Honda Mileage:VID Update July 12

Since my last post (over a week ago) I've been busy with a number of things . But, even while I rebuilt our front steps--- and other duties, I still found a time to do some thinking about modifying the VID design to handle both "saturated" and "peak and hold" type injectors.

This morning I completed changes to the prototype board addressing this---and other issues. While, I am currently testing the changes, I have hopes that the VID can now be user programmed to handle any type of fuel injector, or fuel injector system.

As the readers may recall, I originally configured my YAM "yet another (meter)" to handle only saturated type fuel injectors. After researching the issue, I discovered that other meters/devices offered for this purpose, also do not claim to be able to handle anything but saturated type fuel injectors.

So, that got me thinking: I surmised that in order to handle the saturated type, the input threshold could simply be statically assigned to one voltage---and that would be good enough. (and experience proves this to be correct) However, because of the way peak-and-hold fuel injectors work, assigning a unitary voltage to work with saturated types almost certainly would not work (correctly) for peak and hold types.

So, I took a look at the specification for the PIC 18F2520 CPU and found the device could be configured to provide a (16 step) variable reference voltage output at one of its analog pins.
BTW--the 18F2520 also has two internal linear comparators--but they do not offer the kind of flexibly i needed--so, I use my own.) However, since the range of this voltage reference is limited to a maximum of about 3.3 volts (using a Vcc of 5 volts) --- the injector voltage would also have to be scaled down to match this voltage range.

By installing the appropriate voltage divider on the INJ input, and rewiring the comparator to accept the reference voltage provided by the CPU, the program can now adjust the INJ reference voltage on the fly. With a maximum of 14.5 (high alternator output) voltage at the INJ input, the scaled output reads about 3.3 volts. The processor is now able set the comparator threshold to any value the range of 0 to 3.3 volts. With the 4.4 gain factor this represents a 0 to 14.5 volt adjustment range.

What does this mean? It means that unlike when I used fixed zener diode to provide a constant/stable reference voltage, the user can now set one threshold for the fuel injector current rise time and anther for the fall time. This is most important for the peak and hold type fuel injectors. Because P&H types employ a two step fall time. Obviously, if we did not adjust the threshold to accommodate the two-step-nature of S&H injectors, we might get a very rosy (but inaccurate) impression of the amount of fuel they were allowing to pass.. However, in this way, we will be able to detect exactly the correct pulse for both types of fuel injectors.

I realize the wiring in most cars is designed to allow the ECU to drive one end of each fuel injector to ground while the other is held at battery voltage. I also understand that some cars tie one end of the peak-and-hold type injector to ground, while using the ECU to drive the other end to battery voltage. In either case, the (newly redesigned) VID will accommodate either voltage polarity. In rare cases, certain jumpers would need to reconfigure/switched. As an added benefit, the new design allows the CPU to monitor the analog voltage level on the INJ input, this means that it is (theoretically) capable of automatically setting the appropriate threshold levels itself.

In summary, the changes I made to the VID design this morning allow the user (via the host software) to configure the INJ thresholds levels to match (almost) any requirement. For instance, if the vehicle uses situated fuel injectors, the user might simply enter the number representing 6 volts for both values--- and be done. (Better, yet, the device would simply default to 6 volts.) If, on the other hand, the user needed to configure for peak-hold type fuel injector system, he/she might decide to enter 3 volts for the injector charge time value and 10 volts for the injector off time value. In any case, this design allows the user the flexibility to just "make it work."

I mentioned other changes: In this REV, I have added other features to make it possible to do on the fly communications with laptops and PDAs without having to hook up either a serial or USB cable to the VID. I will be releasing more information on this development as my testing continues. But suffice to say, the final design of the device will provide both wired and wireless communication options. For those who still have their handy HP-Pocket PC with Blue tooth, they may finally have found something they are good for.