Why Low-Pressure GDI Injector Testing Can Give You the Wrong Answer
Can a gasoline direct injector look bad at low pressure — but actually be failing in the opposite direction when tested correctly?
Absolutely.
We recently had a set of four GDI injectors from a Subaru BRZ that provided an unusually clear demonstration of why testing conditions matter.
We tested the same injectors two very different ways.
The results tell an important story about GDI injector testing.
Photo on Left: Tested at 8 Bar on a Port Flow Bench
The photo above shows all four injectors after flow testing at 8 bar on a conventional port-injector flow bench.
The injectors are numbered #1 through #4 from left to right.
Look at the fuel levels in the graduated cylinders.
Injector #1 clearly appears to be flowing less fuel.
Injectors #2, #3 and #4 appear relatively well matched.
If this were the only information available, the obvious conclusion would be that #1 is the weak injector.
That conclusion would be wrong.
Photo on Right: Injector #1 Tested Against the Correct GDI Test Plan at Operating Pressure
We then tested these injectors under the appropriate high-pressure conditions and compared their performance against the test plan for this injector.
Injector #1 was tested at multiple operating points.
At 180 bar, the acceptable range was 75.8–86.4. Injector #1 measured 90.3.
FAIL — HIGH
At 140 bar, the acceptable range was 26.4–29.7. Injector #1 measured 33.5.
FAIL — HIGH
At 65 bar, the acceptable range was 27.9–31.3. Injector #1 measured 33.7.
FAIL — HIGH
Think about that result for a moment.
At 8 bar, injector #1 visually appears to be the injector delivering too little fuel.
When tested according to the proper GDI test plan, that same injector delivers too much fuel at all three test points.
The apparent diagnosis didn't just become more accurate.
It reversed.
What About the Other Three Injectors?
This is where the example becomes even more interesting.
At 8 bar, injectors #2, #3 and #4 appear reasonably well matched.
But when all four injectors were evaluated against the proper GDI test plan, the results were:
Injector #1 — FAIL HIGH
Injector #2 — FAIL HIGH
Injector #3 — PASS
Injector #4 — FAIL HIGH
Injector #3 was the only injector in the set that passed, and its results were near the middle of the acceptable tolerance.
So the low-pressure test gives us two misleading impressions.
It makes #1 appear to be the low-flow injector when it actually fails high under the correct test conditions.
And it makes #2, #3 and #4 appear to be a reasonably matched group even though two of those three injectors fail the actual specification.
Flow Matching Does Not Mean the Injectors Are Good
This distinction is extremely important.
A set of injectors can be closely matched to one another and still be wrong.
Imagine four injectors that are all delivering 10% too much fuel. They could have an excellent percentage match when compared with each other.
They're still four incorrect injectors.
That's why we don't believe GDI testing should stop with:
“How closely do these injectors match?”
The more important question is:
“Does each injector deliver what it is supposed to deliver?”
To answer that question, you need a specification.
High Pressure Alone Isn't the Answer Either
There are now injector benches capable of operating at substantially higher pressures than traditional port-injector equipment.
That's an improvement, but pressure capability by itself doesn't make a GDI test valid.
Suppose a machine tells you an injector delivered a particular volume at 100 bar.
Is that good?
Is it low?
Is it high?
Without knowing what that specific injector is supposed to deliver under those exact test conditions, the number has limited diagnostic value.
A meaningful GDI test requires the appropriate combination of operating pressure, injector command conditions and a valid specification against which the result can be evaluated.
That's why The Injector Shop uses part-number-specific test plans for the GDI injectors we support.
This Is Why We Test GDI Differently
Traditional port fuel injectors operate at relatively low fuel pressures. GDI injectors operate in a much different environment and must precisely meter fuel at high pressures and during very short injection events.
Testing one at 8 bar can certainly make fuel come out of it.
It can even produce a repeatable number.
But this Subaru BRZ set demonstrates why that number should not automatically be interpreted as the injector's actual operating condition.
The injector that appeared low at 8 bar was actually high at 65, 140 and 180 bar.
The three injectors that appeared reasonably well matched at 8 bar contained two failures and only one good injector.
The Bottom Line
There is a big difference between measuring flow and testing an injector.
For meaningful GDI diagnosis, we want to know more than whether one injector flows more or less than another.
We want to know whether each injector performs correctly under the conditions it was designed to operate in and whether its output falls within the proper specification.
This Subaru BRZ set is a nearly perfect demonstration:
At 8 bar, #1 looks low.
At the proper GDI test conditions, #1 is high.
And of the four injectors, only #3 actually passes the OEM test plan.
That's why the test pressure matters.
That's why the test plan matters.
And that's why a set of graduated cylinders alone can't tell the whole story.
