Injector Testing

[SORNagain]

Hi folks,

Finally got round to testing the injectors I plan to use for a megasquirt install. It turned out to be interesting, so I thought I'd share.

One thing I forgot to mention if you are going to try this: This uses a fair volume of fuel and produces loads of vapour, so it's worth ensuring there can be no sparks (i.e. no dodgy connections), work in a well ventilated area, and have a decent extinguisher handy should the unthinkable happen. I even set my phone to flight mode, but that's probably just paranoia... Obvious I know, but best say it anyway.

[FridgeFreezer]

Neat!

What frequency & duty cycle are you running the injector at? It looked quite slow to me.

Nice to see an MSP430 rather than Arduino too.

For those without a driver board, if you've got a MegaStimulator you can use that with a couple of wires soldered on for the injector to run it direct from the MegaSquirt itself.

[SORNagain]

I love a good MSP! Yeah real slow, so I could observe the spray pattern. Oddly enough the launch pad is running the code it was shipped with; when I first plugged it in it flashed the LED's at about the same rate as I wanted so I just left it. All I know is that when I press one of the buttons it stops and does something else - dunno what - and I reset to start it again! Looks about 1Hz at 50% duty to me.

For the flow measurement I wasn't sure whether I was going to run the injectors for a certain time & measure the volume, or measure a certain volume and count the pulses required to get there. I've since dug out a glass flask with a narrow neck that can measure 1l +/-0.4ml so I think I'll be doing the latter.

I want to pulse the injectors to eliminate the time for opening & closing rather than measuring the continuous flow. It will make the timing easier too. Any idea what a ballpark pulse width at WOT will be?

[FridgeFreezer]

General rule is pulse width duty cycle shouldn't be going much above 80% at WOT, over that it generally means you need bigger injectors.

[Chazza]

One thing I forgot to mention if you are going to try this: This uses a fair volume of fuel and produces loads of vapour, so it's worth ensuring there can be no sparks (i.e. no dodgy connections), work in a well ventilated area, and have a decent extinguisher handy should the unthinkable happen. I even set my phone to flight mode, but that's probably just paranoia... Obvious I know, but best say it anyway.

Cool! Thank you for posting.

For not a great deal more investment in equipment, I think you could probably make the test-rig much safer. Petrol has five-times the explosive power of dynamite when mixed with air in the correct proportions, so a fire extinguisher won't be much use, because if it does ignite you will be dealing with a gas explosion not a fire! Good bye garage; house; car and owner!

A friend of mine use to own an injector tester and cleaning machine - it solved the explosion problem by sitting the injectors at the top of a graduated glass cylinder; there was not enough air in the cylinder nor an ignition source, with the added benefit that the operator didn't have to breathe the fumes.

Well done on cleaning the injectors; Redex is a very good cleaner, as are Nulon and Cost Effective Maintenance injector cleaners. My friend's machine used a solvent combined with ultra-sound to dislodge the muck. Land Rover use a solvent which is in a cannister and connects to the fuel line near the tank, which fixed my car once,

Cheers Charlie

[SORNagain]

General rule is pulse width duty cycle shouldn't be going much above 80% at WOT, over that it generally means you need bigger injectors.

I think I'll just work out what the pulse width at 80% & 5000rpm is and use that. I'm not sure if it will matter anyway; I'm only after a comparative measurement between the injectors.

For not a great deal more investment in equipment, I think you could probably make the test-rig much safer. Petrol has five-times the explosive power of dynamite when mixed with air in the correct proportions, so a fire extinguisher won't be much use, because if it does ignite you will be dealing with a gas explosion not a fire! Good bye garage; house; car and owner!

Yikes! I'm not sure I want to go near it again! Now I've verified the spray pattern though, I don't think I'll have to run it like that anymore. I reckon I had over a litre 'float away' during the testing...

Interestingly the fuel collected in the tray had a very milky apprearance. It didn't settle out if left, but did clear when it was recycled back into the tank and the pump started. At first I thought this was contamination getting caught by the filter, but later realised it could be water absorbed from the air. Something about the way the fuel was circulating seemed to separate the water again, as I could see some drops in the bottom of the tank. Not completely sure it wasn't water from the jerry can though. I should be able to test this by having a smaller closed volume as you describe.

It's been a real surprise to me how good the injector cleaner was, and how ineffective throttle body cleaner proved to be. I've been joking about Reddex for years - homeopathy for cars - I guess now I'd better tuck in to the humble pie!

[Chazza]

I think I'll just work out what the pulse width at 80% & 5000rpm is and use that. I'm not sure if it will matter anyway; I'm only after a comparative measurement between the injectors.

I should be able to test this by having a smaller closed volume as you describe.

It's been a real surprise to me how good the injector cleaner was, and how ineffective throttle body cleaner proved to be. I've been joking about Reddex for years - homeopathy for cars - I guess now I'd better tuck in to the humble pie!

I have enjoyed your research, especially about the Redex, but also how you managed to clean your injectors so well!

Thinking about my friend's machine, which he sold some years ago, I remember that it had the capacity to clean and test four injectors at a time; when the test phase was operating, the flow could be measured over a given time and measured by the gradations on the glass cylinders. I am not sure what was deemed a failure when compared to the other injectors, but I can find out if you are interested. For home use, one injector at a time is adequate. I don't know what the cleaning and test-fluid was, but I don't think it was petrol and it may have been safer to use.

The system you have devised is brilliantly simple and a fraction of the price of my friend's machine. To modify yours to a safe level would I think, only involve the following:

* Buy a graduated cylinder from a laboratory supplies merchant, large enough so that the test liquid isn't affected by rising air pressure inside the cylinder.

* Firmly mount the cylinder in a frame of some sort so that it can't topple over.

* Fix a bung in the top of the cylinder to contain the vapour, with a small bleed hose to return vapour to your fuel tank. The bung could have a suitable opening for the injector to push into. The merchant might have suitable bungs, or you could make one with some Sugru, if it can withstand petrol.

* It might be a good idea to make your fuel tank safer as well by using a steel can, which is easy enough to make. I once spilt a large volume of petrol on my groin and the pain lasted for seven hours!

Cheers Charlie

[FridgeFreezer]

No need to run flat out, 50% duty cycle at whatever speed will be fine if you're just comparing injectors to each other rather than trying to make a specific measurement.

Redex is good stuff, I suspect there's not much to it other than some more-volatile-than-average hydrocarbons. We did a sort-of-scientific test back in the day running a karting centre (30+ identical Honda karts), we had computer lap timing and could reliably measure a significant reduction in lap times whenever we ran redex through the karts. They spent a lot of time idling in the pits so stuff would gum up fairly reliably.