2GR-FE 07 Camry

[Rokassileika]

Here Iam. With my problems. Again .. its been probably almost 2 months with my brand new Zawolli Zeta reducer. The economy was so great now that i stopped counting kilometers. I used to drive approx 160-170 miles with full tank now I can do 200 miles easy, 220-230 if im not that harsh on the pedal. Inconsistant power output still hits me, but i had it going so strong that it felt once that it had a turbocharger. But now there is one thing going weird. Once I reach that 160-170 mile mark i get a code for high knock sensor input for one bank. The transmission starts working very harsh until I stop and restart the car, then after some miles it happens again unless I refill the lpg tank. So i do believe that the lpg detonates because of high temperature in the cylinders. So now I think i should richen the mixture a little? To get these cylinders working less hot than they are now? Gonna get less miles now probably and weaker power output. I think im kind of getting to the bottom of inconsistan power output problem. I believe it is the timing advance that messes up the performance of this powerful motor. And transmission problems are very much involved with engine timing advancing I believe. Besides, I find it really complicated setting the right multiplier for lpg according to the lambda readings lpg vs gasoline when the vehicle is moving. I can do this while the vehicle is stationary and make the curves look great but im 100% sure this is not it, there is more. Any suggestions what I should be trying next? I did order a new valve body with correct pressure switch plate will update when I find some time and confidence fitting it in, but what the heck is going on with this knock sens. High input? Any other ideas or suggestions regarding to these annoying problems, maybe auto- calibration? Does that work somehow? Thank You! How can I trick that lpg computer to make lpg injectors opening times longer when the lpg pressure starts dropping? Well, this explains me why on other cars the power felt much better when the lpg tank was getting empty....

[Gilbertd]

LPG pressure won't be dropping on when the tank is getting low. While you still have liquid in the tank, even only a small amount, the pressure will be sitting at around 10 bar and will stay there. That gets fed to the reducer which drops the pressure considerably, down to around 1.3-1.5 bar, so the pressure to the injectors will remain the same until you actually run out.

[LPGC]

I believe the 2GR-FE engines are known for knock sensor problems anyway, and this can effect gearbox shifts.

It is unlikely really detonating / knocking on LPG which has an octane rating of around 110, more likely that after a period of time that sees you use half a tank of LPG the petrol ECU flags up a problem with the knock sensor. Does the same problem occur if you run on petrol for that time period?

You don't set most of the multiplier / map according to lambda sensor readings, you set it so fuel trims are the same across the engine load/rpm range on LPG as they are on petrol. If you don't have ability to read fuel trims (via OBD live data) you can use the alternative method which is to make petrol injector pulse duration (pinj) stay the same when you switch between petrol and LPG across the engine load/rpm range. The fuel trims / OBD method is usually the better of the two methods and this is definitely true of the 2GR-FE engines. You only set high load open loop fuelling using lambda sensor readings and if the rest of the map is correct you should be able to guestimate where the calibration / multiplier should be for those conditions by extending the multiplier line (or if it's a table type map adjusting the numbers) to follow on from the shape of the closed loop map.

This (below) is one of my maps for a different make/model vehicle, using different injectors and behind the scenes allowing for the effectively varying petrol fuel pressure (petrol injectors are fed with constant pressure relative to atmospheric pressure regardless of manifold pressure while LPG injectors are fed with pressure that rises and falls along with manifold pressure so LPG pressure is kept constant relative to manifold pressure, at light engine loads with low manifold pressure effective petrol pressure is higher than at higher engine loads when there is higher manifold pressure, so a petrol injector pulse length of say 4ms at low loads will inject more petrol than the same 4ms pulse length at higher engine loads). Usually it goes open loop before petrol injector duration is 10ms but you can see that even if I had to guess the rest of the map between 10ms and 25ms it would follow the intuitive trend of the usually closed loop section between 1.5 and 10ms. Most of the map above around 15ms isn't used anyway because on this vehicle the peak petrol injector pulse duration is only around 15ms but it's still worth making the 16ms and higher figures correct in case the petrol ECU adds some 'tip in' enrichment when suddenly putting your foot down.

map.jpg (305.77 KiB) Viewed 3700 times

I'm not saying your map should look the same (and it likely won't because you'll be using a different ECU, injectors, pressure, on a different make/model vehicle and you won't have correction for effectively variable petrol pressure dialled in) but if it is correct it should be smooth and remain smooth around the usual transition area between closed and open loop fuelling.

You should also get similar petrol injector pulse length figures for both cylinder banks.

[Rokassileika]

Okay, I will check my map and report the situation here. What else it could be if the same happens while running with petrol? Could it be mechanical damage? If yes then where?

[LPGC]

If the same happens when running on petrol the first thing I'd suspect is the knock sensor that I mentioned above is a common problem on your engine.

[Rokassileika]

Could it be the culprit of inconsistent power output? I dont know whats wrong since it did not happen again. And I drove on LPG always until today.

[LPGC]

Yes it could be.

You need to make sure it runs OK on petrol before trying to get it to run properly on LPG.

[Rokassileika]

Small update. Remember I had a problem with uneven fuel delivery to different banks? I believe it was caused by leaky reducer. Since i replaced it to a new one i didnt notice that the hose going to injectors had a riny crack and the old adjustable clamp couldnt hold strong anymore. I cut the broken hose end, replaced the clamp with a new one and now I dont have any check engine lights. Doesnt matter if the tank is full or almost empty. The car is unrecognasible in the word of power output. It is now 60-70% stable and strong of the time. Before this fix the power was very inconsistant and 70-80% weak of the time. And this was definitely impacting the performance on petrol. I must say. Im really happy with it atm. I wouldnt mind the power output to be stable 90% of the time, but it is so much better now. Just enjoying it. Putting lots of bmws with 3.0D engines into shame . Probbably I will be getting further into this to fully stabilize this engines power output and will keep You updated. Thank You for being connected and helping me out

[LPGC]

I don't remember tbh, and not sure which hose you mean, usually if there's a leak in the feed pipe from reducer to injectors it won't affect the mixture but need fixing because that's a gas leak. If there is a leak on one of the narrower pipes between injectors and manifold then of course that can affect mixture because the leak will allow extra air to the affected cylinder when the manifold and pipe are under vacuum and will also allow gas that is supposed to go to the cylinder to leak to atmoshpere (so a double whammy effect).

Sometimes inconsistent results can be due to worn injectors.

[Rokassileika]

We did talk about my case a lot. I did record the parameters, but it was in previous topics, long time ago. I remember finding that one bank works leaner than another. I was getting a check engine light occasionally (P0420 and sometimes knock sensor high voltage). I don't know how, but tightening and fixing the hose between reducer and injectors did actually do something. I'm on a third tank of LPG- the consumption is so good that its never been before. I had it much worse since installation of the system and I always had problems with check engine light. I forgot to mention that the system that was installed was already a used unit. As much as I believe that the reducer might have been warped all the time (maybe possible? It looked very worn out) letting the gas into cooling system little by little. Anyway, something really did the trick. Either the new reducer or a fixed hose, or both. I drive the car daily so it has been stable for a week or so. I did not remove the springs in the zavoli. Anyway it is really interesting to investigate such cases. I will probably connect OBD one day to see how different these banks are being fueled now. Will update with anything that happens

[Gilbertd]

A bit late on this one but a P0420 code is a catalyst efficiency code. You have pre-cat lambda sensors which flip flop between 0V and 1V and post cat sensors that should stay at a constant 0.5V. If they also flip flop that says the catalytic converter isn't doing its job as they are measuring the same as the pre-cat sensors and will trigger a P0420 code. I've dealt with this same code on two different cars recently, both Toyotas too. To clear it you have 3 options. Replace the catalytic converter (expensive), put a bottle of cat cleaner in the fuel tank (may work if you are lucky) or do as I have done and fit one of these https://www.ebay.co.uk/itm/376439845045. You remove the post cat sensor, screw this in and refit the sensor. I contains a tiny catalytic converter so what the sensor sees is the results of this tiny converter doing what the main one should be doing.

[LPGC]

Does the P0420 occur if running on petrol?

Another reason P0420 can occur even if cats are good is if engine cylinders run different individual mixtures, which is usually if LPG injectors don't flow equally and/or if location of gas outlet spuds on manifolds are not all in the same position (distance to inlet valves etc) as each other. In this situation the cat gets extra work to do because it is fed with exhaust gasses which although on overall average are correct are not individually correct causing output voltage from a post cat lambda sensor to be different to normal. PRE cat lambda readings can still seem to be normal in this situation, with the precat lambda reading something close to the average mixture over the 3 cylinders on the same bank on a V6, though in some cases if you can sample the pre-cat lambda voltage reading quickly enough while the engine is at low enough rpm (fully warmed up no load idle or in Drive idle) you can actually see the lambda voltage fluctuating as individual lean/rich cylinder exhaust gas pulses flow past the lambda. The OBD on a lot of vehicles doesn't sample quickly enough to show such quick changes to pre-cat lambda voltage and even if it does most OBD scan tools don't sample quickly enough but if the vehicle uses a 0>1v pre-cat lambda sensor it is possible to use an oscilloscope to check for this.

Some make/model vehicles (e.g. Mondeos) are more likely to generate P0420 even if injectors are all flowing equally and spuds are all in the correct position. But at least on a 4 cylinder engine'd Mondeo it's relatively easy to fit gas outlet spuds equal distances in the manifold from the inlet valves with short pipe lengths between spuds and injectors.

You recently noted improved results after changing the pressure reducer. If the new reducer outputs a different pressure to the old one that in itself could have made for the improvement for various reasons. One such reason could be if some of the injectors flowed differently to other other injectors at different pressures. Any such difference in injector flow is likely to be less pronounced at longer pulse duration (because when an injector is fully closed it won't flow anything if it's working anything like correctly and when fully open it's condition again doesn't much matter because it's effectively just an open jet) but at shorter pulse durations opening and closing time / response has an increasing effect on total fueling for any given pulse length. But any difference in spud position, manifold nozzle (or 'straw) position / diameter / length can have an effect on fuel dosing / mixture even if injectors all respond and flow the same.

I like the idea of the tiny cat but I'm not sure if there's any cat activity going on or if the apparent fix is more due to it acting like a spacer between the exhaust gas and the post cat lambda sensor, post cat lambda sensor spacers were a fix that could work on Mondeos but there was no cat material in the spacer, the spacer seemed to work mostly by slowing the rate of voltage change / response from the post cat sensor. A completely different method was to wire in an electronic post cat sensor emulator.

[Rokassileika]

Some updates. So as I mentioned, once i fit the new reducer, P0420 was gone, it is gone for a very log time now sonce this fix. Yesterday I connected the OBD to see how it is doing, but i noticed that there is a code saying mixture too rich (there was no check engine light). And indeed it was rich on low loads. So I connected the lpg ecu and corrected the injection times on LPG. What I first noticed- that outstanding engine power is gone now. It does go, but not that good anymore. With overly rich engine it did throw knock sensor codes, rich mixture codes, but 80-90% of the time power was very good, the car was extremely quick I would say. It was all intermittent with fault codes as well. I can bring it back to this point where it runs very good, but i don't think it is a correct fix. Maybe I should switch the injectors with bigger and newer ones? Or try some further LPG adjustments to find that sweet spot between a well running engine and no fault codes?

[Rokassileika]

I'm wondering now. Did the car have power because of consistant line that I had on lpg or was it going good because of how rich the mixture was? What is the idea behind that? By the way, I was reading some literature about rpm correction. Does it mean that on a specific RPM range I can change the lpg injection times aka multiplier? What all these zeros mean on the board?

[LPGC]

There are too many things we don't know about the condition of your engine, its sensors, your LPG system and its calibration for us to be able to advise you properly.

At one point you seemed under the impression you calibrated the LPG system according to lambda readings.. That won't work very well at all, you need to use a better method such as the 'join the dots' method, 'compare pinjs' method, or ideally the 'compare fuel trims' method.

A straight line multiplier calibration probably won't be correct, it usually should peak at around 3.5ms and curve in a similar way to the multiplier chart I posted above.

I don't think we know much about how your temperature compensation factors are setup or the condition of your injectors.

You've said it was running rich at low engine loads according to lamda readings so you adjusted the the multiplier (all of it or at certain points, which points?). You didn't say whether the engine was running in closed loop mode, open loop mode or some sort of error mode (like open loop limp home mode) when it was running rich according to lambda readings - We need at least that info before we can advise you about most things. But one thing we can say is that if you only changed the low load area of the map it shouldn't affect mixture at high loads so shouldn't affect power at high loads and another thing we can say is that if it really is running rich all the time it will cause a P0420 error code (and other error codes).

[Rokassileika]

What I do is I look at the fuel trims. I make the car run on a specific load on gasoline, then i switch to LPG and if it goes to like -10, -15 I correct the multiplier (lower it) so in that case there wouldnt be a jumping fuel trim, so it becomes stable. I don't know anymore, I just literally changed the multiplier on low loads and the power is gone now . Well maybe the problem is in the transmission as I mentioned in other topics about my problems

[LPGC]

When switching between fuels during calibration always allow it a few seconds to settle down before looking at readings because switching fuels can cause momentary changes to pinj and readings even if calibration is perfect.

If you only adjusted the low load area of the map it shouldn't have affected top end power... Unless by adjusting the low load area caused a different problem that makes it run in some sort of error mode.

Have you allowed for temperature and pressure differences when comparing how it ran before you made any adjustements to calibration to how it ran after adjustments? What I'm saying here is that if temperatures or pressures were different when you used to test it at high loads to what they were when you said it felt down on power then it could be an issue with temp / pressure correction that made the difference, not changing the low load area of the map that made the difference.

[Rokassileika]

Sorry for late response. So after a trip to Poland (~1000kms) and some driving in my country i left the car sitting for 2 weeks. Started the car later, on gasoline it was doing fine. Just after switching to lpg i started hearing misfires that were intermittent. Didn't drive it anywhere, the next day i had to change the oil, had to drive it to heat the oil a little. Check engine started blinking on lpg- obviously an injector or two are not functioning properely anymore. I remember they were pretty loud since they were installed used already. The lpg installer mentioned that they might be worn out already. I'm planning buying same AEB injectors, but my installer said i could gain better results with some other injectors that are faster/ more precise. How true is that? Can they really wear out in that way or maybe i could just clean them? Would other injectors make my car feel more responsive or is it just a trick to pay more for fancy injectors? I'm going to go through installations and adaptations by myself as always any information is worth gold for me. Maybe tired injectors were causing all my problems since the beginning?

Thanks in advance

[LPGC]

I forget what injectors you already have but if they're worn out they can create all sorts of mixture related problems due to inconsistent fuelling. There are two types of inconsistency, (A) For a given pressure and pulse length an injector should output the same amount of gas every pulse, if they don't then there's less control over fuelling / (B) Each injector in a set of injectors should flow the same amount of gas at a given pressure and pulse length.

If you mean the actual AEB branded injectors they're very quick at opening and closing, very linear and very accurate, but only if you're lucky enough to buy a good set of 2 x 3 cylinder rails. They don't flow a lot of gas so are no good for engines that make more than around 40bhp per cylinder (seem to have an internal restriction of around 2.7mm) and I found early rails were great while later rails of 3 always seem to have an injector at the end of the rail that flows less than the other 2 on the same rail (B type consistency problems).

Before changing injectors look at pressure and temperature readings to see if anything is different now to what it was when you said it was running great for all those miles. Could even be that your injectors are dirty. But if they're making louder ticking noises that's a sign they're worn out.

[Rokassileika]

I have these. My car has 200 kW (268 bhp) and 350 Nm. Are they good enough? If not, what could I try instead of these? Never really looked into pressure or temperatures. There are no corrections made. All i have to choose from is there: https://sistemosklaipeda.com/produkto-k ... urkstukai/ . Thanks in advance