Topic: We built a LeMons-grade fuel injector tester/cleaner
I'm normally not the build journal / photo documentation type, but I thought this project was cool enough to share. It started as the subject of a random conversation between one of my Lemons teammates and I. Something like: "Hey, wouldn't it be cool if we could test and clean fuel injectors outside of a car?" He's an electrical engineer, and I'm a computer engineer, so we made a deal: if he gave me a driver circuit and built the test stand, I would program a microcontroller and build a board to make it all work. Well, after an order from Digi-key, a couple nights of programming, an afternoon of soldering, and a trip to the junkyard, here it is:
The fuel rail is from a later Quad 4 (Sunfire GT, I think). We were dumb and used a fuel pump from a TBI Chevy truck, so it only made 1 bar (14.5psi), but that was good enough for a trial run.
The board is hiding behind the battery in the picture above; here's a close up shot. It's basically a 5v power supply, a FET driver circuit, and a PIC18F microcontroller I programmed in C. I scavenged a 0-7 binary switch from an old SCSI CD-ROM drive and am using that to select test routines on the microcontroller. It has a couple cleaning and flow testing modes, can ramp frequency and duty cycle up and down, etc. One mode simulates a 15-second quarter mile pass through three gears, which is fun to watch 
Maybe manual control knobs and an LCD display will be in version 2.0.
Despite running at 1/3 of normal fuel injection pressure, we were able to check out the spray patterns and flow rates of several injectors. I proved that the set of injectors I bought off some random guy on a forum for $25 not only worked, but all of them actually flowed pretty close to the same. They were dry when I got them, and a couple of them leaked at first, but after a little exercise with a mixture of gasoline and fuel injector cleaner, they looked good to go.
We were surprised to see that all of the different kinds of used OEM fuel injectors we tested had narrow, jet-like spray patterns. My friend had some brand new 60 lb/hr aftermarket injectors, so we compared those side by side with my mystery forum injectors. His (on the right) produced a wide, finely atomized puff of fuel, compared to the jet that came out of mine. Hmph, maybe mine aren't that great...
By weighing the bottles after running for a while, and using his 60#s as a known value, we were able to calculate the flow rate of my injectors. Surprisingly, they were flowing dead-on what they're supposed to, so I guess the wider spray pattern is just a feature of his aftermarket injectors.
Here's the longest video I took of it in action. It's a one minute ramp from 5-90% duty cycle, from 7Hz (idle) to 50Hz (6000RPM). I'm not sure what's with the hiccup at 20s. I don't remember seeing or hearing it, so I think it's probably the camera's fault, but who knows, might be a little quirk in my PWM code...
As far as safety goes, we certainly could have done better (but we didn't have any judges around to keep us in line). The catch bottles fit snugly on the fuel injector o-rings and didn't leak fuel, but the fumes from the reservoir were probably bad enough to give us both cancer and/or brain damage. I think that at the very least, we'll be switching to something with a flashpoint above room temperature, maybe kerosene. Apparently the pros use a special non-flammable fuel injector test fluid that has about the same viscosity as gasoline. If anyone has a line on where to get some of that stuff, let me know...
