G50 Aftermarket Injector Alternative!!??

[Q451990]

That is understandable that they wouldn't take it back.

I don't disagree, but I certainly won't risk venturing into the risky aftermarket part land without some independent testing to be sure they perform as well as OEM. I'm sure I can work something out with someone...

Heath

[Q451990]

Quick update:

I just ordered 8 of the BWD injectors from O'Reilly Auto... they should be in tomorrow. So I'll ship them to Deatschwerks on Monday. I would say we should have test results within a week after that - depending on how long it takes DHL ground to get them there, and how long it takes to get them tested.

I'll keep everyone posted!

Heath

[Q45tech]

Remember the normal tests are worthless since what we need to know is the matched flow at idle and cruise: say at 7% duty cycle and 34 psi and 12% duty cycle at 34 psi.

Whether 370cc flows at 43.4 is useless because the ecu NEVER asks for more than 60% duty cycle at 43.4 psi or 222cc.

Also the pintle lift time must be the same as oem otherwise the ecu dead time offset will be wrong. Also the voltage offset must be the same when cranking at 9 volts.

Copying an injector must be a perfect copy or something will misbehave.

[Q451990]

I assume by "normal tests" you mean max flow rate and that's it? Here's a link to the Deatschwerks testing page.

http://www.deatschwerks.com/ca...s.php

The test they'll be doing is the "dynamic" test. Do you think this test will tell us if these injectors are worth trying? If not, do you think there's another test option that would be more valid?

In any case I'm not really optimistic that these will be OEM equivalent, but I think these need to be explored - when the best long term alternative is replacing the injectors with OEM new old stock units that will fail again in a year or two. If that's the only way I can get OEM performance, then that's what I'll do.

Heath
Modified by Q451990 at 11:10 PM 7/24/2008

[Q45tech]

I outlined EXACTLY the things one needs to know if the injector is a suitable substitute.

[Q451990]

I outlined EXACTLY the things one needs to know if the injector is a suitable substitute.

Then I will print and send that post with them to Deatschwerks! Hopefully we can get an answer.

I found some training info here...

http://www.smp-training.com/BW....html

if their injectors are as good as the game they talk they might be OK.

Heath

[Q45denver]

Did you go with the phase one's or the phase two's?

[Q451990]

Phase II... works out a tad bit better price-wise, even with new rails and subharness. I was able to get the Phase IIs for about $102 each, while the Phase I injectors are about $177.

Heath

[maxnix]

I found some training info here...

http://www.smp-training.com/BW....html

if their injectors are as good as the game they talk they might be OK.

Heath

Nice little link for the lay man, Heath.

[Q451990]

I thought so... kind of "injectors for dummies" but I was pleased to see at least the mention of testing and quality control. I believe the training is for the counter sales people at the parts places.

Haeth

[elwesso]

heath you are the MAN!!! Thanks so much for doing this...

I cant wait to see what david thinks of the injectors.. If these are truly better, people may want to just convert to phase II and be done with it.

[AlabamaDan]

heath you are the MAN!!! Thanks so much for doing this...

I cant wait to see what david thinks of the injectors.. If these are truly better, people may want to just convert to phase II and be done with it.

Ditto! Thanks and you ARE THE MAN!!!!

[Q45denver]

Remember phase two fuel rails do not fit phase one intakes unless modified. I have a new Dueatschwerks adapter kit I could sell someone at a substantial discount (this allows use of phase one rails and harness with phase two injectors) or I could develop a simple washer and seal set to use with the phase two rails.

Just wondering if the Nico ECU program is optimized to use with the phase two injectors?

I realize the phase two is an improved design but the phase one injectors appear to be a more robustly made which might explain their cost difference. Just glad someone is addressing this issue which I've had to live with for the past ten years.

[Q451990]

Remember phase two fuel rails do not fit phase one intakes unless modified. I have a new Dueatschwerks adapter kit I could sell someone at a substantial discount (this allows use of phase one rails and harness with phase two injectors) or I could develop a simple washer and seal set to use with the phase two rails.

According to Keith (TexasOil) in this thread, the intake runners are fine, he just changed the thickness of the spacers under the tabs that hold the rails down.

I replaced the fuel rail and injectors. The new 2nd gen side feed injectors will not fit in the old(1st gen) rail. The only adaption needed was removing the black plastic spacers at the hold down bolts and replacing with a thinner stack of flat washers to allow the shorter injector+rail to seat squarely on the manifold. One could also grind the OEM plastic spacers thinner. 2nd gen wiring harness (>94) needed. The 92 (168K miles) has been silky and buttery smooth with these FI.

Heath

[Q45denver]

Not sure how he did it, but I just drilled out the OEM plastic or hard rubber phase one spacers and Rtv'd into the center the phase two spacers which have a much smaller diameter. Then I Rtv'd the assembly together with a flat metal backing washer into the intake manifold. The RTV was just to ensure a good seal and probably not needed but remember these cars run terrible with any vacuum leaks. The thickness of the seal makes me wonder if the pintle tip is fitting far enough into the intake but maybe that is not too critical.

[Q451990]

So you're saying you modified spacers where the injectors fit into the runners? That appears to be different than what Keith did... guess I'll cross that bridge if/when I come to it.

Heath

[Q451990]

Q45denver - Does the thickness differ between the P1 and P2 spacers?

Everyone - the injectors should arrive at DW on Wednesday... maybe we'll know something by the end of the week?!

Heath

[Q45denver]

No, I think they are the same thickness.

[Q451990]

Alright... I just got off of the phone with Mike at Deatschwerks. The preliminary tests look pretty good. In his opinion, so far, these look pretty comparable to the OEM injectors. They had four new OEM phase 2 injectors on hand that they ran measurements on, and then compared them to the aftermarket units that I sent to them. The variance at WOT was about 4% vs. 0.5% for the OEM, and on the dynamic test something like 3.2% aftermarket vs. 0.3% OEM. So not perfectly balanced, but good. He also mentioned that the OEM Q45 injectors are particularly well matched vs. other Nissan injectors that they see. So that the 4% variance on other Nissan vehicles would probably be an exact match to their OEM tolerances. I asked him about the dynamic testing, and those flow rates are at fairly wide pulse widths, say 100% and 70%, maybe 60% - I haven't seen the actual test results yet. So that in itself will not answer Q45Tech's questions about the 7% and 12% duty cycle. When I asked Mike about this, he said that they test them at higher duty cycles for flow testing, and then extrapolate the numbers for the lower pulse widths by checking the injector's latency. This is done with a power supply and oscilloscope. The way he explained latency to me was something like this. Say at idle the ECU calls for the injector to be open for 1 second. (Even at WOT the open time is measured in milliseconds, but this is just an example) The injector will not be open for a full second because the injector takes a period of time to open. So say it takes .25 seconds for the injector to open - that means that the injector will only actually be open for .75 seconds. Now, say at WOT the ECU wants the injector open for 10 seconds. (again magnified for the example)... then you would have a .25 second period to open and 9.75 seconds of open time.

So say the OEM injector takes .25 seconds to open, and the aftermarket took .50 to open. Then you would only get 0.5 seconds of open time at idle and 9.5 seconds at WOT. So the performance change at idle and cruise would be drastic, but not so noticeable at WOT.

Mike said that they don't actually do the flow testing at low duty cycles because it would take a long time to get enough fluid through the injectors to measure it and get meaningful results. In their experience, the flow rates at higher duty cycles, and measuring the latency gives them the best measure of how the injector will perform through the entire range.

He still needs to complete the latency test and then he will ship them back to me with the test results. I will scan those results and post them here...

Heath
Modified by Q451990 at 12:11 PM 8/5/2008

[Q451990]

So assuming the latency checks out OK, I guess the question becomes how much will the variance in flow rates be noticeable in real world performance? Would a 4% difference be that different in normal driving, or at idle - especially in that the 4% would be much less at that duty cycle. At least I assume that the variance decreases linear fashion with the pulse width.

Also, would other wear factors on a 12-18 year old engine (coils, overall cylinder compression, etc.) make fine tuned injectors a moot point?

And of course there's the long term performance vs. OEM. If these get wildly unbalanced over time - then that's a problem - but there's no way to know that now.

The big advantage is the claim that they can actually withstand the fuels that we're forced to buy today... so my first thought is that I'm willing to live with wider tolerances in flow rate to get some longevity out of the injectors.

Sort of thinking out loud here, but I would love to hear everyone's opinion.

Heath

[Q45tech]

The time to open LATENCY is more around 0.66 milliseconds............easy to see on oscilloscope the current has a ~ in the middle of an upramp.

The latency varies with voltage so the variance is mostly important during cranking when voltage is below 12.8 volts.

So the aftermarkets are 10-13 x worse than oem 0.3 vs 3-4.

A power balance test will tell the tale as the injector [fuel flow] is 95% of that.

Hopefully they remember to use 34 psi instead of 44 psi [WOT].

Another thing to look at is the shutoff waveform: the bounce and ring that occurs as the injector tries to close...............spring stiffness.

It will be interesting to compare new oem with new aftermarket but that may take years and years.

Has anyone lost a new oem replacement yet? But then again the switchover in mass has only been 5 months in ATL and 15 months in NE.

[Q451990]

Has anyone lost a new oem replacement yet? But then again the switchover in mass has only been 5 months in ATL and 15 months in NE.

From that question, do you think there is some chance that OEM replacements could withstand the E10 fuel with some expectation of longevity? I could live with them making it 5 years and 100K, but I really want to avoid failures every 12-15 months.

According to Mike at Deatschwerks they are seeing some repeat failures of new OEM units in the Phase 1 injectors from the areas that have been dealing with this for a while. He has not seen any of the Phase 2 style.

I believe one member here has mentioned replacing OEM replacements after a year or so... maybe from Denver?

Heath

[AlabamaDan]

So I guess the next test is for someone to put a set in their Q and give it a whirl. Probably wouldn't want to mismatch would we.

What about some of those guys who put the VH in their 240s and stuff? I'll bet they're running aftermarket injectors. Maybe we could drift over to their forum and ask for advice.

[maxnix]

Most are not as the VH injectors are good up to 6 psi boost.

[Q451990]

Mike emailed the latency test results to me. I hope these will be in the report that I post when I actually get the injectors back, but here's what I have at the moment:

VOLTAGE: Aftermarket/OEM6V: 2.39/2.567V: 2.01/2.108V: 1.72/1.839V: 1.53/1.6110V: 1.40/1.4711V: 1.26/1.3312V: 1.23/1.2413V: 1.09/1.1514V: 1.01/1.0915V: 0.98/1.0216V: 0.90/0.9417V: 0.85/0.90

I assume the measurement unit is milliseconds. So from what I can tell, the aftermarket injectors have a slightly faster response time. Mike summarized it by saying:

The two injectors are very close in terms of latency.

From everything we have tested, these aftermarket injectors should work well as replacements. Obviously, we can not judge as to the reliability of the injectors.

and

Yes, I think these injectors would work but the best testing will be getting them in a car. From what we have seen it would not be a waste of time or money to try installing a set.

Dennis, what do you think... are these times close enough together to think that the ECU will be able to compensate for any richness in the mixture by changing the pulse width?

Thanks!

Heath
Modified by Q451990 at 6:40 PM 8/6/2008

[maxnix]

Well, the O2 sensors will provide some trim on closed loop feed back, right?

[Q45tech]

Don't be confused by their version of latency at 12-13 volts and the fact that the normal idle injection time is the same as latency ~ 1.2 ms.

The latency is the time to 100% full flow [some use 66% of full flow]. The injector starts spraying at 0.6 ms as the pintle first lifts so there is some amount of spraying in 0.6 ms.

10 years ago I experimented on injectors to measure the amount of flow and it is very non linear depending on viscosity of gasoline and temperature and air pressure inside runner.

you can reverse engineer the amount of gasoline required in milligrams if you know the engine at idle ingests 8 grams per second from MAF [for 650/60 = ~ 11 revs/sec or 44 injections at 14.7A/F.............8/44 = 0.1818 or 182 milligrams of air/14.7 = 12.37 milligrams of fuel arriving inside cylinder.

Probably 12-16 milligrams of gasoline is the idle flow per injection.

By the way a tiny tiny drop [from the tinest eyedropper] of gasoline weighs about 15-20 milligrams.

The flow at 34 psi is 11.55% less than the flow at 43.4 psi.

After all this hopefully you see why Nissan chose to adjust idle trim by referencing rpm after warm up and not fiddle with a 0.1 millisecond injection.

Get it stable [1.2 ms] then fiddle with air [via IAC] and +- ignition advance to try to maintain rpm.

I expect ecu will learn with a properly set IAC what to do to achieve a steady idle rpm

[Q451990]

I expect ecu will learn with a properly set IAC what to do to achieve a steady idle rpm

Well, I think I'm going to install them and see what happens.

Hopefully I'll see problems within the first year if there is going to be an issue.... what the hell it's just money, right? My next step is to try to rustle up a set of used 94-96 fuel rails. Joe's price is about $175 for the set, so it's not the end of the world if I have to go with new ones.

The one thing that really bugs me is the injector subharness. $223 for the 94-96 style and about $75 for the 90-93 style. I guess they used gold in the wires or something. I may try swapping out the injector connectors if I can find them... I have been looking at that option, but pretty much put everything on hold until I made a call on the injectors. Then it's just another $1000 in miscellaneous hoses and parts and I'll be back up and running!

This will probably take me a month or two to do... the back isn't as good as it used to be and having a garage releives the pressure of getting the torn up car out of the driveway. A lot has changed since I did my last DIY plenum job 8 or 9 years ago in my parent's driveway!

I also need to do some suspension maintenence on my truck before I get started on the Q, since I need one of the two operational. I suspect every peice of rubber on that truck is 11 years old and has lived a hard 135K miles in service, with the first 105K as a Terminix fleet truck. I'm starting to get some scary vibrations and slop in the steering that makes me almost afraid to drive it... my truck maintenence has been much different than on the cars

I will keep everyone posted as I make progress - and I'll be sure to post up some .pdfs of the actual test results when I get everything back.

Heath

[Q45denver]

Do you know if BWD has made these injectors for awhile? I do have the connector adaptors if you decide to go that route. They are part of the adaptor set I got from Deatschwerks that I did not use.

[Q451990]

According to the person I talked to at BWD, the injectors have been made in their Greenville, SC plant since sometime last year. Prior to that they came from China. I posted the codes that should be etched on the connector area to tell the US version vs. the older China version earlier in the thread.

Could you post a picture of the adapters? I'm curious to see what they used. There are posts on some of the Subaru forums, and a vendor that specializes in injector kits that indicates that they are available separately, but I need to do some more research.

Thanks!

Heath
Modified by Q451990 at 8:45 PM 10/4/2008