Capturing the Data (continued)
As mentioned earlier, the LM-1 has six built-in channels, one of which is pre-populated by the wide-band oxygen sensor. In order to utilize the other channels we used the LMA-3 "Aux Box" device. This device is really the core tool for data-logging with the LM-1 because it comes with several built-in sensors and functions. For instance, the built in 3-bar MAP sensor simplifies measuring boost pressure. We simply T'd the vacuum line from our boost gauge and connected to the port on the LMA-3. The unit also has a built in rpm converter, so we just tapped in to the MSD tach-ouput lead. We were also interested in monitoring fuel pressure. For this we replaced the analog pressure gauge in the fuel line with a pressure transducer. This sensor sends a linear voltage signal, 0.5V for 0 psi to 4.5V for 100 psi. The steps below show how to hook up the LM1 and LMA-3 and begin logging data.

The first step is to hook up the auxiliary unit. We connected a vacuum line to the built-in MAP sensor on the LMA-3 to measure boost pressure. The terminals on the left are for the other channels. Channel 1 is default for RPM input, so that is where we connected our tach lead from the MSD unit. We followed the instructions to set the unit for 8 cylinders. The unit should be mounted inside the cabin, and level if using the built-in accelerometer.	Sensors, such as this 0-100psi pressure transducer, are used to send data to the LMA-3 in 0-5V signals. The sensor has three wires (5V input, ground, and 0-5V output.) The LMA-3 has a convenient 5V output (bottom terminal) so you do not have to create one. The output lead from the sensor connects to an open channel on the LMA-3. Sensors like these are found on many Ford and GM cars, and also in aftermarket products like Autometer electric gauges.
The next step is to install the wide-band O2 sensor. Either weld in a bung (see side bar) or alternatively use an exhaust clamp from Innovate, as seen here on our '01. Note that for the sensor to read accurately you must locate it before the catalytic converters, or as in our case have no converters to deal with. Because condensation in the pipe can damage the sensor it is good practice to only install the sensor when you plan to data log.	Make the connections to the LM-1. The unit comes with a cigarette lighter power adapter, this must be connected as it operates the O2 sensor heater. The O2 sensor cable plugs in to the port marked "sensor". The auxiliary unit (LMA-2 or LMA-3) connects to "Aux In". The "Serial Port" connection is used to connect to a PC, and the "Analog Out" port can be used to send data to another system or external A/F gauge.
Once the connections are made, turn the unit on. The very first time the unit is powered it will perform a calibration for the sensor. After that you can start the motor and the unit. The LM-1 LCD will display a "warm up mode" for about 20 seconds then it should display the air/fuel ratio of the exhaust or 20.9% oxygen if the sensor is in free air. A calibrate button is provided in order to calibrate the unit due to air density changes.	If you have a laptop you can view the data in real time using Logworks. Because voltage itself is not terribly meaningful you can configure each channel to what the voltage converts to for your particular sensor. AFR, RPM, and the built in sensors on the LMA-3 come pre-configured for ease. When ready to log data, hit the red record button. Press the button again when you want to stop recording.

Analyzing the Data
We went out to our favorite stretch of desolate road and made a half-dozen runs in various gears, full and part-throttle. The nice thing about the LM-1 is that unlike other tools or track testing, you do not have to run through all the gears or a full quarter mile to get results. We got the car up to a good cruise rpm hit the red "record" button on the box and then stabbed the throttle. The red button is pressed again to stop recording the session. The LM-1 will assign a sequential file name to each session until the 44 minute storage capacity is used up. Then you must download the data to your PC to clear the memory.

We opened up the LogWorks program to review the sessions. Here we can not only analyze the data in more ways then a ballot recount in Florida, but we can also replay the session and watch all the gauge needles dance around as they did during the drive. Below is a screenshot of what one of our runs looked like while under boost. Each color trace on the graph corresponds to a data channel. We were logging on 4 channels: AFR, RPM, Fuel pressure, and boost pressure (MAP). We turned off the remaining two channels because we were not logging data on them.

What our data shows is that as the supercharger builds boost (blue line) the air/fuel ratio goes into dangerously lean territory (pink line.) Note that we have set the Y-axis to show air/fuel ratio (AFR) units. This can be changed to show the units for any of the channels. The X-axis always shows time in seconds. We have set a call-out just past the 8 second mark which lists the data for all channels at the point. Note the AFR is at 17:1 when boost pressure is at 4psi. Looking at the fuel pressure (green line) we see the likely source of the problem - not enough fuel. Baseline fuel pressure without boost is 10psi. Our boost referenced fuel pump should be rising 1psi above baseline for each 1psi rise in boost. We would expect 14psi or more of fuel pressure, which we are not quite achieving. Another observation is the major fluctuation in fuel pressure as boost comes on. With this data we are able to conclude with fair certainty that fuel starvation is the problem, and the next step will be to upgrade the fuel system. We can also view and sort the data in several different ways (see side bar) to get the information we need.

Endless Capability
Air/fuel ratio, rpm and fuel pressure are very fundamental tuning parameters. However they are just scratching the surface of the tuning capabilities with data logging. The possibilities for what you can monitor are virtually endless. One need not look further than many late model vehicles, which have sensors to monitor all sorts of engine and vehicle conditions. Chances are there is a 0-5V sensor out there that can measure whatever it is you are interested in logging. If there isn't, then someone on the Innovate Support Forums has likely put together a circuit diagram for how to build the sensor using Radio Shack parts. In any case the exciting thing about the LM-1 is that there is a rapidly growing online user community from which you can learn, share ideas, and get technical help. In fact, the LM-1's principal engineer, Klaus Allmendinger, is routinely on the forums providing technical assistance.

A few ideas we're starting to play around with, and which you will see in future articles, are using the LMA-3's internal G-force sensor to measure rate of acceleration. We can make changes to the engine or vehicle, log a run, and then using LogWorks overlay function to compare before and after G-force data. In this way we are not slaves to the track or dyno for determining performance improvements.