There has been a popular graph floating around that shows the negative temperature coefficient response of the temperature sensor used in the 2 valve K100 bikes.  It has been difficult to read a given resistance given a particular temperature on this graph. I thought I'd improve that graph.

Today, I recorded a set of data from a new temperature sensor that was obtained from Euro Motoelectric.  I've included the resulting graph as a PNG graphic file.  The raw sensor data is here.



The vertical axis is resistance in Ohms between one of the terminals and the body of the NTC temperature sensor, the horizontal axis is temperature in °C.

Just want to say thanks for this  

Now, how does the injection react to this? Is it constantly changing based on the temperature from this sensor, or does it hit a certain temperature and not get any leaner? Is there a certain resistance we should aim to get under, or is lower always better? Not that mine seems to be getting richer and richer or anything

The Bosch L-Jetronic EFI system is fully analogue in nature.  In 1980's, there wasn't digital technologies as we've come to experience and expect now a days.  So, this NTC resistor device would have driven some sort of analogue amplifier or comparator that would have changed fuelling by adjusting injector timing (total guess on my part as I don't have any documentation that describes exactly what goes on inside the ICU).  

In short, there's no low resistance factor you shoot for.  There's no AFR tables or ignition timing tables or temperature compensation tables in an L-Jetronic. You're correct in saying it is constantly changing based on the current coolant temperature being converted into a resistance reading.

The barn-door air mass sensor also contributes to over-lean or over-rich running conditions.

I've always found it interesting how the various systems operate in telling the EFI how much fuel to add. The early K bikes operated with mostly a mechanical system with minor inputs from the temperature sensor in the water jacket, to adjust for different temps. The TPS on these bikes was really just an anti backfire device in that they cut fuel when you rolled off the throttle.

The first generation R1100's operated with an air sensor in the air box and an 02 sensor.  The air sensor controlled the mixture until the 02 sensor got up to temp and then it adjusted mixtures.  The R1100 only had one 02 sensor so it wasn't as accurate as the later R1200 with two 02 sensors.  The TPS on the R1100 also sent a signal and its adjustment was fairly critical.

I know and have used some of the aftermarket systems that are out there to trick the 02 sensor into sending a different signal to the ECU, so that it would tell EFI to add more fuel, in an attempt to get rid of surging etc.  I've used them on BMW's & Harley's that I've owned and found them of limited value.

I'm curious if anybody has used an alternative to the stock temperature sensor on a K100 2v to send the resistance reading to the EFI. I would think a later air temp sensor would work equally well. I've played with one years ago when I had some R1100's but don't remember the resistance readings, but do remember that it was very sensitive to temp changes. Any comments or thoughts out there.

Not hard to tell it is winter up here. Too much free time.

robmack wrote:The Bosch L-Jetronic EFI system is fully analogue in nature.  In 1980's, there wasn't digital technologies as we've come to experience and expect now a days.  So, this NTC resistor device would have driven some sort of analogue amplifier or comparator that would have changed fuelling by adjusting injector timing (total guess on my part as I don't have any documentation that describes exactly what goes on inside the ICU).  

In short, there's no low resistance factor you shoot for.  There's no AFR tables or ignition timing tables or temperature compensation tables in an L-Jetronic.  You're correct in saying it is constantly changing based on the current coolant temperature being converted into a resistance reading.

The barn-door air mass sensor also contributes to over-lean or over-rich running conditions.

plus the air temp sensor inside the mouth of the afm (barn door device ) adds some to the mix calculations ...likely more compensation  but still an input to the Icu

Thanks for the curve!

Glenn

charlie99 wrote:
plus the air temp sensor inside the mouth of the afm (barn door device ) adds some to the mix calculations ...likely more compensation  but still an input to the Icu

I've heard of that sensor before but I can't find in on the microfiche for the 2 V K100RT. 4V & K1100 had what is called an idle regulator valve, but that wasn't on the 2 V models.  Is that what is referred to as an air temp sensor.

That air temperature sensor is integral to the AFM and is not available separately.

Here is an inverted view of the airflow meter with the air temperature sensor in its inlet indicated by the arrow.

Thanks.

cycleman wrote:The first generation R1100's operated with an air sensor in the air box and an 02 sensor.  The air sensor controlled the mixture until the 02 sensor got up to temp and then it adjusted mixtures.  The R1100 only had one 02 sensor so it wasn't as accurate as the later R1200 with two 02 sensors.  The TPS on the R1100 also sent a signal and its adjustment was fairly critical.

K1100s are very similar.