12 Hole Injector Upgrade - Finally Tested

[2001LC]

Using good gas always. Along with fuel system cleaner products like 44k, Chevron Techron, Seafoam, etc. occasionally. As well as the on-board cleaning systems, like Dealerships offer. Along with driving regularly. Are all good practices to keep fuel injectors happy.

But the best way to check for excessive leak-down and poor spray pattern. Is having fuel injectors tested, cleaned and rebuild.

When to Clean or Replace fuel injectors

Please post your experiences and thoughts on the subject. I’ve been considering cleaning or even replacing all 8 fuel injectors, is this wasting time and money. I understand keeping fuel injector clean by adding a can of additive to the gas tank, now & then, is a good practice. This keeps them...

forum.ih8mud.com

 

[2001LC]

is there any advantage to buying new injectors vs having them cleaned/ serviced? can an injector be beyond service or fail in such a way that it can't be repaired?

Cost!

 

[Brandon Ryder]

Cost!

my question is why wouldn't i clean them with this service? cost and convienience aside, will new injectors be better or last longer or can original injectors be infinitely rebuilt?

 

[suprarx7nut]

is there any advantage to buying new injectors vs having them cleaned/ serviced? can an injector be beyond service or fail in such a way that it can't be repaired?

Short answer: no.

The life of OEM-grade injectors is generally well beyond the life of the car (supposedly 1 billion cycles which is around 500k miles if my math is right). With the 100 series lasting longer than a normal car, maybe the injectors start to become a wear item, but I think that's a bit of a stretch.

my question is why wouldn't i clean them with this service? cost and convienience aside, will new injectors be better or last longer or can original injectors be infinitely rebuilt?

New OEM-grade injectors are $700-1500 for a set. When people talk about buying "new" injectors I believe 99% of the time, they are talking about buying used, re-manufactured or off-brand injectors. None of those options are notably better than just having your existing injectors cleaned. Some of those options will be worse.

My input is that you should buy high end injectors (plan on $750+ for a V8) for a race car or have your OEM injectors cleaned by a reputable shop for all other uses.

 

[2001LC]

New is always nice, but not necessary. The Million mile Toyota 2UZ, had originals.

Have you looked at list price of just one injector! For the 2006-07 2UZ-fe VVT. OEM Toyota Denso list is $198. Denso ~$100.

Sometimes FIS here in Denver has found some bad or marginal, in sets I've had them service. It's why I use them. Then I know!

Excessive leaked down, can cause oil wash of cylinders. Making start-up, hard on cylinder walls and rings.
Poor spray pattern, reduce performance and drops MPG.

 

[dirtydeeds]

With the help of the ih8mud community, we were finally able to test the 12 hole injector upgrade. See the video below. Huge thanks to @RiverRatMatt @BuckeyeFan and @J1000 for stepping up to the plate and ponying up with contributions.

As far as I can tell we were able to produce the most complete set of data related to this topic to date. I've still got all the logs and dyno charts so we can dive into anything else people want to see in this thread. I'm happy to share the data via Google Drive as well. I'll try to get that setup soon.

HINT: There's a bonus at the end that's going to ruffle some feathers. It relates to engine air filters...

How does the truck drive on the street with 12hole injectors Versus 4 hole?

 

[dirtydeeds]

Did any of this ‘science’ include driveability testing on the street? Or was it all to see if there was minimal gains in Hp?

Shame that actual money had to be spent on this. Some folks just don't want to believe simple science and physics. Definitely Mythbusters-level quality of effort. Good job!

 

[dirtydeeds]

Oh absolutely, I'm counting on it. That thumbnail of the video is basically aimed at those people. If you believe there's a difference "show me your data".

I'm hoping the video garners some controversy and boosts it on Youtube's algorithms. I've been painfully close to the monitization limit for a couple years. I'm hoping this video finally tips over that edge and starts a tiny trickle of passive income so I can justify doing more fun car stuff for business.

At least we’ve found the purpose of this one sided video

 

[suprarx7nut]

Did any of this ‘science’ include driveability testing on the street? Or was it all to see if there was minimal gains in Hp?

Our drive-ability test of sorts was the partial throttle, loaded dyno roller test. That was the best method we (and the professional tuning shop) could determine to test closed loop performance and provide objective (numerical) results.

We've got 30,000 rows of datalogs and professional analytics software we can comb to search for any subtle differences in engine response. I wasn't able to find them, but I'm happy to run some analytics on any hypothesis you can suggest.

The following was plotted for difference in the two populations:

1. Fuel flow rate per power
2. Fuel flow rate per torque
3. Ignition timing
4. o2 sensor variance between commanded and measured AFR

At least we’ve found the purpose of this one sided video

I haven't been shy regarding my skepticism on this topic. That doesn't preclude me from maintaining scientific integrity during the test or drafting its results summary. Myself and 3 others had varying hypothesis on this topic and were interested in objective results and an unbiased scientific method.

As the thumbnail indicates if you've got a counter data set, by all means, "whip out your data." We're interested in the truth here, not proving anyone's preconceived opinion.

 

[dirtydeeds]

Our drive-ability test of sorts was the partial throttle, loaded dyno roller test. That was the best method we (and the professional tuning shop) could determine to test closed loop performance and provide objective (numerical) results.

We've got 30,000 rows of datalogs and professional analytics software we can comb to search for any subtle differences in engine response. I wasn't able to find them, but I'm happy to run some analytics on any hypothesis you can suggest.

The following was plotted for difference in the two populations:

1. Fuel flow rate per power
2. Fuel flow rate per torque
3. Ignition timing
4. o2 sensor variance between commanded and measured AFR

I haven't been shy regarding my skepticism on this topic. That doesn't preclude me from maintaining scientific integrity during the test or drafting its results summary. Myself and 3 others had varying hypothesis on this topic and were interested in objective results and an unbiased scientific method.

As the thumbnail indicates if you've got a counter data set, by all means, "whip out your data." We're interested in the truth here, not proving anyone's preconceived opinion.

The point of 12 hole injectors is, and has always been, for better on road driveability.
None of your testing included that since your test vehicle was strapped down to a dyno and was stationary. Also, our trucks have an adaptive memory ecu. If your before and after test was back to back without giving the ecu time to adapt to the new mod, it was inconclusive.

However, none of your data surprises me and I’ve installed several sets of 12hole injectors on several trucks. I’ve also road tested said trucks with great detail. After road testing, my conclusion was that there was no power gains from simply installing 12 hole injectors. However, the driveability was much better and that they are well worth the small investment.

I also have feedback from literally hundreds of 2uz owners running 12 hole injectors. Better driveability isn’t an opinion. It’s either better or it’s not. You’ll never see somebody with 12 hole injectors say ‘dang, these suck, I’m putting my 4 hole injectors back on”. At least I haven’t ever heard anything like this after talking with literally hundreds of owners of 12 hole injectors.

In other words your testing may prove something that I’ve told people to be true since I first discovered the option to install 12 hole injectors back a couple of years ago, there is likely no power gains.

What is unfortunate is that your testing is incomplete and people will see it and think that it conclusively rules out this great mod as being worth it. While there is no ‘data’ showing more power. The fact that your truck downshifts less, stays in a higher gear more of the time, has better driveability, is more responsive, and just overall picks up the way the truck drives, makes them a very worthwhile mod.

The differences noted above are subtle, but if you’re a ‘driver’ you will notice them.

BTW, I agree with your research regarding air filters. I doubt they lose power, if you had given the ecu time to adapt rather than doing back to back tests on a dyno you’d have found that either filter made the same power. The benefit of a washable air filter is that you never ever have to buy another air filter. If you’ll notice, I have been posting on tundra forums since 2003. Never once have I said that an air filter or cold air intake makes power or is worth buying (cold air intake). Matter of fact I don’t post air filters or cold air intakes on my site. But I absolutely recommend 12 hole fuel injectors for your Land Cruiser or Tundra with 2uzfe

 

[suprarx7nut]

How might you define improved driveability in terms of a measurable parameter? If it "feels" better, what are you feeling? To me, that means increased engine efficiency (power per fuel, power per torque) or increased absolute power or torque. Most notably at partial throttle and lower in the rev range. I think we effectively concluded we couldn't measure any differences that significantly favored the 12 hole injectors on those metrics. If you've got a different metric in mind, let us know what it is so we can plot it from our data set. We logged everything we could think of.

ECU memory is a good point. We did pull ECU fuse during one or two of the injector swaps (Jimmy might recall better than me when we did or did not do that). We did not pull the car off the dyno so the re-learn period was covered while on the dyno idling and run at low RPM to speed up warm up. We did not time the learn period, but did ensure the engine was fully warmed again. I expect that if there were significant impacts of the ECU not having learned the new injectors that we would find that in the data somewhere. Admittedly, I don't know exactly what parameter would highlight that, but I'd expect to see some sort of population variation in AFR variations, timing, power, torque, etc... There were no significant population drifts I could find in any of those - Keep in mind the tests were run on a car with 12 hole injectors normally installed. The first batch of runs was our baseline with 12 hole injectors which had been in place for months (years?) and an ECU that was fully "learned". If anything the 4-hole tests were at a disadvantage/risk of being run without a learned ECU.

Anecdotes from customers is tough because while you might be able to find customers that praise the 12-hole upgrade, we can find an army of similar people exclaiming the benefits of the K&N filters (and open element filters) as significant power adders. Data doesn't lie, but anecdotes do sometimes so we have to stick to hard numbers here.

 

[Nick T]

@dirtydeeds ,

In the link below, you state that low end torque is very noticeable :

Chips/CAI for better MPG?

Hey guys, 2005 Sequoia SR5 4x4. About to hit a 3,500 mile roadtrip and wondering if anybody has had any significant MPG gains simply by installing a...

www.tundras.com

Inferring there is a gain in low end torque which this testing debunks.
How can driveability be better, downshifts less, stay in gear longer etc etc that you state if the engine is not making more torque aka hp?

 

[dirtydeeds]

@dirtydeeds ,

In the link below, you state that low end torque is very noticeable :

Chips/CAI for better MPG?

Hey guys, 2005 Sequoia SR5 4x4. About to hit a 3,500 mile roadtrip and wondering if anybody has had any significant MPG gains simply by installing a...

www.tundras.com

Inferring there is a gain in low end torque which this testing debunks.
How can driveability be better, downshifts less, stay in gear longer etc etc that you state if the engine is not making more torque aka hp?

The low end torque is better. The testing does not debunk this fact. Even an OEM ROAD TESTS their changes. Only dyno testing is not complete testing.

that said, answer me this, have you personally tested 12 hole injectors against 4 hole injectors? Or are you getting your info solely off the internet?

 

[suprarx7nut]

The low end torque is better. The testing does not debunk this fact. Even an OEM ROAD TESTS their changes. Only dyno testing is not complete testing.

that said, answer me this, have you personally tested 12 hole injectors against 4 hole injectors? Or are you getting your info solely off the internet?

Did you watch the video so you know what tests we ran and how they work mechanically?

 

[dirtydeeds]

How might you define improved driveability in terms of a measurable parameter? If it "feels" better, what are you feeling? To me, that means increased engine efficiency (power per fuel, power per torque) or increased absolute power or torque. Most notably at partial throttle and lower in the rev range. I think we effectively concluded we couldn't measure any differences that significantly favored the 12 hole injectors on those metrics. If you've got a different metric in mind, let us know what it is so we can plot it from our data set. We logged everything we could think of.

ECU memory is a good point. We did pull ECU fuse during one or two of the injector swaps (Jimmy might recall better than me when we did or did not do that). We did not pull the car off the dyno so the re-learn period was covered while on the dyno idling and run at low RPM to speed up warm up. We did not time the learn period, but did ensure the engine was fully warmed again. I expect that if there were significant impacts of the ECU not having learned the new injectors that we would find that in the data somewhere. Admittedly, I don't know exactly what parameter would highlight that, but I'd expect to see some sort of population variation in AFR variations, timing, power, torque, etc... There were no significant population drifts I could find in any of those - Keep in mind the tests were run on a car with 12 hole injectors normally installed. The first batch of runs was our baseline with 12 hole injectors which had been in place for months (years?) and an ECU that was fully "learned". If anything the 4-hole tests were at a disadvantage/risk of being run without a learned ECU.

Anecdotes from customers is tough because while you might be able to find customers that praise the 12-hole upgrade, we can find an army of similar people exclaiming the benefits of the K&N filters (and open element filters) as significant power adders. Data doesn't lie, but anecdotes do sometimes so we have to stick to hard numbers here.

The difference between a random customer anecdote and me, I’ve built or had part of building literally thousands of Toyota’s since 2003.

The way you define improved driveability in a measurable parameter is to put an expert behind the wheel and let him drive the vehicle before and after. That’s exactly what the OEM does after they’re done testing on the dyno. That’s why only testing on the dyno is inconclusive.

Admittedly your test is incomplete. As you’ve stated you did not allow the ecu to fully adapt. That doesn’t happen until about 50 miles of regular driving. It doesn’t even come close to happening just in the time it takes to warm up. The difference in driveability quickly after an ecu hard reset compared to 50 miles after a hard reset, even with the same injectors is big. The truck drives way better after the ecu has adapted than it did before. Then on top of the ecu hard reset you’ve thrown in an entirely different set of injectors as another big variable.

With skewed testing, and a desire to ‘stir the pot’ for views on YouTube, I suggest guys test 12 hole injectors for themselves, on the road, then decide if it drives better. After all, the purpose of these injectors is for better driveability, not to make bigger numbers on a dyno.

 

[J1000]

@dirtydeeds we've been soliciting both cash and advice for months preceding this test and you were nowhere to be found. Now you are here pointing out all of the flaws and details we missed. There is a whole other thread about these injectors and how they feel on the road. That horse has been beaten to death. We tried to put numbers to it and this is the result. I agree, I think there are things that the dyno can not measure but that's not what this thread is about.

 

[dirtydeeds]

@dirtydeeds we've been soliciting both cash and advice for months preceding this test and you were nowhere to be found. Now you are here pointing out all of the flaws and details we missed. There is a whole other thread about these injectors and how they feel on the road. That horse has been beaten to death. We tried to put numbers to it and this is the result. I agree, I think there are things that the dyno can not measure but that's not what this thread is about.

If you’re the same guy whom called me asking for a set of injectors to test, then I said no not unless you bought them, that’s because I knew gains would be immeasurable a dyno. As I’ve told you and everyone else since adding 12 hole injectors was an option for our trucks.

I also told the person on the phone that unless you lay into the throttle exactly the same way each time you run a particular test, there’s another unaccounted variable skewing your testing. Did you put a block under the gas pedal to make sure it was pressed down the same amount every time? Not only that, but throttle position isn’t the only variable to consider when testing with a block under the gas pedal, how quickly you apply the throttle is taken into consideration by the ecu when it determines how much to open the throttle blade. None of this was taken into consideration during your testing.

Again, that’s my point, while you’ve certainly reached your goal of ruffling feathers, there were variables you’ve not considered. And that cannot be measured by a dyno. There’s two sides to this and swaying a guy from making his truck drive better for attention doesn’t do anybody any kind of a service.

 

[suprarx7nut]

The difference between a random customer anecdote and me, I’ve built or had part of building literally thousands of Toyota’s since 2003.

The way you define improved driveability in a measurable parameter is to put an expert behind the wheel and let him drive the vehicle before and after. That’s exactly what the OEM does after they’re done testing on the dyno. That’s why only testing on the dyno is inconclusive.

Admittedly your test is incomplete. As you’ve stated you did not allow the ecu to fully adapt. That doesn’t happen until about 50 miles of regular driving. It doesn’t even come close to happening just in the time it takes to warm up. The difference in driveability quickly after an ecu hard reset compared to 50 miles after a hard reset, even with the same injectors is big. The truck drives way better after the ecu has adapted than it did before. Then on top of the ecu hard reset you’ve thrown in an entirely different set of injectors as another big variable.

With skewed testing, and a desire to ‘stir the pot’ for views on YouTube, I suggest guys test 12 hole injectors for themselves, on the road, then decide if it drives better. After all, the purpose of these injectors is for better driveability, not to make bigger numbers on a dyno.

If you can feel a difference, there is a difference to be measured. If you can't measure the difference you may be feeling the placebo effect. The placebo effect is very real and seasoned experts are not immune. This is why data matters and why we only considered objective data in this testing.

When OEMs tune engines and test system performance components like fuel injectors, I believe they're doing this (engine dyno). We can't do an engine dyno easily, but we can do a rolling dyno, so we did.

If the ECU learning period caused any sort of significant impact, we should see it in the data. We have all the engine sensors logged and the output AFRs. There's no population variance.

@dirtydeeds we've been soliciting both cash and advice for months preceding this test and you were nowhere to be found. Now you are here pointing out all of the flaws and details we missed. There is a whole other thread about these injectors and how they feel on the road. That horse has been beaten to death. We tried to put numbers to it and this is the result. I agree, I think there are things that the dyno can not measure but that's not what this thread is about.

To be fair to @dirtydeeds he did offer up a small discount on a 12-hole injector set in 2019 after I chatted on the phone with him. We discussed the test plan and agreed the testing would be valuable at partial throttle if throttle could be accurately controlled. No other advice was offered related to the test method on gasoline. He had E85 advice, but we never performed those tests so that wasn't relevant.

I believe we succeeded in testing partial throttle - even better than I anticipated because of that nifty test mode the Mustang dyno allows with variable load to maintain RPM and partial throttle.

We tested partial throttle, low end torque. There is no discernible difference.

Regarding ECU reset time: I respectfully disagree, Keith. We ran the 12 hole injectors first. ECU was fully programmed and learned and ready. We then changed to 4 hole injectors, then back to 12 hole injectors with one or two ECU resets in between. There's no way we covered 50 miles of drive distance on the rollers between tests. There was ZERO difference in the ECU data or produced power or torque. The ECU learning can have an effect when changing some parts, certainly. Throttle position sensors, crank sensors, O2 sensors, etc.... come to mind), but I feel confident we can rule it out in this case. We showed that the runs before and after the reset were no different with the 12 hole injectors. I'm not sure what else to say on that. We have as thorough a data set as you'll find.

 

[suprarx7nut]

If you’re the same guy whom called me asking for a set of injectors to test, then I said no not unless you bought them, that’s because I knew gains would be immeasurable a dyno. As I’ve told you and everyone else since adding 12 hole injectors was an option for our trucks.

I also told the person on the phone that unless you lay into the throttle exactly the same way each time you run a particular test, there’s another unaccounted variable skewing your testing. Did you put a block under the gas pedal to make sure it was pressed down the same amount every time? Not only that, but throttle position isn’t the only variable to consider when testing with a block under the gas pedal, how quickly you apply the throttle is taken into consideration by the ecu when it determines how much to open the throttle blade. None of this was taken into consideration during your testing.

Again, that’s my point, while you’ve certainly reached your goal of ruffling feathers, there were variables you’ve not considered. And that cannot be measured by a dyno. There’s two sides to this and swaying a guy from making his truck drive better for attention doesn’t do anybody any kind of a service.

I was the guy that spoke with you regarding testing last year.

Go watch the video, please. I explain in detail exactly how the test works. Throttle was controlled extremely tightly - far better than a block under the pedal, even.

We have 30,000 data points from this test. Give me something to plot. Give me a concrete hypothesis related to data and I'll run it for you.

*add on: Our point wasn't to ruffle feathers. Jimmy was on your side prior to testing. He and I were both interested in data, despite our different opinions. We worked together and collected some awesome data. When the results contradicted his previous opinion, he didn't fight the results or find a way out. He updated his opinion as you can see in this thread and we're both walking away with a better understanding of how this mod affects the 2UZ in the 100 series.

I spent half a day off work, another 10-15 hours studying data, recording narration and editing together a video to support my YouTube channel and brand. That's where the ruffling feathers comes into play - not in the data itself. The data is what it is. I didn't modify it and Jimmy witnessed it unfold in real time.

 

[suprarx7nut]

Sorry for all the back to back posts, but I've updated post #1 with the raw data set for anyone to download and play around with.