In the first part of this tech special, we got a bit of a primer on semiconductors, fuel injectors and pulse width modification. Now check out the testing process and see what we found.
Since the primary purpose of these tests was to quantify the individual effect of voltage changes on specific system components, we used two power supplies. With one connected to the PCM and the other to the fuel pump, we were able to alter voltage as necessary for each test group.
To simulate in-car conditions as closely as possible, our test fixture incorporated a signal-generating box with the outputs needed to convince the PCM it was connected to a living, breathing engine. Among other things, “the box” allowed us to alter rpm, manifold pressure and throttle position. In turn, these adjustments provided the PCM with the inputs required for it to change commanded pulse width in response to any simulated engine operating condition. Adequate fuel flow is most critical when an engine is spinning at high rpm and wide-open throttle, so we set the controls to simulate 6,000 rpm and 100 percent throttle position.
We verified those settings by firing up EFILive’s FlashScan program on a laptop PC, connecting it to the PCM and monitoring (and recording) each test on the virtual dashboard. The tune portion of FlashScan allowed us to program the PCM as necessary for the requirements of specific test sequences.
During the first group of tests, we supplied the same voltage to both the PCM and fuel pump. And although we varied it from 14.2 to l0.88 volts, discharge volume stayed constant at 64ccs for each 15-second test run. That clearly demonstrates the effectiveness of the pulse width voltage adjustment table, which enables the PCM to lengthen pulse width as system voltage drops (or increase it as voltage increases).
For the second group of tests, we held PCM/injector voltage at 13.7 and varied fuel pump voltage in steps from 9.5 to 15.5. We also reprogrammed the PCM to eliminate voltage-related pulse width adjustment. Consequently, any change in discharge volume would strictly be the result of variations in fuel pump voltage. As the accompanying chart shows, discharge volumes ranged from 58.5cc to 64cc (for 15 seconds). Converted to pounds per hour, the range is from 22.88 to 25.04. That’s well below the 27.77 pound per hour rate we recorded when testing the injectors for maximum flow rate.
During the third test series, we held fuel pump voltage at 13.7 and varied PCM/injector voltage from 10 to 14.2. Although discharge volume was definitely affected by voltage, it didn’t change as much as it did when fuel pump voltage was varied. With PCM/injector voltage varying between 10 and 14.2, discharge volume ranged from 60.5 to 63cc (23.67 to 24.64 pounds per hour).
At first glance, it seems that our tests simply proved the obvious: a drop in voltage equates to a reduction in discharge volume. They certainly did that, but more importantly they quantified the effects of voltage variation. A 1-volt variation at the fuel pump equates to a .4 pound per hour change in fuel discharge volume. That may not seem significant, but if fuel delivery volume is marginal at normal voltage levels, even a slight drop can lean out the air/fuel mixture enough to cause problems ranging from reduced power to engine damage. Also worthy of consideration is the fact that these tests were run with LS1 injectors, which are rated at 28 pounds per hour. With larger injectors, differences tend to be more dramatic.
In practical terms, data relating to injector output versus system voltage is primarily relevant to highly specialized installations. Unless a vehicle has been rewired, as is typical of legitimate race cars (and some illegitimate street cars), chances are slim that voltage at the injectors will be substantially different from that at the PCM. On the other hand, differences in voltage at the fuel pump and at the PCM aren’t all that uncommon. That’s good information to keep in mind when you’re looking for fuel system gremlins and can’t find any.
Text and Photos by Dave Emanuel