Sensors

Throttle Position Sensor

The throttle position sensor (TPS) is mounted on the throttle body.  It is a 3-terminal variable resistance device also known as a potentiometer.  Although it is a resistance device, the ECU uses it in conjunction with an internal fixed-value resistor to form a variable voltage divider.  This voltage divider is powered by a precision 5-volt reference inside the ECU.  This is why we can measure a voltage on the TPS that is about 0.6 V at idle.  As the throttle opens, the voltage increases.

When the TPS is adjusted to produce 0.62 volts with the correct idle speed, backing the idle-stop screw all the way out results in 0.342 volts.  WOT yields about 3.9 volts.

Note that the voltage does not vary all the way from 0 to 5 volts as the throttle moves from closed to wide open.  This allows the ECU to detect a failed throttle position sensor.  In automotive applications, a failed TPS puts the ECU into “limp home mode” in which a “moderate” throttle opening is assumed and appropriate fueling is provided.  The engine will run very poorly, but it will run.  I have no reason to believe the TR280i would not adhere to this convention but have not tested it.

The K-Scan diagnostic software reports throttle positions that range from 5 degrees to 82 degrees when rotating the twistgrip from stop to stop.

Setting the TPS

It is important to follow the factory procedure to set the idle speed at 1400 rpm.  While working with a local OSSA dealer, I was allowed some brief contact with a factory EFI technician.  He admitted that the strange idle-setting procedure is simply because it yields a throttle setting that provides the easiest starting (and had nothing to do with EURO3 compliance).  I also learned that OSSA was looking into using a different throttle body that would permit customer alteration of the idle speed.

On my own, using K-Scan, I noticed that the ignition timing changes hugely at the slightest movement of the throttle.  It is about 20° BTDC while idling, and jumps to approximately 30° BTDC just off idle (exact behavior is map dependent).  This produces exactly the behavior desirable in a trials bike – instant throttle response just off idle.

This is also the reason OSSA says to use no throttle when kickstarting.  I have found this to generally be a beneficial practice – but once the engine fires, using a little throttle will encourage it to keep running.

It is a bit difficult to determine, but I believe that at kicking speed the ignition timing is retarded to about 8° BTDC.  That is a number that makes sense to me for starting.  As the ignition advance increases, there is a higher probability of “kickback” due to combustion pressure increasing enough to prevent rotation past TDC in the running direction.

I have experimented with different initial idle speeds and TPS settings only to find that the factory procedure yields the best power-delivery characteristics.  Any increase in the TPS zero-point alters the ignition timing from 20° BTDC to 30° BTDC.  This causes a rough idle and potential stalling at inopportune moments.  It is possible to reset the TPS to allow an increased idle speed but with the trade-off of less-than-stellar running elsewhere.  The slow idle was, to me, preferable to imperfect fueling elsewhere.

Since the recommended 1400 rpm idle speed is too low for my liking, I came up with a solution that gives the best of both worlds.  See the section on “Idle Air Bypass”.

The engine speed readout in K-Scan is not filtered heavily enough.  This causes the reported engine speed to fluctuate greatly while setting the TPS.  I always use the engine speed reported by my Trail Tech TTO Tach/Hourmeter instead.

Coolant Temperature Sensor

The coolant temperature sensor (located in the head) utilizes a thermistor (a 2-terminal resistance device that exhibits different resistances at different temperatures, however in a highly non-linear manner).  It is nominally 2k ohms at 25 degrees C.

It works in conjunction with a precision resistor and a voltage reference inside the ECU.  Knowing the coolant temperature also allows the ECU to control the fan via the chassis relay.

Below is a table I made using K-Scan and a decade resistance box.  I set the resistance box for an even multiple of 100 ohms and read the temperature reported by K-Scan.  I could have made this table a lot more precise by showing the resistance at which the temperature transitioned from one value to the next, but that was unnecessary for my purpose.  I just needed a rough idea to see if I could find a replacement thermistor for a sensor that had failed.  I chose US Sensor's part number 202FG1J.  These cost under $2 each in small quantities.

As with other sensors, the full range of the thermistor is not used in order to provide a means of detecting a failed sensor.  If the coolant temperature sensor fails, a default coolant temperate of 80° C is assumed by the ECU – this makes the fan run continuously.

 Air Temperature Sensor

The air temperature sensor (located above the throttle body) is also a thermistor.  It is not the same as the coolant temperature sensor and is nominally 10K ohms at 25 degrees C.  If the air temperature sensor fails, a default air temperature of 0° C is assumed by the ECU.  I have measured its characteristics in the same way as the coolant temperature sensor.  Values for temperature versus resistance are in the same spreadsheet above.  Because none of my air temperature sensors have failed, I have not determined a suitable replacement thermistor. 

The OSSA sensor is marked 174H.  The Arctic Cat sensor is marked 562H.  Although not identical, it probably could be substituted but would produce readings several degrees C different at typical ambient temperatures.

Barometric Pressure Sensor

The barometric pressure sensor (located in the airbox) is a solid-state device that responds to changes in atmospheric (barometric) pressure.  It is powered by the ECU's 5-volt reference and produces a voltage proportional to absolute pressure.  Note that it does not measure the pressure in the intake manifold.  OSSA called it a MAP (manifold absolute pressure) sensor.  This is incorrect.

At one point in my experimentation, I attempted to modify the fueling by attaching a long hose to the baro sensor so I could suck/blow on it while riding.  In this way, I expected to lean the mixture by sucking (lower atmospheric pressure necessitates less fuel) or blowing (higher atmospheric pressure necessitates more fuel).  But the results were highly inconclusive.  I suspect it is difficult to maintain a constant vacuum/pressure.  But I also wonder if the barometric pressure is only sensed when the ECU first boots.  I have heard of EFI vehicles that would run more and more poorly when climbing a mountain, for example.  But stopping and restarting their engines would make it run correctly again (well, as correctly as possible given the reduced atmospheric pressure).     

Low Fuel Sensor

The low fuel sensor is located in the fuel tank near the throttle body.  It has a 14mm hex head for screwing it into the tank.  Its electrical connection mates with a 2-pin JST JWFP connector.  The sensor is used to illuminate an optional warning light when less than 3/4 liter of fuel remains.

The sensor's logo appears to read “TC!” but is a stylized version of the letters “TCI”.   That logo is also found on the final OSSA wiring diagram.  A bit of Googling revealed an electrical manufacturer in Spain called TCI Conexiones, sl. 

The sensor itself is a negative temperature coefficient (NTC) thermistor.  A characteristic of all NTC thermistors is that as its temperature increases its resistance decreases.

The complete circuit is just the thermistor, an incandescent light bulb, and 12-volt power all connected in series.  This idea, although new to me, has been used in motorcycle and automotive applications for a long time.

The thermistor self-heats when power is applied and its resistance decreases.   A lack of fuel allows sufficient current to flow in the circuit to illuminate the low-fuel lamp.   

However, the presence of fuel conducts that self-produced heat away from the thermistor, causing its resistance to increase thereby extinguishing the lamp.  Simple and clever.

For reference, my thermistor measured 1560 ohms at 65 degrees F.