, and installed it in my blower manifold.
Stripping the sensor:
I simply used a razor blade (not the pocket knife in the pic) to carefully shave away the plastic. What you will find is a ball of yellowish epoxy which encases the RTD element itself. I did not shave any of the epoxy off of my sensor. If you could find away to get it off too, however, you would really have a fast sensor!! I didn't want to risk breaking it because it is my only one, and I need to be able to drive my car...
Why is this better?
It is better because it now has a faster response time. In fact, it is faster than the GM sensor by quite a bit, and it will read the actual temperatures rather than ones that are "close." I tested them by holding each sensor in front of a heat gun for a minute or so while datalogging with Kpro. The temperature response is shown below.
Time constants:
The time constant of of sensor tells you how fast it will respond to a given temperature input. It is the time it takes for the value to change to 63% of the final value that it will reach.
OEM - 11.5 seconds
GM - 7.25 seconds
OEM altered - 4.375 seconds
As you can see, it is definitely fast responding. The oem-altered plot is with the amber epoxy still covering the element, would be considerably faster once that is off.
Stripping the sensor:
I simply used a razor blade (not the pocket knife in the pic) to carefully shave away the plastic. What you will find is a ball of yellowish epoxy which encases the RTD element itself. I did not shave any of the epoxy off of my sensor. If you could find away to get it off too, however, you would really have a fast sensor!! I didn't want to risk breaking it because it is my only one, and I need to be able to drive my car...
Why is this better?
It is better because it now has a faster response time. In fact, it is faster than the GM sensor by quite a bit, and it will read the actual temperatures rather than ones that are "close." I tested them by holding each sensor in front of a heat gun for a minute or so while datalogging with Kpro. The temperature response is shown below.
Time constants:
The time constant of of sensor tells you how fast it will respond to a given temperature input. It is the time it takes for the value to change to 63% of the final value that it will reach.
OEM - 11.5 seconds
GM - 7.25 seconds
OEM altered - 4.375 seconds
As you can see, it is definitely fast responding. The oem-altered plot is with the amber epoxy still covering the element, would be considerably faster once that is off.
Installing the OEM sensor into the manifold:
As you know, the oem sensor is not threaded so I had to get a little creative in order to install it into my jacson racing manifold. I went to Menards and bought a brass 3/8 to 1/4 NPT bushing, and also a tube of epoxy.
Next, I simply roughed up the surface of the plastic a bit with my knife, or sandpaper would work well. This is to help the epoxy get a good grip - you don't want you boost pressure to blow the sensor out or anything (it only has to hold 1.1 lbs force at 10psi of boost, so nothing to worry about). Mix up the epoxy and coat the inside of the bushing and the outside of the sensor, put them together, and let cure.
The finished product. Note that the sensor element is 'rough' looking because I sanded it slightly - to increase the surface area in hopes that it might promote more heat transfer.
All information pulled from ClubRSX via this thread: http://forums.clubrsx.com/showthread.php?t=697367. Lots of graphs, tables and logs and a ton of good tech discussion to digest in there. I pulled all the relevant bits out, but if you'd like to reference everything, there's the link.
When I complete this, I'll be sure to post up results. They will most likely be in my upcoming build thread.
Registered User
Joined: May 2011
Posts: 99
Likes: 0
You could also tap the intercooler piping and just cut the tabs on the sensor for the ap2. Then just thread the sensor its self. That is what I plan to do to my ap2. As I did this on my rsx and have had no issues.