temp-sensors

SE416 EFI Air and coolant temperature sensors.

(see applicable engines)
THE ECT AND IAT are near the same design , only the environment (water/air) is different (and speed of reaction) (they are located here)
This sensor data below (SE416) does not match up to any GrandVitara SQ416 from 1989 to 2004 and beyond. nor any Grand Vitara JB416,
The SQ 1.6L, uses Dash 50G01 (this alien sensor reads 176 ohms at 50C +-20%) (seen also on some G16 Esteem engines too! 95-97' )
This page does not apply to any carburettor fueled engines at all. (eg: 1989 SE413s)

The below chart, are the IAT and ECT , Intake Air Temperature and Engine Coolant Temperature Sensors.

This data applies to USA ECT sensors only, using 16valve G16B engines. Suzuki P/N. 13650-57F00 (there are over 5 other ECT sensors world wide that do not behave like the below charts. This data, is USA data ONLY !
This data also applies to G16A TBI engines, from 1989 to 1995 in the USA. Suffix dash -61A00 and -61B00 and -84101 sensors. And the below covers all 90-95 SJ413 Samurai sensors (water and air)

This device is just a Thermistor (a resistor with a predictable NTC curve)
By the way, the Water (coolant) sensor does not read coolant temperatures correctly, if you let the coolant level drops below the tip of said sensor. (cool the engine FULLY, open the radiator cap, look down inside, got coolant?)
Same with upside down thermostats or the wrong thermostat gasket.
If you look at the chart, on a cool morning ,say 68ºF, both sensors read about 2500 ohms (yes, disconnected) Does yours?
A hot engine , will hold this ECT at 300 Ohms (about) The IAT may show near 140ºF or 600 ohm, a scan tool , best shows what the ECU see's.

Any hand held DMM " ohm meter" scale (resistance range on any DMM) do this. (so can a voltage checks for ECT)
To check hot temps? (no scan tools in shop) (ohms) stop engine, key off, pull the connector off the ECT , and measure ohms now (inside 5 minutes on hot engine) What you see is what you get (got).
The old OBD1 ECU's measure volts at the sensor and reports volts to scanner, (the old scan tool must do this conversion accurately, many such tool don't) So for those folks, with THAT tool? , I offer the last column, VOLTS.
It would go with out saying, if using the wrong ECT , the ohms and volts will be WRONG.
The new OBD2 ECU (1996+) reports in Degree's C' PID 05. This scan tool only needs to report what it saw !
The plotted curve is logarithmic. (useful for interpolations) -57F00 data.. The normal hot temperature and nomal themostat (EPA regs) is 180F in the USA. Seen in RED below. (data in blk/white below is my testing but data on yellow is FSM data)

Factory spec.	Spec Ohms		Data measured in Fahrenheit Ohms				medium	Calc.
Temp. C° (F)	Coolant ECT	AIR IAT	ECT mine ohms	ECT Temp. F	IAT mine ohms	IAT Temp. F	tested	Volts ECT
-18(0)	14650							4.4v
-7 (20)	8100							4.0
0 (32)	5800	6100	5600	33	5600	33	H²O	3.7
4 (40)	4780							3.5
20 (68) room	2450	2600	2450	69	2550	71	air	2.8
21(70)	2350							2.7
38(100)	1250		1724	81	1900	81	H²O	1.9
40 (104)	1400	1200	5030	36	5200	42	air	2.1
60 (140)	580	620	--	--	--	--	--	1.1
71(160)	400							.83
80 (176)	320	340	--	--	--	--	--	.69
82(180)	300?	300?						.65
99(210)	190							.43
pan boiling water	unk	unk	205	212	210	212	H²O	.46 ?
FSM GM & Suzuki	FSM	FSM	new		new			? estimate

Tests: (ECT removed from engine)

My freezer holds 0º F & my Refrigerator does 40F and Ice water near 32F, the 212 was done in boiling water, all at 500ft altitude .

Both Sensors are simple NTC Thermistors.
The IAT is just faster to react and works in air best., but otherwise same readings.
At about 180ºF the resistance is near 300 ohms. (0.65volts) If yours is near that, then it is good. do not throw away good ECTs. ± 10%? but some Suzuki books show ± 20% spec. accuracy variances.
GM (Geo)) Books tend to show a single average readings.
The sensors form a classic voltage divider. (the ECU provides the voltage source and the divided value is scaled to temperature.)

GM data : (USA. GeoTracker FSM book 1995 (92-98 same) How odd GM, is not showing the 180F normal running temperature data. (see my curve below for interpolated values)

How many volts should I read, on my ECT pin? (see right side if table)

The voltages are from backprobing the ECT connector (connected) and a running engine.

The ground wire at ECT ,Gray-yell is 0 volts at all times. or this wire is broken open.
If you pull the connector (has locks), to the sensor , key on, the hot pin Red-Yel, goes to +5vdc. (if not ? the ECU may be bad)
This happens due to the ECU internal pullup 2000 ohm resistor.
If you ground this Red-yel pin, key on, the ECU , throws a DTC for too low voltage ECT readings , fast.
If you float this pin (cut it ) the ECU reports, ECT open (too high voltage)
On all engines this ECT, sensor is of great importance. The IAT is too, on the 8v engine only. The first 16v had no IAT at all in 1992. (the MPI engines only are a fine adjustment to AFR, due to a very accurate MAF)

ECT: (first off, if you use a scan tool, you will see what the ECT reads, and can be checked against an IR pyro gun for accuracy )

The Hot Red-Yel wire, about .65volts at 180F, my 96' ECU has 2000 pullup resistor to 5vdc. (+5vRefr pin is the power source inside ECU)

The math: the voltage is 5v, (some ECU can read 4.7v) so do you own math at your cars actual voltage.
E = volts
I = Amps
R = resistance. Ohms law covers it.
The ECT current flowing is I = E/R or I = 5v / 2000+300 or 2300 ohms total and 0.0022 amps total current, (Kirchoff's laws for series circuits,,, on current, state current the same in this path)
I then do E = I x R on the ECT ,to get the sensor voltage, that being E = .002amps x 300 ohms, or about 0.6vdc.

With water (coolant ) at near room temperature. The voltage is about 2.5v, about 1/2 due to a 2000 / 2000 ohm ratio.
The one critical temperature is 158F , this is where the ECU changes modes, cold to hot. This happens about at 1.0 volts across the ECT.
Let me do the 158f temp calculations (400 ohms the approximate critical cross over ohms (temperature)
So, 2000 + 400 is 2400 ohms It = current total, = 5v/2400 (5 divided by 2400) is 0.0021 amps, so E= I x R, or .0021 x 400 or 0.84 volts
That means if the voltage dont drop below ! 0.84volts, then the ECU will be locked in cold/warm startup for ever. and will suck fuel like mad.....
The car left the factory at 180F and 0.65volts is NORMAL. Mine runs a 195F (for more fuel economy and mine reads near 0.5vdc at this temperature.
If the ECT does not cross the 0.84v barrier and lower, either the thermostat (installed wrong is top reason) is bad or the ECT is bad, both are easy to test. (IR gun and Ohm meter in hand)
The ECT or IAT can removed at any time and tested in hot or cold water.
The ECU operates in 4 modes (and more) cold start, cold run, warm run and hot run.

Fahrenheit

My Megasquirt custom ECU project needs, the 81 degree F , temp. spec., so you can calibrate the tables. The readings above are on a brand new sensor , carefully measured.

-- = not measured.
unk = Suzuki, doesnt say.

Negative temperature coefficient Thermistors. (what's inside )
91 Tracker 1.6L 8V
All values are on OHMS.
Factory spec. is out of 1991 FSM 8V manual Chapter 6E97-98

all readings at Sea-Level + 450 ft.

ECTs can age poorly and the IAT almost never fails.

My tests were:
room temp, Freezer, Refrigerator, and boiling water, Ice water (32) and bath water 81F (let warm tap water drop naturally) and last Icewater with Ice cubes test.

Used a new IR gun to take measurements.

The newer IAT sensor is of the lower mass type , so it will be far quicker and will help 8v TBI engine performance and economy.
1992/93 16v MPI cars had no IAT until 1994, (but calif started in 1993) The IAT don;t do much on any 16v. MPI.. (multipoint injection)

This MS. Excel plot with ("Sum of least squares fit" equation) the excel file is here, if you want to play with it more. {the black line below, is Excel doing a match up for, LOG(x)}The data is a full mix of GM/Suzuki.

The funny aberration is Suzuki data and GM , (errors, in that one company uses exact numbers and GM averages them, or ?) but is of no concern at all.

Other SE416s world wide ECT including Santana's from Spain (deep ECP searches) revealed : Only EFI never carb.'s.
Most SE416s G16A or B world and with EFI only, run this above curve, but not any SQ416. Many are still running CARB's outside USA., and do not use this sensor at all.
I examined the 28 different SE416 world wide, and all G16A or B engines (non carb.) all match the above curves. Market codes E01 to E98)
The page above does not cover any JDM cars made. (made, and sold in Japan, for use inside Japan)
For funs and giggle's I added the 1989 Suzuki SE416 ECT graph, and again they fail to show 180F. (oddly) once below 300 ohms, it's a DON'T CARE any way. (0.3k = 300 ohms.) this 89/90 sensor below, is the dash 84101 sensor, by Suzuki.
The sensor is also inaccurate below 300 ohms....

v3 1-30-13 , added volts test. revised for GM data and volts on the ECU pin. and the Excel plot of USA ECT. 4-10-2015 I did overkill on this page, to solve problems with bad scan tools. (by request)