Different MAT sensors?

[JeepsAndGuns]

So now that I have my test harness set up, I decided to start testing a couple sensors from my jeep 4.0 so I would know what I can and can not use.
I pulled the jeep MAT from my engine and tested it against the gm MAT I bought new at the parts store. The gm sensor reads correctly on my datalog dash, and I used a heat gun to test it. The jeep MAT uses the same plug, but the ohm range is incorrect. When its hooked up, instead of reading 70* like the gm one did, it read closer to 50*.
So I am going to have to swap to a gm MAT. Only problem is, the gm one uses larger threads than the jeep one. The gm sensor has the same thread size/count as the CTS (I think 3/8 npt?). However the jeep mat has smaller 1/4-18 npt threads. Other than the thread size, they are almost identical.
So I was wondering, is there a gm MAT from some other car/truck that is compatable with a 7727/$8D that has 1/4-18 npt threads?
I suppose if there is not one, you might be able to re thread the gm sensor. Looking at them, the "cage" around the elements are the same size on both, just the gm one has more material around the main body for the larger threads. I think you might be able to chuck it in a lathe and turn it down a hair and re thread it. But only going down that route if there are no other compatable mat's.
Anyone know of any?

Heres a pic, jeep on the right, gm on the left.

[EagleMark]

I don't know what size it is but 1996 4.3L CPI has a MAT in intake like the ones you pictured.

Only others I know of are plastic and go in air intake, IAT. Which may be easier and more accurate if that is what your bin/mask/xdf used from factory. Temps in intake manifoild are differant/hotter then air in air cleaner or air tube to air cleaner. 1991-92-93 Caprice, Roadmaster has one in air filter and 1994-95-96 Caprice Roadmaster, Camaro etc.. LT1 engine has one in air intake tube.

But CTS from just about all GM EFI all read same as IAT and way cheaper, last FOREVER there, just a little slower to respond, it's what I've always used and can't feel a difference or see one in data?

[dave w]

Years and years ago, I used the Standard Motor Products Illustrated Parts Guide to figure out sensor interchangeability. The guide would show thread pitch and thermistor ranges. I poked around the online version of the SMP guide and it's not as easy as flipping pages in a book that has about 1200 pages.:mad1:

Anyway, the challenge I see is most newer / smaller sensors are going to be metric thread pitch.:mad1:

I'd try machining a CTS down to the correct O.D on a lathe, and try cutting the tread pitch I needed on the machined CTS. Maybe the CTS has enough metal to be machined and re-threaded?

dave w

[EagleMark]

I did shave down a three wire CTS to fit into an oil pan drain plug. Then someone else did it and ran out of metal? Look up the Oil Temp Sensor in drain plug thread...

If it fails just plug the hole and add an IAT...

[dyeager535]

From now on if I intend to try and rethread sensors, I'm going to have a junkyard spare to practice on first. As Mark mentioned, the three wire (brass) sensor in question was too thin ABOVE the threads. I tried the same on a two wire CTS, and it turned out to be steel with a brass core, and solid all the way through. I have another CTS, same application, and it is brass. But I had that one turned on a lathe and it didn't snap like the three wire. Just be careful.

If I wasn't worried about leakage/sealing (pipe thread) then I'd ONLY cut the sensor down where there are threads, no further up the body. Still won't guarantee it doesn't twist off, but the portion of the body where there are threads on the one that broke, was thicker than the portion above the threads.

[JeepsAndGuns]

I will be running $8D, witch is set up for a intake mounted MAT.
I remember the thread about the oil temp sensors. This one I only need the threaded part re threaded. I just wanted to know if there was any already avalable with the correct thread size before I have one machined. I will drop this one off at the machine shop and see what they can do. If it gets distroyed, then I'm only out $18 for another sensor.
I will only swap to a IAT if there is no other options, as the ecm/bin will be expecting a MAT, not a IAT. Like you said, different temps between the two.

[EagleMark]

Just chuck up the sensor end in drill mounted in vise and spin it, use a file to shave it, get close to diameter and thread it with die. I did that one in 20 minutes or so and never done it before.

[RobertISaar]

could recalibrate the conversion table to match the jeep sensor.

[EagleMark]

could recalibrate the conversion table to match the jeep sensor.Now, now, don't make it complicted! Were trying to bolt a part on here... :homer:

Never thought of that...

[dave w]

could recalibrate the conversion table to match the jeep sensor.
I hadn't thought of that either. :homer: I'm not sure I'd know how to recalibrate. Is there an 8D table to change? Maybe the hex code has to changed? Just wondering?

I tend to build / parts source an EFI conversion setup by GM Make / Model / Year.

dave w

[RobertISaar]

i may be a machinist, but if i can avoid potentially scrapping good parts, i avoid it.

as long as you could hold the two sensors in identical conditions(let's say throw them in a pot of water of something?), then just need to get the sensors around to a calibration point(there are 17), see what temperature the GM sensor is showing, then switch to the Jeep sensor and move the value at that portion of the table around until it matches what the GM sensor was showing. maybe switch back and forth again to confirm it's still the same.

obviously, this is done much easier with an emulator or at least a bank switch.

[Six_Shooter]

Why has no one suggested drilling the intake for the larger thread? That's what I do in a situation like this.

[EagleMark]

Probably to avoid removing the intake, it's a running vehicle... but yes that is a simple solution.

[Six_Shooter]

I've drilled intakes on vehicles for this, you just have to be careful, use rags, some grease, a vacuum and your head.

[JeepsAndGuns]

could recalibrate the conversion table to match the jeep sensor.

Wow, that can be done?
I would have no clue how to do it. Only problem I see, is finding a controlled environment to put the sensors in to get proper readings from them. Would the ADX also have to be changed?

I would also rather mod a $18 sensor than the intake on my jeep. So no drilling it. I might give what mark said a try, though I dont have any type of drill that will hold this thing.

[RobertISaar]

if the datastream pushes out a raw A/D value, then yes, it would need changed. if the normalized value comes out, then it's already accounted for in the PCM.

[EagleMark]

Wow, that can be done?
I would have no clue how to do it. Only problem I see, is finding a controlled environment to put the sensors in to get proper readings from them. Would the ADX also have to be changed?

I would also rather mod a $18 sensor than the intake on my jeep. So no drilling it. I might give what mark said a try, though I dont have any type of drill that will hold this thing.Yeah I did a coolant temp which was easy to chuck, just like a drill. A file makes quick work of taking off brass. NPT is an easy to die/tap, just file on a slight angle. Are those senosr ends to flimsy to chuck?

[phonedawgz]

You don't have to mount the MAT in the manifold. 3800, 3100s and many others have it in the snorkel. Fiero's have it in the air filter can (and GM still calls it a MAT sensor). Leave your stock MAT where it is and add your new one in your air filter housing.

[dyeager535]

The ends of the sensors apparently vary as well.

Three wire is a solid piece of brass, at least what I tried on.

Two wire CTS with steel body had a hollow "nose", which the chuck crushed.

I had the brass CTS turned on a lathe, so no idea what the nose is on it.

My *guess* is that if they turn the body from brass, there is no reason to remove any more metal than necessary at the tip. When they used a brass tube insert (steel body) they used the thinnest possible.

Only chucked two up like this, so if the money isn't really a concern, you could try it and see. As thin as the one I crushed was, light pressure with pliers would have been enough to know if it was solid or not.

[JeepsAndGuns]

Took the mat and chucked it up in my drill clamping onto the "cage" around the element, it held it pretty good. Worked on it for quite a long time with a file and got it turned down to where I thought it was enough to thread. Went to fastenal and picked up a die and started threading it. Got about halfway and couldnt go any more. I didnt turn it down enough near up next to the hex part. So I go to chuck it back up and the cage comes off. When I cut the threads, it went through the very tip and cut off the tip that was crimped to hold on the cage. So I cant hold this one any more with the drill. Guess I will keep it for my test harness, and go pick up another one and start again. Dont guess the cage is 100% required to use the sensor, if the new one breaks off too. It would be nice though if I could find someone with some type of temp control box or something simillar and just get the data to use the jeep sensor like robert suggested. Anyone know of a place or person that could do this?

[Six_Shooter]

Seriously, just drill the intake and thread it for the larger thread, you will save yourself lots of headache doing so.

[EagleMark]

It would be nice though if I could find someone with some type of temp control box or something simillar and just get the data to use the jeep sensor like robert suggested. Anyone know of a place or person that could do this? I've never tried that but I don't think it'd be as hard as you think... shop manual would probably have the info to test sensor, voltage, ohms whatever to 256 count... just thinking out loud...

[JeepsAndGuns]

I've never tried that but I don't think it'd be as hard as you think... shop manual would probably have the info to test sensor, voltage, ohms whatever to 256 count... just thinking out loud...

DOH! :homer:

Why didnt I think of that. Searched through my FSM and found a chart for it! The Jeep MAT and CTS have the same ohm readings. I think the gm MAT and CTS have the same readings as well dont they? With this info, can we make a conversion to what we need to change?
What exactly would we need to change, and where is it? Like davew said, is it in hex? If so thats gonna be so far over my head....

[RobertISaar]

a small patch would need to be made, since both the CTS and MAT use the same normalization table, but i can take care of that.

anyways, what would need changed is a 17 value table. it may or may not be present in your current XDF, but it can always be added in. the table will be referenced in A/D counts and the value you'll be changing will be in temperature, it's simple once you do it. i'll have to look into creating a basis for the changes later.

[EagleMark]

Looks like this is the place in $8D...


**************************************************

* Coolant var table #1, 4K PULL UP

*

* COOL = ((DEG C)+40) * (256/192)

**************************************************

ORG $F105

; TEMP DEG C A/D

; --------- ---

LF105: FCB 225 ; VERY HOT 0 151 1

FCB 215 ; 121 6

FCB 155 ; 76 22

FCB 132 ; 59 38

FCB 117 ; 48 54

FCB 106 ; 39 70

FCB 97 ; 32 86

FCB 88 ; 26 102

FCB 81 ; 21 118

FCB 74 ; 15 134

FCB 67 ; 10 150

FCB 60 ; 5 166

FCB 52 ; -1 182

FCB 44 ; -7 198

FCB 34 ; -14 214

FCB 22 ; -24 230

FCB 00 ; -40 246

************************************************** *

**************************************************

* Coolant var table #2, 384 OHM TBL

*

**************************************************

ORG $F116

; TEMP DEG C A/D

; -------------------------------

LF116: FCB 255 ; VERY HOT 0

FCB 255 ; VERY HOT 16

FCB 249 ; 147 32

FCB 223 ; 127 48

FCB 205 ; 114 64

FCB 191 ; 103 80

FCB 178 ; 94 96

FCB 167 ; 85 112

FCB 157 ; 78 128

FCB 147 ; 70 144

FCB 137 ; 63 160

FCB 127 ; 55 176

FCB 117 ; 47 192

FCB 104 ; 38 208

FCB 89 ; 27 224

FCB 67 ; 10 240

FCB 00 ; VERY COLD 256

**************************************************

[RobertISaar]

whoops..... i spread a little bit of false info there. the CTS and MAT do NOT use the same normalization table... the MAT has it's own dedicated table based on a 1K resistor, the coolant temp uses a 348 and 4K table. the oil temp sensor uses the 348 ohm table as well.

anyways, i'm doing some resistor calcs now to try and estimate the proper voltage drop that you will see based on that spec sheet for the jeep sensor.

[RobertISaar]

well, i'm done with the resistor calcs.... good news and bad news. good news is that it will work, bad news is that the sensor values for that specific sensor are crap for use as an IAT/MAT, but excellent for coolant temp... take a look at this table(assuming it posts correctly):









16




32




48




64




80

238



96

220



112

203.9



128

188.9



144

174.7



160

160.6



176

146



192

130.5



208

112.9



224

91.5



240

60



256

-40







on the left are A/D counts, on the right, the actual temperature that results from that A/D value in *F. notice how there are a lot of A/D counts at higher temps? that gives great resolution. from 112-176 counts, there are roughly 14* between breakpoints. there are 16 values between breakpoints, so each A/D count is a little under 1*F difference. now look at the range where you would expect MAT/IAT to be: ~112*F and less. note the gigantic jumps between the breakpoints. the 60-90 range would get used a lot, i would think, and over a 16 value range, the temp changes by ~32*.... so each A/D count is roughly a 2*F difference. i mean, it's a workable resolution, but it's not great. you'll notice i didn't fill out the higher temp ranges of the table.... well, the original datasheet didn't have that info, so i can't without guessing.



anyways, your choice, i personally would avoid using a sensor like this for MAT. would be fine for coolant and oil temp, but not MAT.

[EagleMark]

If it were IAT I would agree, but it's MAT so temps are going to be hotter do to engine/intake manifild heat and in the more defined range...

[RobertISaar]

probably. i use the terms interchangeably and often switch between them multiple times in one paragraph. i was also thinking about it some more.....

the resolution used by the ECM is X * 1.35 - 40. so as long as you have a resolution of 1.35* per A/D count in the range that it will be in often, that's really all possible resolution that you'll ever use with an 8 bit A/D.

looking at the 113-92*F range, that's a difference of 21* over 16 counts. so 1.3125* per count. in other words, works to the limit of 8 bit resolution from those temperatures and upwards.

[JeepsAndGuns]

The jeep table posted above is for both MAT and CTS. So the cts works good, but not a mat? That sucks as the jeep cts is same thread/size as the gm, so I can swap it no problem. The mat is the not so easaly swapped sensor and its the one that wont work that great....lol.

So basicly what you are saying is, it will work, but not that great. So I would be better off sticking to converting to the GM MAT?

[Six_Shooter]

If it were IAT I would agree, but it's MAT so temps are going to be hotter do to engine/intake manifild heat and in the more defined range...

There isn't that much of a difference in temp. The air moves quite quickly through the intake tract and the engine, and has less time to get heated than most people want to believe. In many cases there is a boundary layer of air that is pretty stagnant, and due to how most MAT sensors are installed, they tend to measure this boundary layer more than the actual air being used. This is why it's important to get the sensor as far into the air path as possible.

[RobertISaar]

well, check my revised statement.... from 92*F and up, it's actually at the limit of 8 bit A/D resolution. from 60*F to 90*F, it's 32* per 16 counts, so ~1.97 A/D counts per *F(decent resolution), 60*F and under however.... there's 100*F that gets blown through in 16 counts and it's probably very non-linear in that region.

so, in any application where it's measuring from 60*F and higher, i'd probably use it, but under that and to me it's worthless as anything other than a realistic guess at the real temp.

[JeepsAndGuns]

Well it does get colder than 60* here in the winter, so I think I will go back to modding the GM sensor to use it. Sounds like less of a headache.

[EagleMark]

Where is it? Could you loosen intake and drill so shavings can come out? Or close to throttle body to vacuum them out? Didn't I read that $8D guys put in an IAT instead?

[JeepsAndGuns]

Its just behind the throttle body. If you remove the throttle body completely, you could probably vaccume out the shavings. But I would rather turn down the sensor than drill my intake. Gonna try to mod another sensor before I drill intake.

Where did you read about the iat instead of mat?

[EagleMark]

http://www.thirdgen.org/techboard/diy-prom/265869-mat-pw-compensation-8d.html