problems with "Injector Evolution" article

I just read the "Injector Evolution" article in the Feb 2013 issue of GRM, and see some major flaws in the test. I'll go over my concerns one-by-one. The biggest issues first... then the smaller gripes.

1. The claim that the stock injectors were maxed out was false. I've personally dyno tuned multiple normally aspirated 1.6 miatas on stock injectors that have made over 130 rear wheel horsepower. I have charts from many different dynos around the country of 1.6 miatas making 120-130hp on stock injectors. And as many miatas as GRM has dyno'd over there years... I'm sure they do to!! Heck, 1.6 Spec Miatas make what... 125 to the wheels? I think it's pretty safe to say that the miata in the test did not run lean because it maxed out the injectors. So, the car had some other problem that was causing it to run lean. By moving to the bigger 1.8 injectors they simply patched/masked the other problem, restoring performance.

2. The scientific method was not followed. In order for it to be a viable experiment, only one thing should change inbetween tests. In the experiment, GRM wanted to show that newer tech injectors could outperform old tech injectors. But in the tests, GRM changed injectors that were both newer tech (the thing being tested) and a different CFM. They should have not used injectors that had a different CFM. If there was no other option, then they needed to at least retune and have a COMMON AFR between tests. The test is worthless if the AFR is different between runs.

3. Very little detail was given on the other modifications that were previously done to the car, or how GRM went about "tuning" the car. Were they tuning the AFM spring, or were they using an ECU, etc?

4. We need AFR curves before & after.

To sum it up... we basically have an article about someone with a broken miata, who accidentally fixed it by putting on bigger injectors.

All that said, I really would like to see a legitimate comparison between old and new tech injectors. They mentioned that they've tried something similar in the past with a Ford Escort. Was there an article?

Justin

I'm sorry you disagree with the idea that the injectors might have been maxed out. That was something I came up with when Alan called after doing the tests, it was a best guess based on what he could tell me. I'm always a little skeptical about Spec Miata dynos myself, they seem to read higher than our own dyno does. So it was a theory, but one that was never tested.

With a little more time to think about it, I've got a 1.6 that was making about 140 rwhp on our dyno on stock injectors. 160 peaks a set of 1.6 injectors with an extra 6 psi of fuel pressure.

Having not read the article yet did the test the fuel pressure under load? I wonder if the pump was going weak..

In reply to JustinHoMi:

Haven't read the article, but suspect that a lot of injector changes are not actually needed, just better understood.

But, about the "scientific method"- while one factor testing is valid, being done correctly, multiple factor change testing is also valid. It's part of the "design of experiments" method, where you use testing to understand what changes cause what effects. It works quite well, if the procedeure is done correctly. It's especially good when you are doing a lot of stuff that needs changed. So one factor "scientific method' isn't required. (although, I'm sure in this case, a DOE was not actually followed, so the issue is correct).

Taguchi test methods work (array analysis of multiple changed variables), but induce noise into the data.

To see if the injectors were actually maxed out, one could put a scope on the injector circuit and check the duty cycle. <100%? Not maxed out.

i'm easily influenced. the article inspired me to buy a set of injectors from an '07 miata for my '94.

You too, eh?

I've got two sets of NB injectors on the way with a view of getting them cleaned and flow matched.

Dr. Hess wrote: To see if the injectors were actually maxed out, one could put a scope on the injector circuit and check the duty cycle. <100%? Not maxed out.

I put a scope on my injectors and the signal was so noisy it was hard to isolate the injector pulses. The scope was picking up all the ignition noise and other electrical noise that cars are infamous for. Do you know what I was doing wrong? How should I set up the scope? (sweep time, scale, etc)

Sadly I used to be an electronics tech - hence owning the scope - but just couldn't get a clean signal out of the injector pulse.

In reply to dculberson:

Shielded Test leads?

Lots of comments from people who haven't read the article :)

One thing I did find interesting was the difference in the tested injector sizes. The tests I have here at my desk show a 1.6 injector as 230cc and a 1999-00 injector as 238cc*. That makes the newer one pretty close to a direct drop-in, and that's what I told Alan when he first called and asked about doing this test. I was surprised to see that my numbers were so different than his tests, 210cc and 250cc.

The Ford that was mentioned in the article was an earlier car of Alan's, before he moved to GRM. At least, that's what I remember from our discussion.

This was a one-time test. The runs were made, then Alan tried to explain the results. Of course, the time on the dyno was limited. Ideally, there would have been a complete investigation of exactly what was going on. But there wasn't, probably due to the fact that GRM writers are trying to put out two magazines and not answer every question on the internet. So, should GRM have just abandoned the article? I don't think so. At the very least, it illustrated why you might want to get your injectors clean!

• Why would the 1.8 NB unit be so close to the 1.6 part in size? The NB runs at a higher fuel pressure.

In reply to dculberson:

Besides shielded test leads, I would think that turning the gain way down would help too, and the sweep time would have to be pretty slow. You can do teh maths on 50% duty cycle and see where to kinda start. Just listen to the injectors with a screwdriver or mechanic's stethoscope to get an idea of how often they are firing. Actually, the mechanic's stethoscope would give you a pretty good idea of the duty cycle as well, but might be kinda hard to do when the thing is strapped down on the dyno WFO.

You could also tap into the injector signal at the ECU, which would probably be quieter than at the injector under the hood.

Dr. Hess wrote: In reply to dculberson: You could also tap into the injector signal at the ECU, which would probably be quieter than at the injector under the hood.

this^^^

My test leads are shielded but the car's wiring obviously isn't. I'll have to try the ECU end suggestion, that might help.