2GR-FE 07 Camry

[LPGC]

They're AEB injectors, whether or not they'll flow enough probably depends on what pressure they're running at and how much pressure drop when you put your foot down.... What reducer are you running?

It would be easier to replace like with like (AEB injectors again) so the new ones would just fit in the position of the old ones, no need to adjust piping or be concerned about refitting engine trim cover etc. But as said, I used to use AEB injectors for some applications then stopped using them when they started having the injector at the end of the rail flowing less than the other 2 injectors on the rail... If you can buy good AEB injector rails that are consistent and if your reducer runs enough pressure and is pressure stable when you put your foot down it might be a good idea to replace like with like, if not I'd prefer to replace them with something like the Baracuda's sold on the website you linked to.

But before changing anything at least find out what the problem is with your current setup.

[Rokassileika]

As i mentioned before, I replaced the reducer with a brand new one Zavoli Zeta (and left the springs inside). I have some screen records that i showed in other topics how pressure changes during acceleration. The setup was the same at that time, but the reducer was leaking gas into coolant with a very little amount. I had a Peugeot 407 with baracudas, but failed, one of the injectors were stuck open. Our LPG quallity is very questionable here in our country. But, of course, if the car had more power than it has now I would consider risking with more expensive baracuda injectors. What does the rail have to do with AEB injectors? Aren't they all the same? And You said that the AEB injector in the back flows less LPG sometimes than the other two. So this is pressure issue or just a design issue of injectors that fails overtime?

[Rokassileika]

Checked the video, on low loads the pressure was ~ 1.15, when accelerating hard the pressure dropped to ~0.9 that time. Then I made a video running only on gasoline. Noticed, that LPG pressure would fluctuate too? Isn't it supposed to stay steady when running on gasoline? Maybe the injectors were leaky already? But what about the valves that close LPG entrance to reducer when running on gasoline? Or maybe it was just vacuum related differences?

[LPGC]

Rail / injectors are the same thing, you have 2 x rails of 3 AEB injectors, they are also available in rails of 2 or 4 etc. What I was saying is that I stopped using AEB's when I noticed newer batches arrived brand new with the injector on the end of the rail of 3 flowing less than the other 2 injectors on the same rail, seemingly a manufacturing issue.

There have been problems in the UK with a lot of 'heavy ends' in the LPG too. The heavy ends (thick oil dissolved in LPG) can affect performance and longevity of components like injectors, reducers and solenoids, it can make injectors stick open / closed, make them perform incorrectly, etc.

1.15 Bar, especially 0.9 bar, is a bit too low for the AEB injectors to be able to flow enough fuel for a 6 cylinder 270bhp engine. Mind you, the Zavoli Zeta reducer probably won't flow enough for 270bhp either. Is it a Zeta or Zeta S reducer?

Your LPG pressure reading will fluctuate even when you're driving on petrol because the pressure it shows on screen is the difference between gas pressure and manifold pressure... You see the reading change because manifold pressure changes not gas pressure.

[Rokassileika]

Okay, so let's say i install Barracuda injectors. Will they work better than AEB with such pressure? Im planning to leave the lpg to petrol switch function on higher rpms like somewhere between 4-5k as it was. I believe the pressure readings should be different now vs then since the reducer is brand new and not leaking. I can't really rememeber but it might be a Zawoli Zeta S. And, i believe I will have to buy a "Y" form fitting to connect separate rails is that correct? Will this knock off pressure stability somehow or am I good with that?

[LPGC]

Yes Barracuda's can flow more than AEB, particularly at bigger nozzle sizes, because AEB have an internal restriction of around 2.7mm where-as Barracuda's have internals wider than 3mm. Barracuda's are available with slightly differet specs / flow rates of the injector itself (nozzles aside), for example some are marked '120' and will flow 120nL/minute at 1bar, some are marked '130' and will flow 130nL/min at 1bar.

You could plumb Barracuda's similarly to how your AEB's are plumbed with gas flowing through one rail to feed the other but you would need to buy at least an extra fitting to do so. Or you could use a Y or a Y filter. How you plumbed the feed from the reducer to rails wouldn't make a big difference.

But if you intend on setting up the LPG system to switch back to petrol above 4000-5000rpm the bigger flow of the Barracuda's won't be much of an advantage. It makes sense that your system was setup to switch back to petrol at higher rpm's given that it is doubtful your reducer could supply enough gas for 270bhp but the real fix for that aspect would be to fit a reducer that could flow more gas and higher flowing injectors.

Plenty injectors will flow more than the AEB's, not just Barracuda's.

As said, it would keep things simpler and make less work for yourself if you replace like with like... But I don't know if AEB 3 cylinder rails still have the issue I noticed when I stopped using them.

[Rokassileika]

Not sure if I should risk with AEB if they have such issue. I really want my car to have as much power as it can. That all makes sense since in other topics I mentioned noticing one bank working leaner than other on LPG. And it was all comming from the last injector rail (bank 2). I just checked. I have Zavoli Zeta S. I don't use the fancy engine cover so it shouldn't be a big deal. But I still somehow want it to switch on gasoline on high revs, i have a feeling this preserves engine valves somehow (maybe a myth?) But what about low RPMs and high engine load? The Baracudas wouldn't still make a diference here vs AEB? And let's say if I buy AEB again. How can I spot if the rail is defective?

[LPGC]

The 2GRFE engine might have different fuel trims on different cylinder banks anyway even running on petrol - some V6 engines that feature variable valve timing do that, especially among those that use wide band lambda probes. Not sure why, in some case it seams it is because the 2 cylinder banks are programmed with different base maps, probably because the base map is used if the ECU resorts to limp home mode and having 2 different base maps serves as a way to check if lambda probe readings/response seem viable in case limp home mode has been activated, in some cases it seems the ECU is programed to adjust valve timing to make one of the two cylinder banks do most of the work during engine idling, this might be to slightly lower fuel consumption during idling though it could reasonably be expected to effect idle quality to some extent.

You could have 2 injectors rails, each rail has 3 injectors, one of those rails could flow 10% more gas than the other rail while all injectors on each rail flow equally to all other injectors on the same rail... this would be consistent flow on each rail (because all injectors on each individual rail flow equally to each other) but inconsistent flow between rails (because all injectors on one rail flow slightly more or less than all injectors on the other rail). With this situation you could set up the LPG ECU up to compensate for the difference in flow between rails by setting which cylinders are on bank 2 and adding some sort of compensation for bank 2, examples of types of compensation you could use are rpm correction for bank2, flow correction or offset for all cylinders on bank2. Or you could have injectors on individual rails than flow differently to other injectors on each rail. For example the 3rd injector on a rail of 3 could flow 10% less than the other 2 injectors on the same rail. In this example if both rails have one injector that flows 10% less than the other 2 injectors on each rail then the 2 rails still flow an equal amount of gas in total but both rails are inconsistent because one cylinder on each cylinder bank gets 10% less fuel than the other 2 cylinders. The best situation is to have all 6 injectors over 2 rails all respond and flow equally to each other but if that isn't the case then it is better to have rails consistent with themselves (all injectors on a rail flow equally to each other) than to have 2 rails that flow equally to each other but only because they're both inconsistent with themselves to the same extent. It is best not to have to do any bank trims or individual injector correction because all 6 injectors flow equally. Failing that it is better and easier to correct for rails that have all 3 injectors flowing the same on that rail but two rails don't quite flow the same than to correct for individual injectors on a rail flowing differently to other injectors on the same rail. With a lot of ECU's including all but the earliest Stag ECU's (I think yours is a Stag) you can set up corrections for individual cylinders using an offset (say add 0.1ms to the pulse time) or a percentage difference (say add 1% to injector pulse time) or both... But if injectors on a rail don't all flow equally you would be very lucky to be able to set even a combination of offset and percentage that would see injectors on the same bank all flow equally at all points in the engine's rpm/load range even if you set the best possible combinatoin of offset and percentage. The engine has 2 front lambda probes but 6 injectors, this is the underlying reason why it is better to have all injectors on one bank/rail flow equally to each other. If all injectors on a bank flow equally to each other the closed loop response to lambda readings will see each cylinder on the bank gets equal and correct mixure under most conditions even if fuel trims are different across the 2 cylinder banks.

The way to test if injectors flow equally is to first calibrate the system so fuel trims on LPG are the same as fuel trims on petrol, then with the engine switched to run on petrol (all 6 cylinders running on petrol) in turn (one at a time) switch each cylinder to run on LPG while watching fuel trims. If switching any cylinder to LPG effects fuel trims more than switching the other 2 cylinders to LPG effects fuel trims then that injector is suspect. You can do it the other way around too, have all cylinders running on LPG and switch individual cylinders to run on petrol while watching for changes to fuel trims. But both these methods rely on other factors being properly setup too, such as pipe lengths between injectors and manifold being short and equal, spuds in the manifold being in the correct position, no leaky petrol injectors, equal compression and good ignition on all engine cylinders, an LPG ECU that is truly sequential and triggers LPG injector pulses when it receives petrol injector pulses. If your car's OBD port and your OBD scan tool have fast enough sample rate it is sometimes possible to watch the lambda readings in real time, in which case you can sometimes tell if cylinders on a bank have different mixture if lambda doesn't change as smoothly as it should at low rpm. Another way is interpretation of post cat lambda readings. If average mixture on a cylinder bank is close to correct but some cylinders run richer than others on that cylinder bank then both excess air and excess fuel get pumped into the exhaust, this is what can cause the front lambda sensor not to 'flick' as smoothly as it should and it can also cause the cat to respond differently to normal, reflected in unusual post cat lambda readings/response.

To preserve engine valves it would be a good idea to fit a valve lube system on your engine, but for sure if your reducer and/or injectors cannot deliver enough gas for rich enough mixture at high engine loads you should set up the ECU to add petrol to the mix at high engine loads or if your ECU isn't capable of petrol addition should set it up (as it is) to switch back to petrol at high engine loads.

[Rokassileika]

My ECU is a really old Stag 300-6. It doesn't have much functions. So if I understood correctly if my engine has different maps for different banks i have to make corrections for one of the bank injectors? So at the end it doesn't matter if it is AEB or Barracudas? And in the end it would be better to have separate rails instead of a plumbed 2 in 1 rail? The AEB injectors might be old with a defect You mentioned before. The whole system is really old, probably dissasembled from another car in 2016. I still have to buy the injectors first since current AEB are really done, just not sure anymore which ones

[LPGC]

Ideally you shouldn't need to add any cylinder bank trim or individual injector trim, because ideally all LPG injectors flow the same and you mapped the LPG system properly so that fuel trims on LPG are the same as fuel trims on petrol.

It won't make any difference if you plumb the injector rails in series or run a Y of some type and run a separate feed to each rail.

I suppose you could buy the AEB injectors, see how it runs, use the individual injector flow rate test comparison I mentioned to see if they do still have the problem of the last injector on each rail flowing less than the other 2 injectors on each rail, send them back if they do or keep them if they don't.

Another thing I'd do before changing injectors is test that the ECU is driving the injectors correctly because some of the old Stag's used to suffer broken injector output stages. You can test the ECU's output by swapping which injectors are driven by which ECU outputs, you do that by swapping the gas pipes (between injectors and manifold) on injectors and swapping electrical connections on the same injectors... Then if the bad fuelling / misfires follow the injectors you know the problem is injectors, or if the fault follows which ECU channels are driving injectors you know the fault is the ECU.

Some ECU's such as KME Nevo feature a built in injector test facility, it works by opening the gas valves to pressurise the system, close gas valves so no more gas can enter the low pressure side of the system, pulse just one injector until gas pressure falls to 0.4bar, time how long it takes pressure to fall to 0.4bar when that single injector is pulsing, repeat for all injectors on the system, plot a chart that compares injector flow rates based on how long it took pressure to fall whilst taking into account aspects such as minor changes to pinj and rpm during the pressure drop stage. You can run the test at any rpm or load but it is usually used just at idle (to keep conditions the same, no changing engine loads etc) or at idle or very light load if the car has a conventional slushbox automatic gearbox. This method isn't infallible because it cannot prevent some things changing during the test such as gas vapour temp but it is a useful additional method for comparing injectors, another tool in the box.

[Rokassileika]

I'm really short on time, so can't really try that method with AEBs. And the shop would probably refuse to take back used injectors after these tests or would tell me that they dont care. Can never tell, but in my experience it is always a problem with returns. People do wild business usually here. But I will check if ECU is giving good or bad signals to injectors. So probably going to go for Barracudas? What about other injectors? Are they worth looking?

[LPGC]

I've remebered you sent me the link to the shop you intend buying injectors from. On that shop's website I think the only real contenders for you are the Barracuda's. Hana's can be OK but they're not a favourite of mine because they don't like working properly when cold, some like the Landi MED's are expensive for what they are, others can be a bit hit and miss in terms of quality and consistency between batches or individual rails. The Barracuda's are very consistent, buy a set today and convert a specific make/model vehicle, fit nozzle size to suit that vehicle, in 6 months convert another same spec make/model vehicle and fit the same nozzle size, upload the map you created for the vehicle you converted 6 months ago and the map usually won't need adjustment to suit the current vehicle being converted - I know this in itself won't be of much use to you but what will be of help is that it implies consistency of manufacturing tolerances etc, meaning you'll get 6 injectors on 2 x 3 rails that all match in terms of response and flow.

[Rokassileika]

Beautiful. Gonna go with Barracudas then if it is not an ECU failing and probably leaving a 2 in 1 plumbed rail variant. And now I thought, what if I bought a second filter case that stands before injectors? I mean that I would leave the old case and added one more would this help to keep the injectors as clean as possible for a longer time? Besides all that what hole diameter should I choose for Barracudas since they are flowing a different amount of LPG? Maybe they have an information sticker about that, which comes with new injectors? I havent adjusted any maps recently, just want to make sure about the basics. After installation of injectors, when engine is fully hot i will need to make an automatic calibration and only then I will have to adjust the multiplier according to STFT and LTFT in different loads? For example, I always have trouble holding middle or high load, switching from gasoline to lpg and catching the numbers so I can make an adjustments, how do You cope with that? I don't have any help, just me, computer and the car. Thanks!

[LPGC]

The Barracudas are often supplied with very small nozzles, you remove the nozzles from the injectors, drill them to your required size and refit them to the injectors, the nozzles just unscrew. If you can choose a supplied nozzle size I'd suggest you err on the small side at 2.7mm because you can always increase the nozzle size by drilling them but you obviously cannot decrease nozzle size if you buy/drill them too big. If your reducer could flow enough gas to supply the engine with correct mixture at full engine power I would have suggested you go for 3mm nozzles but since we don't expect your reducer can do that you might as well err on the small side for nozzles and this may help make calibration a little easier for you... The smaller nozzle size will mean the injectors need to pulse longer for any given engine load including idle and the longer the pulse the closer the injectors are to linear response. If I were converting your engine I'd fit 3mm nozzles but I'd be fitting a reducer that could supply the engine with enough gas for full engine load and calibration isn't a problem for me.

I calibrate cars by myself, use a straight bit of road, use the gears and accelerator to get to the engine load I want to check/calibrate, glance at the screen to check pinj and/or fuel trims, without moving accelerator position switch fuels, wait long enough for lambda to flick again, glance at the screen again to recheck pinj and/or fuel trims. You could remember the changes you want to make to correct fuelling then park up and change the settings, drive again to recheck after making adjustments. It becomes a lot easier when you've mapped as many systems and become familiar with as many types of injectors as I have hehe.

[Rokassileika]

Absolutely forgot to ask. Since I have an old Stag 300-6. I'm probably not going to find Barracuda injector type. What should I choose instead?

[LPGC]

I don't remember what injector settings are available in old Stag300 ECU's but it likely has several settings that will work OK. What setting was used for the AEB injectors?

If AEB are selected in software keep with the AEB setting because they're close in terms of coil resistance (AEB have 2Ohm coils, Barracuda have 1.9Ohm coils) and the AEB setting won't automatically set a high minimum pulse duration. Other than AEB you could use Hana / OMVL settings, could suggest a few more.

You could list what options you have available, I might be able to suggest others from that list.

[Rokassileika]

Quick update. Purchased the Barracudas, tested Stag ECU for failure. ECU is doing good, injector number 2 is most of the time over-fueling. After some driving it stops misfiring, but with STFT it is obvious that it is sticking, and working the loudest from all. I'm just trying to imagine for how long this one injector was kicking on my nerves.... Adding the list of injectors

[LPGC]

I've looked at the list you posted.

My original advice was to go with the AEB setting (which would be the AEB INJR I-Plus setting on your Stag300 ECU) or another that is close to 2Ohms. Now I'm having second thoughts because I've seen how some other ECU's such as KME Nevo drive AEB's differently to how how they drive most injectors 'normal' type injectors and 'normal' seems good. Nevo lists a wide range of injectors but the main difference with Nevo is that the calibration file is stored in XML format and by viewing the XML file you can see behind the scenes data such as the peak and hold times and pwm frequencies, you can also edit the XML to use custom settings for such aspects, you can't do that with most systems including Stag so you're stuck with whatever peak and hold and PWM etc the ECU uses to drive the type of injectors selected... and unless you use an oscilloscope to check how it's driving them you're left in the dark about how it's driving them.

Matrix injector selection was a go-to on some of the AEB ECU's that didn't have the exact type listed in software but that was because the Matrix setting didn't impose a minimum injector pulse duration (while some settings would impose a minimum pulse duration which could end up being a problem). The Nevo drives the Matrix differently to how it drives most injectors too but that is probably because Matrix are quite a different design to most injectors. Unless we checked using a scope, if we selected injectors such as AEB or Matrix then we couldn't be sure how it was driving them.

There are a few aspects to get right, particularly the peak pulse length and hold PWM current, also the minimum pulse length the ECU might automatically set behind the scenes for a selected injector. If the peak isn't long enough then problems might not be noticed straight away but could occur at cold switchover and/or if pressure is high because both cold temps and higher pressures place more demands on an injector's ability to open. If it were an AEB or other table based calibration ECU there would also be the map shape to consider because with such systems the 'shape' of the map is pre-programmed for given types of injectors and auto-cal in such systems assumes those map shapes will be near correct, you can make changes to the shape of the map manually but it would be less work for an amateur if the shape of the map were roughly suited to the type of injectors selected (assuming other factors such as nozzle size and pressure were also close to correct). But your system isn't AEB or similar and doesn't have such assumed map shape for a given type injector...

All considered I now advise you initially select the Reg Rail SL Fast Reg injector type, because I'm pretty sure that even if Stag300 is sophisticated enough to change the peak and hold length and PWM hold frequency for different types of injectors it will drive that type of injector 'normally' and normally is good for driving the Baracuda's, despite the difference in coil resistance 1.9Ohm Baracuda's vs 3Ohm REG SL. If anything I might expect the Stag's REG-SL setting to have peak pulse length that is a little long for the Baracuda's but if so it won't damage the injectors or cause a problem and in the real world it is better that the peak pulse is a little long rather than a little too short. But at an early point I would watch for the Stag300 imposing a minimum pulse length using the REG SL setting and if there is any evidence of it doing so I would next try the AEB INJR I Plus setting like I originally advised because it is very unlikely that the ECU would impose a minimum pulse length for those injectors. If there were still problems with minimum pulse times then I'd try the generic Matrix setting.

[Rokassileika]

So first things first. Tomorrow I'm going to try and remake injector holder for Barracudas, set up additional secondary filter and try to setup the injectors as Reg Rail SL Fast Reg. Besides that, do I need to setup anything else? As much as I understand, the PWMs peak pulse doesn't last long and that should be OK, but knowing that the injectors keep going on and off consistantly aren't they going to overheat eventually? And secondly let's say I have a problem with minimum pulse lenght. I will probably get rich mixture codes and maybe an engine that is misfiring is that correct?

[LPGC]

The injectors won't overheat and will be fine.

Yes if for correct dosage at idle injectors would need to pulse for (say) 3ms but the ECU imposed a minimum pulse duration of 3.3ms then it would idle 10% rich regardless of how you tried to calibrate the multiplier, this would have the knock on effect of sending fuel trims full negative. But this is extremely unlikely because 1. It is unlikely the ECU will set such a minimum pulse length or if it does it is unlikely to set it as high as that said 3.3ms. 2. Your engine has 583cc cylinders, you'll be fitting nozzle size of max 3mm and your reducer pressure is only 1.2odd bar, so we'd expect idle pinj to be higher than any minimum set by the ECU. The biggest concern isn't actual idle, it is during over-run when manifold pressure, engine load and injector pulse length will fall below idle values but even then it is unlikely any imposed minimum pulse length would be shorter than the necessary pulse length for correct fuelling.