Henrik

iTrader: (5)

I've got a bunch of injectors lying around, and I want to get them flow tested. Now, obviously, I could send them off and pay someone to measure them, but what's the fun in that?

Instead, I've started developing a little electronics project where I use an Arduino to drive a mosfet, which opens and closes the injector. I've also got a little **** that lets me adjust the "rpm" from 600 to 8000rpm.

Here's the little test bed indoors, with a bread board circuit just driving the injector on is own:

Here's a shot of it hooked up to an impreza fuel rail and injector housing driving a 650cc sard injector:

A 565 blue injector has a totally different pattern:

A decapped 380cc injector (standard blob wrx) actually has a reasonable pattern:


I'm waiting on a graduated measuring cylinder so that I can see how much they spray.

Henrik

iTrader: (5)

got the cylinder, and bodged a test setup.

Obviously, it's not the completed item (given that it's a couple of bits of wood held together with some screws...), but still interesting to play with.

I only had time to mess around with the decapped injectors I have, but the flow pattern looks better than the Sard injectors for sure. They also flow pretty close to each other.



My tests were done at 3 bar, and I pulse the injectors like: 50ms ON, 150ms OFF, 50ms ON, 150ms OFF, etc etc. This makes the injector open as many times per minute as at 600rpm, but it also flows 100% for the 50ms, which is way too long for idle, of course. I.e. what I'm trying to say is that whilst the data below is accurate for my comparison of my injectors, it doesn't mean that the _actual_ flow rate as measured by a proper rig (at 100% open time?) will be the same numbers.



injector flow per minute CC
injector 1 588
injector 2 576
injector 3 564
injector 4 556


mean flow: 571 cc / min


Difference against mean:

injector 1: +3.0%
injector 2: +0.88%
injector 3: -1.2%
injector 4: -2.6%

BrownPantsRacing

iTrader: (6)

This is great and very interesting, I love it.

Thanks for sharing your testing.

Are you going to do some pattern tests? I'd love to see how a de-capped injector pattern compares to standard uprated capped version.

Henrik

iTrader: (5)

If you compare the last two videos in the first post, one is a blue STI injector, the other a decapped wrx injector, but I appreciate the quality is a bit rubbish of those videos, and the angle is weird etc etc, so will do some better ones.

having said that, the spray pattern of the decapped ones are MUCH better than eg the sard ones (same pattern as power enterprise I think), which is basically two fat jets of droplets... No wonder PE ones have a bit of a reputation for poor idle (from reading online).

Henrik

iTrader: (5)

BrownPantsRacing: Here's the spray pattern... should have really used a black background, but this is all I had to hand

BrownPantsRacing

iTrader: (6)

To be honest it looks pretty good.

Henrik

iTrader: (5)

I also have a set of blue STI injectors l, so thought I'd measure them as well to make sure they were OK, and I discovered that one injector is broken... It will run fine for about 5 seconds, then stutter and completely pause for another second before starting again!



Also, and here's another "Subaru wisdom" that I'd like to dispel: the injectors do NOT flow 565cc/min, as the common ok knowledge goes. They in fact seem to flow 505-510cc!!
No wonder my car was running lean on the way to the MOT (afr 16-17), because I took out some pinks, which I believe flow the advertised 550cc (I haven't measured, I sold mine many years ago). Luckily it was only a gentle pootle to the MOT place, so hopefully I will have gotten away with it.

However, I think there is a lesson here about accepting common gospel.


Separately I've also decapped some more injectors... I now have 10 in total (4 were sold to me as lateral injectors, but who knows where they actually came from).
The process I've taken is to strap the injector up in the lathe and grind the cap collar off with a cut off wheel on a grinder. It looks a lot better than the free hand mods, of course... Think I might even make a jig for them so I don't have to spend so much time ensuring the injectors run true.

For remap I'm going to go with in 4 decapped matched injectors. Three of them flow 771cc, and the fourth 760cc (1.4% diff). I'll also send along the sard ones which all flow exactly the same 641cc/min so that the mapper can change them if he wants.

Will be interesting to see what happens! Maybe they'll all be rubbish and I'll have to get some id1000s like everyone else, but whats the fun in that?

BrownPantsRacing

iTrader: (6)

Nice work. Good that the tests have managed to highlight a faulty injector and very interesting that the blues don't flow as well as predicted.

Henrik

iTrader: (5)

I know I'm raising my old thread, but just thought I'd link it to the updates.

Posts 23 and 24 in this thread https://www.scoobynet.com/projects-4...th-a-td04.html contains an updated and much improved version of the injector tester.