MSefi.com read only forum (Jan.15,2018)

I've got a supercharged toyota motor (22R).
The blower is a positive displacement roots type supercharger.

I've got a Holley type IAT sensor, basically the same thing Holley uses on their digital fuel injection units (TBI). It's made to be put above the TBI -somewhere in the air filter.

I don't know how much the air intake charge is heated with my blower, but I'm guessing a good deal... The question is - do I have to get this sensor - or some sensor like it, into the manifold tract - on the compression side of the supercharger? If so - anyone have a suggestion for a very thin IAT - NPT thread, if possible.

Basically I'm using a GM TBI in a draw-through setup for my EFI.

I would think you need it on the compression side also. That blower will heat up the air.

FYI - NICE set-up!!

dcg9381 wrote:but I'm guessing a good deal...

And your guess is correct, a good Roots blower is 65% efficient at peak, most of the time they are running 50% or lower, so you get lots of heat from them...

You'll want an "open cage" sensor, other than that not much matters. You don't really need the security of an NPT fitting, you can just use a small clamp to hold it in a hole in the manifold (the force exerted by 20 psig boost on a 1/4"/6mm hole is only 4 lbs/18 N).

Eric

Anyone have examples of smaller open element IAT sensors? Something relatively small, still compatible with GM resistance values (preferably)?

I could probably get a temperature sensor in there made for coolant - but I don't think it would work well for a wide range of temperature variations and it'd also deal with a lot of heat soak.

References appreciated...

Ended up "building" my own.
I have a holley IAT sensor - Lance, I believe it's the same one you use.
Anyway, I cut off the GM weather fitting, drilled out a 1/8 NPT plug, inserted it into the plug, JB Weld and should do what I need..

I stumbled across:
CYB-SENSS7E (cyberdyne) also. 1/8NPT.
Can't find a photo or resistance / temp values though..

Still a newbie with my MS project. I am also loking for a small AIT for my turboed Yamaha RX snowmobile. What exactly did you guys come up with and is it better to have the sensor down stream of the turbo to pickup the actual temperature of the compressed air. My setup is a draw through system also and because of space limitations a 1/8 screw in unit would be perfect to screw into the exit side of my turbo. Did your CYB-SENSS7E (cyberdyne) 1/8 screw in sensor work? Does it work fast enough?

I am converting this http://www.lightningprollc.com/?page=ga ... RT%20C.JPG to an efi sytem and with the stack up of a throttle body and air-filter I don't need any more restrictions on a 300 hp snowmobile.

What a great site and source of information.

LightningPro wrote:What exactly did you guys come up with and is it better to have the sensor down stream of the turbo to pickup the actual temperature of the compressed air. My setup is a draw through system also and because of space limitations a 1/8 screw in unit would be perfect to screw into the exit side of my turbo.

Bob,

I'm thinking that you might want to make your own IAT sensors. From what I can figure, the GM sensor is just a run of the mill NTC thermistor with a B-constant of about 3975 (here's a challenge: first one to confirm or refute this gets a nickel ), and resistance of 3.2k ohms at 25C. Looking on digikey.com, I can see a couple of parts with B=3950 and r=3.3k, so pretty close. These are SMT parts, so are tiiiny, and would therefore probably have excellent response.

Someone who knows anything about this, please check these out and tell us what you think...

490-2425-1-ND
490-2447-1-ND

http://dkc3.digikey.com/PDF/T053/1242.pdf

Thanks,
Eric

Thanks Eric,

Everything is starting to come together with the gathering of sensors and making a compact setup. I don't mind the idea of building my own, but considering that I am looking at making quantities of EFI upgrade kits to our turbo kits, do you still think it would be better to make my own IAT unit or to buy and slightly modify. I am trying to stay away from making electronics as my business is cnc/machine parts and I want the reliability of manufactured components in our Turbo/EFI kit.

Right now I am depending a lot on pictures to help me learn what some you guys are talking about.

So Eric when you say build your own do you mean I should purchase the small thermal wire component & leads, drill out a 1/8 -27 pipe plug, epoxy the thermal wire into and protude into the turbo exit air passage and "tada" I have a small compact IAT that is fast responding and can withstand possible harmonic vibration and is resistant to heat and gas/alcohol/nitrous. Can this be done? (Hmmmm, me likee!)

i think he means that if you get a non-screw-in IAT sensor, just drill the guts out of a plug, screw that plug into the intake somewhere and then glue in the sensor into the newly drilled hole.

Ok, what I'm thinking is that you already have a hole in the exit of the turbo, right? Well, find a hollow fitting that fits the hole, purchase some thermistor devices of the appropriate characteristics (both electrical and mechanical), then solder wires (any wire will work, these aren't thermocouples) to the sensors, and pot the device (say, with epoxy) into the fittings, such that the thermistor is in the air flow and the wires are going back through the fitting. Add a robust connector to the wires poking out the back and off you go. The connector is by far the most expensive part in the whole package.

The two thermistors I reference above are simple surface mount resistors, the small one is about 1 mm long, so as I mention, should have very low thermal mass and should be extremely responsive. The temp range is good, -40-150C, the only thing I wonder about is their tolerance of petrol...

Eric

So I've got a thermister going through an 1/8 NPT pipe fitting (brass) - insulated from the fitting with JB Weld.

I think I still have heat soak. My MAT readings are around 190 F during normal driving. Not a lot of variation...


I'd go to a plug-in type, but 12+ psi on a plug is a lot of pressure.
Maybe a plastic thread in type plug?

RS Components (www.rswww.com) here in the UK sell a gm compatible thermistor under part number:151-215 (we call them 'sperms')

Ive attached the datasheet for them (page 5 onwards)

I'm also looking for a fast reacting IAT sensor that fits in a small hole and can handle a rough environment. I'll try to explain what I'm thinking of doing. Any other ideas are appreciated.

Solder a wire to one side of the SMD thermistor (mentioned above) and pass it through a 1/8" brass tube. Heat shrink with adhesive would be on the tip of the thermistor and wire to protect from shorts. A bare wire will act as a strap and also ground the thermistor. It will be soldered to one side of the brass tube, over the top of the thermistor and soldered to the other side of the tube. Somewhere along the line, I would pot the inside of the tube with RTV or jb weld.

Any fitting can be soldered to the tube.

I hope I explained that well enough. What do you guys think?

Kirk

efahl wrote:I'm thinking that you might want to make your own IAT sensors. From what I can figure, the GM sensor is just a run of the mill NTC thermistor with a B-constant of about 3975 (here's a challenge: first one to confirm or refute this gets a nickel ), and resistance of 3.2k ohms at 25C. Looking on digikey.com, I can see a couple of parts with B=3950 and r=3.3k, so pretty close. These are SMT parts, so are tiiiny, and would therefore probably have excellent response.

Someone who knows anything about this, please check these out and tell us what you think...

490-2425-1-ND
490-2447-1-ND

These are surface mount units. Does anybody know of a lead-type thermister with these specs that is available in North America? Could this be converted to a lead-type by soldering leads onto it? I'd be concerned about the solder joints breaking and it getting sucked into the engine, although it's so small, it probably wouldn't do any harm to the engine, but you'd be left with not IAT.

Surface mount or not, the leads can break.
It IAT that I used from Holley simply wrapped the whole thing up in some heat shrink tubing. If the leads break, the tubing still holds.

Standard NTC sensors will work very well , have used quite a few homemade IATs in mS installs .
The sensors used by OEMs ( the actual NTC resistor) look pretty similar to what you buy in from electronic supply houses , but it may be that the OEM sensors are special.
But I have not had any problems so far with std NTCs . Have immersed them in gas for days on end , no problems .

I use a NTC sensor with 2700 ohms at 25 Centigrade , B value of 3977.
Time constant is some 0,9 s.

Quite a close match to the standard GM sensor , reducing bias resistor from 2490 to 2300 ohms makes it even better.

(Fabricating an enclosure with a good connector is the tricky part , especially if you intend to sell it to the public . )

Eric´s SM NTC resistor is probably faster , but more challenging to
connect to .

Heribert