Cheating on the IAT sensor

[Luma46]

So the snow finally melted away from the back alley in front of my garage and I could test my latest "mod". I have installed a 680 ohm resistor in series with the IAT sensor on my 2009 RT (got it only last fall). BTW, I measured the resistance of the sensor, it was about 2600 ohms at 7 degC. I have done this mod previously on three fuel injected bikes including my 2006 1200GS to my full satisfaction. Here is what I have noticed so far:

Easier start in cold weather (I am talking just above freezing point)

I can idle the bike after start as little as 15 seconds and it is ready to ride

After my first 900 kms, my average fuel consumption is 4.1 l/100 kms

 

One of the benefits of this mod, which makes the fuel mixture a bit richer, is that I can ride on 87 octane gas without any adverse effects like knocking. I am sure some will disagree with me, but I am backing up my statement with past experience.

[roger 04 rt]

Here is Boston the snow has finally cleared and they're running the street sweepers so there's less sand to contend with. We should be into steady riding now.

 

Like you I wanted to richen the mixture on my bike (which is an 1150) and have learned a great deal about

 

I am surprised that your sensor measures 2.6K at 7C—maybe a bad sensor if you measured it correctly. Here is a chart with the stock curve (blue) for the R1200XX and after the user tried some resistors. At 7C you should have measured closer to 9K ohms.

 

 

Recently terryofperry has provided a wealth of GS-911 and LC-1 AFR data for his R1200. I can say for sure that 680 ohms isn't enough to matter immediately and even if it were the R1200 has a powerful adaptation capability which would "learn" about this sensor error fairly quickly and develop Adaptation Values to ignore it.

 

There are really good ways to get an increase to fueling and there are great benefits if you do. But on Closed Loop motorcycles like the R1200 IAT changes have no lasting effect. Maybe your prior experiences were on bikes that were Open Loop fueled.

[dirtrider]

Morning Luma46

 

You are correct in that I have to disagree with some of your findings.

 

First off, as Roger mentioned The 1200 hexhead BMS-K has a great learning ability so quickly learns from the 02 sensors to adapt out any richer fueling added by the IAT (Intake Air Temp) sensor.

 

Also that 680 ohm resistor is pretty low to do much at cold start. At 7°c that only shifts the intake temp a few degrees. Even the Booster Plug shifts it about 20°c.

 

On the thing about adding that 680 ohm resistance allowing 87 octane gas. You should be able do that with or without any added intake temp resistance as your 2009 hexhead has a knock sensor on each side so automatically removes spark advance until it quits knocking. You lose performance but c-a-n use lower octane gasoline.

 

As long as your 02 sensors are hooked up & working then about any spoofing of the intake air temp will be adapted out as you ride. The BMS-K on the hexhead has a very sophisticated adaptive learning algorithm that can adapt out most intake sensor spoofing.

 

One of the things that make it very difficult to A/B intake spoofing when adding Intake Temp Resistance on the 1200 hexheads is that when first added the open loop part of the fueling does feel better for a short time. As the bike is ridden that intake spoofing is slowly learned (adapted) out so the rider notices the immediate addition of the added IAT resistance but really doesn't notice any drastic change as it slowly learns it's way back to normal fueling based off of the 02 sensors.

 

On the other hand adding that little of resistance to the intake sensor loop really doesn’t hurt much so isn't enough to cause cylinder fuel washing at very cold (below freezing) starts.

SO, if it makes you feel better about your bike then leave it there as it does so little it sure won't hurt anything.

 

[Luma46]

I do agree that once the bike warms up to operating temperature the fuels control system starts working in a closed loop. But it does work in open loop when cold and the added resistance to IAT sensor is not compensated for by O2 sensors.

 

What got me thinking however is the graph presented by roger 04rt. Maybe my rather poor starting at cold temperatures and hesitation to go shortly after start is caused by out of spec IAT sensor. I think I will take the tupperware off again and remeasure this sensor under more controlled conditions.

[roger 04 rt]

I do agree that once the bike warms up to operating temperature the fuels control system starts working in a closed loop. But it does work in open loop when cold and the added resistance to IAT sensor is not compensated for by O2 sensors.

 

What got me thinking however is the graph presented by roger 04rt. Maybe my rather poor starting at cold temperatures and hesitation to go shortly after start is caused by out of spec IAT sensor. I think I will take the tupperware off again and remeasure this sensor under more controlled conditions.

 

It turns out that the Open Loop fueling is corrected by Adaptation Values learned when the bike is warm. So once you're driven for a while warm, the next time you start the BMS-K knows about the "error" created by the 680 ohm resistor and takes it into account even when the bike is too cold for Closed Loop.

 

Regarding your sensor, I did some looking around to complement some data I had and I'm pretty sure the chart is correct for you 1200, so you may have a bad sensor. When you measure the sensor, had you removed the connector fully? I ask that because the BMS-K has some resistance that could have affected your answer.

RB

[dirtrider]

Clip--- But it does work in open loop when cold and the added resistance to IAT sensor is not compensated for by O2 sensors

 

Afternoon Luma46

 

Ah, but it is compensated for from PAST 02 feedback from the last few times you rode the bike.

 

This is what is rarely understood about modern adaptable fuel injection systems. They learn from the closed loop 02 operation then apply that learned off-set (adaptive) to a lot of the open loop operation.

 

It takes a while to re-learn the offsets so that means when you FIRST add the IAT resistance the fueling computer does react & add a bit of fuel in the open loop areas but as you ride the bike, it (the fueling computer) just learns the off-sets (adaptives) & slowly puts the open loop fueling right back close to where it was before the IAT resistance was added.

 

[Luma46]

Regarding your sensor, I did some looking around to complement some data I had and I'm pretty sure the chart is correct for you 1200, so you may have a bad sensor. When you measure the sensor, had you removed the connector fully? I ask that because the BMS-K has some resistance that could have affected your answer.

RB

 

I kept thinking about this IAT sensor resistance vs. temperature graph and tried to backtrack in my mind how I actually measured the sensor's resistance. I am ashamed to share it, but I realized that after disconnecting this connector on top of the air box, I actually measured the contacts on the cable plug and not the sensor itself! So god knows what was I actually reading. This should not happen to a 66 years old engineer, but I might be getting senile, who knows...

 

As for the effect of the 680 ohm on fueling the bike, I do accept the comments that you guys made, mind you, I am not expert in this field. However, I am quite positive on this better cold start and almost instant driveability after cold start.

As for the same mods that I have done earlier, they were on 2002 V-Strom 1000, 2006 Kawasaki Nomad VN1600 and 2006 1200GS. The first two bikes did not have O2 sensors and closed loop fueling and I remember that with 1k ohm on the Nomad I eliminated knocking in hot weather.

Cheers,

[roger 04 rt]

I kept thinking about this IAT sensor resistance vs. temperature graph and tried to backtrack in my mind how I actually measured the sensor's resistance. I am ashamed to share it, but I realized that after disconnecting this connector on top of the air box, I actually measured the contacts on the cable plug and not the sensor itself! So god knows what was I actually reading. This should not happen to a 66 years old engineer, but I might be getting senile, who knows...

 

 

Been there, done that ... you were measuring a resistor inside the BMS-K. Interestingly that resistance measures about two and a half K ohms on the Motronic as well ...

[eej1]

The December issue of MCN has an article about Booster Plugs and seem to like them. Their test was on a F650GS and I don't know if that bike relearns the offsets as described previously. Here is an quote from the article.

 

This was verified during prolonged testing, during which our F650GS test mule was operated first in stock condition, then with the Booster Plug installed, and finally with the Accelerator Module installed. Approximately 1800 miles were covered in each condition. Average gas mileage in stock condition was 48.3 mpg. With the Booster Plug installed, 52.7 mpg was achieved, and the Accelerator Module recorded 53.5 mpg. Both devices delivered an approximate 10% improvement in fuel efficiency compared to stock.

[Fathom]

I also read with great interest the MCN article about the Booster Plug (and a competing device, the Accelerator Module, both of which purport to address rough idle, remove unsightly flat spots, and generally address the -- some would say -- overly lean fuel/air mix on our late model bikes).

 

As the testimonials on the Booster Plug site rave about the product across a wide range of applications, I'm wondering if anyone on the board has tried this (or the Accelerator Module) over any span of time, and if so, could share your experiences. In addition to whether the results are "as advertised," I'm wondering if using one of these would void the warranty, result in blackened / fouled plugs (from a too rich mixture), or vary in performance [read: flake out] when it's cold out to begin with. Both devices appear to work by fooling the EFI system into thinking the ambient air temp is much cooler (68 degrees cooler on the Booster Plug) than it actually is, resulting in a richer mix.

 

FWIW, I see a slightly rough idle on my '13 R1200RT, and have (very) occasionally stalled after downshifting, when on returning to throttle, the bike seems to cough/sputter a moment and dies before the revs start coming back up. I don't purport to have even a fraction of the tech knowledge of many on this board, but it would seem that this could be due to a lean fueling condition ("Not enuf to burn here, so... C'ya!). And who wouldn't want a bit crisper throttle response, as attributed to both models in the MCN review?

 

 

 

[smiller]

This is what is rarely understood about modern adaptable fuel injection systems. They learn from the closed loop 02 operation then apply that learned off-set (adaptive) to a lot of the open loop operation.

I've seen a lot of back-and-forth as to whether/which BMW ECUs support fuel trims to the open loop map. I think this does not exist in the R1100 management system, but I'm not sure about later models. Does anyone know if any definitive information exists on this topic? It's pretty critical to the question as to whether these IAT-spoofing devices have any potential value or not so if it were me I'd want to know that before purchasing one. These things are all the rage on the K-bike forums but for the above reason I've wondered whether they could really be effective or not (at least for more than the first few miles.)

[racer7]

Hunt up roger04 at ADVrider or MOA and ask him- he's the guy who gathered all the data for AF-XiED device that allows setting the desired offset for CLOSED LOOP operation. Bet he tells you the same thing as DR.

IMO the AF-XiED is a more fundamentally useful approach than simple IAT device but it is capable of working with them.

 

Or hook up your GS-911 and learn how to study the real time data- it will tell you all about injection times, lamda, and about anything else you want to know.

 

The BMS-K (on most everything since 05) is a far more powerful system than the prior Motronics.

 

The AF-XiED totally transformed the running of the SOs 04 R1100S (Motronics system)- to the extent that if I hadn't done it myself I likely wouldn't believe it. And hers was running as well as ay oilhead I've ever seen in stock form before the change. Now, roll on throttle tests against my stock 08 RT come out nearly dead even, something that wasn't even close before. Most cost effective change Ive ever seen for a tuning improvement.

Wouldn't expect as dramatic an effect on BMS-K bikes but plan to test a few to find out.

[roger 04 rt]

The December issue of MCN has an article about Booster Plugs and seem to like them. Their test was on a F650GS and I don't know if that bike relearns the offsets as described previously. Here is an quote from the article.

 

This was verified during prolonged testing, during which our F650GS test mule was operated first in stock condition, then with the Booster Plug installed, and finally with the Accelerator Module installed. Approximately 1800 miles were covered in each condition. Average gas mileage in stock condition was 48.3 mpg. With the Booster Plug installed, 52.7 mpg was achieved, and the Accelerator Module recorded 53.5 mpg. Both devices delivered an approximate 10% improvement in fuel efficiency compared to stock.

 

I don't mean to sound like the grinch that stole christmas, but there really isn't a mechanism by which something like a BoosterPlug can improve gas mileage. Jens at BoosterPlug is careful to say that a motorcycle will use the same or a little more.

 

This is what is rarely understood about modern adaptable fuel injection systems. They learn from the closed loop 02 operation then apply that learned off-set (adaptive) to a lot of the open loop operation.

I've seen a lot of back-and-forth as to whether/which BMW ECUs support fuel trims to the open loop map. I think this does not exist in the R1100 management system, but I'm not sure about later models. Does anyone know if any definitive information exists on this topic? It's pretty critical to the question as to whether these IAT-spoofing devices have any potential value or not so if it were me I'd want to know that before purchasing one. These things are all the rage on the K-bike forums but for the above reason I've wondered whether they could really be effective or not (at least for more than the first few miles.)

 

Here's something that I wrote for another forum question earlier today.

 

Roger,

 

Please correct me if I am getting this wrong. You are saying that eventually the BMSK self corrects for any attempt a unit such at the BP makes to modify the mixture.

 

That's exactly what is happening, and that's what you and the designers would want. Here's a partial list of things that might not be precisely in spec: fuel composition +/- ethanol, fuel pressure, fuel injector flow rate, air filter cleanliness, carbon build up on the valves, pistion or CH, TPS reading, air temp reading, oil temp reading, battery voltage reading, injector dead-time vs battery voltage, on-board barometor, actual air humidity, etc. All those things can change, be inaccurate or have aged.

 

As a result of those possible variances, and given the high accuracy of the stock O2 sensor (by far the most accurate on the motorcycle related to combustion) and the position it has of measuring the actual result of combustion, the BMSK and Motronic both keep internal records of how much fuel was actually needed to hit an AFR of 14.7:1. Then it analyzes those deviations and comes up with correction factors that Bosch calls Adaptation Values--invented in the late 1970s.

 

As for any sensor, if it gives a signal that leads to more or less fuel than ideal (whether intentionally like an IAT modifier or unintentionally as in an error), the BMSK learns that "error" (without knowing which sensor is wrong) and saves up those records which it uses as corrections.

 

The Boosterplug is a well made product by a guy who knows what he's doing. It's pretty accurate too. It has an interesting application together with an AF-XIED or LC-1 (Jens at BoosterPlug runs both BP and LC-1 on his BMSK-bike, an F800S). In tandem, the BP says the air is colder use more fuel and the AF-XIED when set rich says to the BMSK, that's the right amount--so it speeds the Adaptation Process. But if you give it time, the AF-XIED or LC-1 will do the job on its own.

 

The BMSK is so clever that if you disconnect one O2 sensor, it will use the estimates from the connected cylinder and apply them to the other side. I have the data, watching the mixture and seeing it happen. If you disconnect both, it still tries to keep the Cat working by varying the fueling on its own but that leads to a messy, wide fueling spread, which is why the PC V isn't a great pair with the BMSK. The PC V instructions say to get rid of both O2s and the BMSK says, okay then deal with this fueling spread. Not a good combo.

RB

[WestyLancs]

Without getting technical, a couple of years ago I fitted an Accelerator (compact version) to my RT in an attempt to address a random cut

out when slowing down. (subsequently turned out to be a stepper motor issue)

On first fitting there seemed to be a less jerky response at very slow speed on opening the throttle, but over time the bike feels exactly as it was

initially, seemingly having learned to compensate. (as others have technically explained)

GS-911 readings show the module to still be 'fooling' the IAT by -15 or 20C, so it's still working.

Fuel consumption remained unchanged, and with my mainly two up riding style I haven't noticed much, if any change to the bike otherwise.

I'm NOT saying it was a waste of money, at (then) 35GBP it was only slightly more than the cost of a tank of gas, and an avenue to explore to sort

the cut out, but I am saying it may not be as long term improvement some are looking for.

 

As always YMMV.

[sardineone]

First a high five for all the collective knowledge on this board

 

You can fool a IAT some of the time, but the BMWST will catch up with you so it seems. Many thanks to all that have contributed including those that are looking for reasonable worthwhile mods.