I've been wanting to experiment with leaner fuelling on my 2008 Suzuki SV650S. I've had good success with a lean burn tune on my Subaru, running as lean as 17.5:1 AFRs with more advanced ignition timing at lower engine loads. Ideally I'd like to try the same on my bike, but I'm not willing to drop $500 on engine management software quite yet, so I'm going to start with a sensor resistance hack to tune the AFR. I may add an ignition timing hack as well. My spark plugs indicate that my AFRs are normal, but I like to tinker. I'm not sure how the engine will respond to running leaner. I hope to achieve better fuel economy in highway cruising.
Fuel injection control overview:
The SV650 uses the speed density method to estimate airflow, running in open loop with no closed loop adjustment from MAF or O2 sensors. Injector duty is primarily affected by throttle position (TPS), intake air pressure (IAP), atmospheric pressure, and rpm. The intake air temperature (IAT) sensor provides compensation, leaning out the fuelling as air density drops with rising temperature. Modifying the pressure sensors or throttle position sensors tends to cause engine codes, and the fail-safe modes for these sensors (if there were a problem with the wiring) would have a big effect on engine operation.
IAT sensor resistance hack:
I am going to modify the resistance of the IAT sensor circuit since it doesn't have these issues. Installing a resistor in parallel with the sensor reduces the resistance of the circuit, tricks the ECU into thinking the intake air is warmer and less dense, and injects less fuel. On some vehicles, IAT sensors will also be used to retard ignition timing at higher air temperatures, but the service manual for this bike doesn't show this sensor having an effect on ignition timing.
I've decided to use a potentiometer (variable resistor) so I can change the AFR on the fly and see how the engine responds. The downside of using a resistor or potentiometer is that it is a fixed resistance, whereas the IAT sensor is a thermistor with a non-linear resistance curve. This results in more leaning at cool IATs, and less leaning at higher IATs. The solution would be to use a thermistor with a similar curve (B-value) to the IAT sensor. However, since IATs will only vary by about 30 degrees C, I'm not concerned about this effect. In fact, less leaning at high intake temperatures may be a good thing to keep cylinder and exhaust temperatures lower, and to help lower the chance of spark knock when riding in hot weather.
I'm using a 0-10 kOhm potentiometer (Mode 62-284-1) in series with a 750 ohm resistor, with this combined resistance in parallel with the IAT sensor. The 750 ohm resistor is included to set a reasonable minimum resistance or maximum leaning effect. I've included an on-off switch as well. The effect of the mod on IAT readings and fuelling is as follows:
10.75 kOhm at 20 C IAT: 6 degrees C warmer, 2% leaner
10.75 kOhm at 40 C IAT: 3.5 degrees C warmer, 1.1% leaner
0.75 kOhm at 20 C IAT: 42 degrees C warmer, 12.5% leaner
0.75 kOhm at 40 C IAT: 28.5 degrees C warmer, 8.3% leaner
I mounted the hardware in a box that I'll mount to the handlebars or dashboard. I've left room for another potentiometer that I may use to advance the ignition timing.
Ignition advance hack:
I may decide to advance the ignition timing by connecting a 1 kOhm potentiometer in series with the engine coolant temperature (ECT) sensor so that the ECU reads a lower coolant temperature. According to the service manual, this sensor can affect ignition timing as well as fuelling. In my Subaru, ignition timing is advanced by about 2 degrees for every 10 degrees C decrease in coolant temperature below operating temperature, up to moderate engine loads and rpm. I'll assume that this engine is similar in this regard, but I will have no way of measuring the effect on ignition timing. I should be able to test any power gains or loss by using my throttle lock and varying the potentiometer.
In terms of fuelling, I assume that the ECT sensor's only purpose is to determine how much to enrichen the mixture during cranking, for a time after startup, and to provide an additional smaller amount of enrichment up to a certain coolant temperature. On my Subaru, there is little to no enrichment above 50-60 degrees C coolant temperature. I'll assume this engine is similar and that I should have some wiggle room to lower the coolant temperature reading without richening the mixture. I can also temporarily turn down the potentiometer resistance to aid cold starting and running if necessary.
Anyway, advancing the ignition is a future project at this point.
Tuning and Testing
First of all, I'm not sure if I'm going to destroy my engine with leaner fuelling. Leaner fuelling will increase combustion chamber temperatures and increase the risk of spark knock, though these engines are apparently quite resistant to knock. With a closed loop fuelling design, I'm thinking there is probably a lot of margin in the engine design for running less than ideal AFRs and ignition timing. The other risk is high exhaust gas temperatures and burning an exhaust valve. I would like to lean the fuelling only up to moderate engine load and rpm, but this would be a bit complicated to do without a Powercommander or ECU tuning software. I'll certainly turn off the device during extended hard riding.
Without a fuel economy meter or a wideband O2, I guess I'll be tuning by feel. I plan to use my throttle lock while varying the fuelling to measure the effect on engine power. I'm not sure what this will accomplish exactly. If the bike is already running rich when cruising, I would expect similar power levels while I lean it out, until it is running leaner than stoiciometric, after which point power will start to decrease while fuel economy may increase.
I'll also read the spark plugs, and will monitor fuel economy by the tank. I'm open to any testing suggestions, or any other advice. Are there any good phone apps that measure acceleration and engine power without needing to connect to the ECU?