IS250 - cranks, just about to start, but doesn't

[bernieeccles]

Previously, I had a VW Bora TDI 130 which came with 2 sets of keys. I thought I had lost one set so being a cheapskate ordered one from Aliexpress which contained all the electronics but a blank key blade. Had the blade cut but in the meantime found the "lost" key. One day tried to start the car as usual. Key in all the dash lit up, engine turned over and started but then within a second or so, just died. It kept doing this and I could not figure out why as it had run fine before.

It then dawned on me that I had been using the new key fob to start the car but it actually did not contain the transponder to transmit a code to the car. Substitute one on the original keys and voila! no problems.

Which in a long winded way, I ask if your key fob is working correctly or maybe you have 2 and one is good and one is bad which would explain the car running fine at first with the good key and trying and dying with the other.

Just a suggestion, but sometimes there is a very simple explanation while we look for other more complicated reasons.

[Linas.P]

49 minutes ago, bernieeccles said:

Previously, I had a VW Bora TDI 130 which came with 2 sets of keys. I thought I had lost one set so being a cheapskate ordered one from Aliexpress which contained all the electronics but a blank key blade. Had the blade cut but in the meantime found the "lost" key. One day tried to start the car as usual. Key in all the dash lit up, engine turned over and started but then within a second or so, just died. It kept doing this and I could not figure out why as it had run fine before.

It then dawned on me that I had been using the new key fob to start the car but it actually did not contain the transponder to transmit a code to the car. Substitute one on the original keys and voila! no problems.

Which in a long winded way, I ask if your key fob is working correctly or maybe you have 2 and one is good and one is bad which would explain the car running fine at first with the good key and trying and dying with the other.

Just a suggestion, but sometimes there is a very simple explanation while we look for other more complicated reasons.

Car came with only one key and I have confirmed previously that immobiliser get's deactivated (red light in the centre console goes off). Good shout on the immobiliser, but sadly I wish it would be that easy.

2 hours ago, peniole said:

Assuming both pumps are running, then sounds like a lean run (too much air, too little fuel). Large vacuum leak, really off sensor reading, clogged fuel filter or injectors are causes that come to mind.

It is definitely this, the annoying part is to find which one of few dozen parts are causing it and so far I am having no luck with it. 

[peniole]

10 hours ago, Linas.P said:

It is definitely this, the annoying part is to find which one of few dozen parts are causing it and so far I am having no luck with i

well, I would think wide first. Injectors are last on my list as I would expect a cylinder or bank running lean rather than a no start. I would start with filter. A way to test it is to put car in acc, but not start it, let the pump prime for at least a couple of minutes (you do say you hear it running continuously). Theoretically the longer the prime, the more fuel can get through a clogged filter, allowing for a longer run before it dies. Which kind of fits as when you start it immediately after it dies you get nothing but crank. If the longer prime doesn't work, then I would be looking at the too much air side, checking all the vacuum tubing by hand (you won't hear any hiss at this point), and checking the sensor readings on techstream once you get it.

[BC99]

I looked at the manual yesterday, the LP pump has two speeds. a Relay switches the current to pass either through or past a resistor. when starting it's supposed to be at the higher level (bypassing the resistor). I assume you've tried sitting with foot on brake (green light) for a longer period before hitting the starter. but then a longer cranking time should be enough to do that anyway.

You mention earth issues are possible? in the wiring or with the components, earth them directly using a jump lead?

There isn't a filter as such, it's part of the fuel pump in the tank, at least not on mine (2007)

Could always disconnect maf to force into "safe" mode?

[Linas.P]

Yes, it let the pump prime for short time, for long time, I have "primed" it manually a little bit when testing if it work (just applying 12V to pin 4 and 5), it still does the same thing. It doesn't seems to respond to accelerator i.e. when it ran for 5s (maybe I am exaggerating the time a bit, but it turned over several times and fired on all cylinders to the point where one would consider it "started"), I was pressing accelerator hoping I could sustain the idle, but it just burn what it had and shut-off.

I mean that fuel is not getting delivered to chamber I am 100% sure, but I can't figure out why. I know the easy answer would be supply pump, because it is what causes 99% of cases, but it doesn't seems to be the case here yet. I guess the answer here would be for me to get right tools and not to waste anyone's time i.e. techstream should come tomorrow and I just need to get right gauge to test it delivers right pressure. In theory I can just take reading from the sensor, but that could be faulty itself.

Earth issues - is just general bug-bear in electric system e.g. maybe short sensor, maybe broken wire, maybe water ingress in one of billion harnesses and the reason I am saying it would be "interesting" is that it would be extremely difficult to troubleshoot as it may be something completely unrelated and intermittent. i.e. maybe injector driver is shorting and all the rest system is fine.

Yes I believe there is only strainer in the tank, only diesel has separate fuel filter.

I have tired to start it in "safe" mode, but it didn't help much and I kind of got unexpected results. Without MAF is no difference at all, but when I disconnected fuel pressure sensor (which could cause all this) then I got not crank at all... and that is not what car should do. So it just all very weird and doesn't do what it should do.

5 hours ago, peniole said:

well, I would think wide first. Injectors are last on my list as I would expect a cylinder or bank running lean rather than a no start. I would start with filter. A way to test it is to put car in acc, but not start it, let the pump prime for at least a couple of minutes (you do say you hear it running continuously). Theoretically the longer the prime, the more fuel can get through a clogged filter, allowing for a longer run before it dies. Which kind of fits as when you start it immediately after it dies you get nothing but crank. If the longer prime doesn't work, then I would be looking at the too much air side, checking all the vacuum tubing by hand (you won't hear any hiss at this point), and checking the sensor readings on techstream once you get it.

The sound which I am confident was coming continuous from high pressure fuel pump only happens after failed start. It does not make any sound before I attempt to start it. 

The vacuum theory is good shout, but I don't believe it would cause engine to stall in the fashion it does. I would expect rough idle, lean/rich mixture and loads of error codes - currently I am getting 0 error codes. Seems as if ECM tries to start and dies before being able to run any self-diagnostics, hence I am looking into "exotic" issue like injector driver ground or short fuel pressure sensor (just disclaimer -  for somebody who may be using this thread in future - I would not recommend doing what they are doing or what I am doing, I am just looking at these videos for ideas of what else to check ) :

or:

 

[Linas.P]

I am still stuck where I was, but finally got Techstream working and got one code which OBD2 reader didn't show and it cant be cleaner. B2799 - that is Immobiliser error. The examples of car with such error looks exactly like mine - cars would crack but won't start. As well few people reporting it appeared from nowhere e.g. they decided to move the car to different parking sport, car was driving fine and then suddenly it won't start - exactly like it was for me!

Just to be very clear - my key works, it lock and unlocks the car, when I turn the IG-ON immobiliser light on the dash goes-off so it seems like car is recognising the key and turning off the immobiliser.

Looking at some other people trying to diagnose it, it could be linked to bad ground which in result interrupts communication between ECU and ID chip (immobiliser) - again what of the thing I was suspecting, but difficult to diagnose. 

I don't want to jump the gun here, but it matches my issue quite well, so it may be it!

[ColinBarber]

It could be faulty wiring, or corrosion/dirt on the connectors if the ECU has been removed and the connectors left exposed for a period of time.

There is a mechanism to register the immobiliser code device to the engine ECU but if that isn't correct then it shouldn't have been running at all previously. 

[Linas.P]

Yes, I would have thought so as well, but I have read at least for 3 people with B2799 code it was exactly like for me - car was driving fine, parked and it won't start again.

I was trying to test for short somewhere, going with the theory of loose wire in harness or something and I am not sure what normal reading should be, when car is locked/unlocked I have about ~0.6A draw which doesn't sound that bad, with ACC-ON it jumps and settles ~7.92A is that normal? And finally with IG-ON it is ~20A+, sometimes drops to ~18-19A, sadly my multi-meter is only rated for 10A and only displays 20A max. I waited for ~3-4 min to see if it drops down, but wires were starting to get warm and it was still ~20A+. Again - I am uncertain what the draw should be with with IG-ON and whenever this indicates massive short or is it just normal considering all electric gizmos in the car?

I mean my next step is to test Certification ECU and ID chip circuits to best of my ability (I can't do oscillometer stuff), but just wondering if there is something obvious that is shorting everything.

The final, think I am a bit worried about is that I only have one key and I am not sure if it is master key i.e. if it is not master and somehow something got corrupted in Certification ECU for sub-key IDs it may not be recognising the key properly. 

[BC99]

Isn't the back up for if the car doesn't recognise the key in the car to hold the fob against the starter button? presumably though won't change things in your case as it's recognising the key.

[bernieeccles]

A quick Google and this video popped up. Not sure if or how it is supposed to work but a lot of comments say it worked for them.

 

[Linas.P]

8 hours ago, BC99 said:

Isn't the back up for if the car doesn't recognise the key in the car to hold the fob against the starter button? presumably though won't change things in your case as it's recognising the key.

No that is if immobiliser doesn't recognise the key i.e. the key Battery is low and immobiliser light remains on. I have actually tried that just in case and replaced Battery in key, but that doesn't work.

[Linas.P]

7 hours ago, bernieeccles said:

A quick Google and this video popped up. Not sure if or how it is supposed to work but a lot of comments say it worked for them.

 

That is quite interesting. Maybe somebody know is there is a way to do the same in techstream? I have looked at few examples of this "trick" and they mention that dealer can do it on their programming tool, which is basically something similar to techstream, so maybe would be a bit fancier way of achieving the same.

Actually thinning about it, on thing which I remembered now - once I got the car from auction the OBD2 port was pushed into the dash, I kind of needed to fish it out. This indicated that somebody were clearly using the port. I wonder if this happened before and they maybe reset ECU with this method, so it ran for some time, but then went back into where it was?!

[Linas.P]

Basically my concern with above reset is that it removes the keys and leaves only master key, because I don't know if the only key I have is master it would be too risky to do for me. If it fails and I happen to have only sub-key, then the only option would be to get new ID Box with new keys from Lexus, which is £2000 in parts only. Although in other hand that may already be the only option anyway.

[bernieeccles]

I'm not sure that there is a master key. I have 2 keys and there is nothing to distinguish between the two. They are identical.

[Linas.P]

From what I am reading there is master key to ID Check Box, which remains on the box no matter what, then there could be many more keys registered. Basically if Master key is lost then the only option is to change the ID Check Box and buy all new keys. 

I have checked again with Techsteam and the only thing I get in summary is that there are 2 keys registered and 5 more slots in memory for keys to be registered. But it does not give me details of what the 2 keys are, obviously I have one of them. As well I have checked ID Check Box to Steering lock connection (Techstream provides the wizard to do this) and it works fine, which is in line with expectations i.e. immobiliser light goes off and steering unlocks as it should. But then there is separate wizard to check "remote engine start" which I assume is connection between ECU and Certification ECU and this test fails - again in line with the issue I am getting - because if Certification ECU does not certify the key then ECU cuts the fuel to the engine.

Now in terms of this second connection it could be just programming error, although I have no clue how that could have happened considering car worked in the past or it could be actually bad connection. The first issue in theory could be caused by jump starting the car and connecting and re-connecting the Battery, which definitely have happened many times to this car recently and it should be fixable by the trick you posted above (jumping pin4 and 13 on OBD port). The second issue will be pain to diagnose and again I am a bit confused as to how it was working all fine and suddenly connection between two ECUs are broken. I am wondering if there is any possibility that washing could have caused it e.g. car has sunroof and maybe if drain is clogged it could flood harness or Certification ECU? I am just speculating but I have heard similar stories on other cars in the past where such critical components are placed in strategically idiotic places e.g. on Audi Q7 if sunroof leaks that is exactly what happens - it floods stereo amp and if you unlucky enough then body ECU as well, which controls central locking.

[bernieeccles]

Had another thought. Have you tried to unlock one of the doors using the key blade to see if the key you have actually belongs to your car?

[Linas.P]

Yes, I have used the key blade for most of the time in last coupe of weeks, because I have removed Battery for charging few times. As well when I go through key registration wizard I get error saying "key is already registered". And there is separate wizard which checks all key amplifiers (which to my surprise are like 9 different spots in the car), basically you select the spot and bring the key somewhere close and car beeps. I checked most of amplifiers and my key beeps, so it is definitely registered to the car and recognised.

[ColinBarber]

You won't get a B2799 code if there is a problem with the key or the transponders - that would be properly diagnosed by the system and the corresponding code generated. B2799 is a fault condition - communication error between the ECU and the ID box.

[Linas.P]

Yeah - have the description, the problem is that one needs oscilloscope to check the signal between ECU and ID box. 

But I guess it still means it could be either out of sync ECU + ID Box, something to be fixed by reset or actual harness issue, which requires identifying faulty wires somehow? 

[ColinBarber]

Rare for a Toyota wiring harness to cause problem unless it has physical damage which is most likely to be in the engine bay (from mice etc.). Check condition of the hardness and also check for water ingress into the engine ECU connectors.

[bernieeccles]

Another interesting video of crank but no start( solved).

 

[Linas.P]

Yes interesting, so in his case it was fuel pressure sensor, which I kind of tried to diagnose but could not completely do it due to it requiring 5V to do the test which I can't provide. 

The other thing I read was that simply unplugging the sensor should allow to start the car in the "limp" mode, but in my case it didn't start. The behaviour was slightly different - with sensor plugged in it starts and immanently stalls, with sensor unplugged it doesn't start at all, just cranks indefinitely and by itself. 

In either case, my sensor is plugged in and unplugging it, clearing the codes and plugging it back on didn't make a difference. As well it seems that I have fuel pressure, but it would be good idea to do what he did and see if it gets to 588-1900Psi when cranking.

Interestingly he didn't do anything with B2799, I assume he just cleared that DTC and it did not appear again. 

[Linas.P]

I am still no further with my issue. I still have B2799 and I just realised I need to remove all dash to check the connector to ID Box... certainly not keen on that. So on one hand I am trying to find reasons not to do it, but as well to find something else I can do in mean time.

As per maintenance manual I have completed individual active tests to all components (injectors, distributor, fuel pump etc) and all of them work as per test conditions. At this point I am fairly confident there is no problems with any parts, maybe with slight suspicion on fuel pressure senor, but it reports pressure fine, so nothing obvious. 

I have as well double checked all the fuses in all junction boxes and all of them are good and correct, except one. Somebody put standard 15A fuse in RH JS cigarette lighter slot, where they should have used low profile fuse - point is fuses are mostly untouched and all present and good.  

Today I have noticed two things which I thought were weird:

1. Maintenance manual suggests to clear DTC B2799 before trying again and checking whenever it appears again, before checking the harness, replacing ID Box, Certification ECU or (Engine) ECU. The problem I have is that I can't clear it, no matter what I do DTC stays and as per my understanding that not suppose to be the case. I have even reset ECU (with 30 min jumper trick) and the DTC is still there. Besides as I have mentioned before it does not show on generic OBD2, only on Techstream, which is again bizarre - if there is DTC why wouldn't OBD2 report it?!  
2. One of the first things I have noticed in Techstream was that my ECU does no report any VIN. I am not sure how it should be on Techstream but I was under impression that it should be able to read VIN from ECU. I have tried scanning in separately via specific option within Techstream and it reported that "either VIN is not available or connection to ECU was lost". Can anyone confirm if VIN is by default available on IS250? I have then manually entered VIN via separate function which allows to write VIN to ECU (which by the way stated that all DTC will be lost in progress) and it successfully completed this function, and now VIN is available in Techstream... however dreaded B2799 is still present.

In short, it isn't right that DTC does not clear at all and I don't know what to do about it. I am not sure if VIN read is at all relevant to this issue, but if ECU did not have VIN registered then I guess that can't be good. To add to this - I don't think anyone messed with ECU, it seems like it as untouched or at least not touched recently - all bolts holding the cover were very rusty and dirty looking, like it was never opened from new.

As well, I am just confused how could any of the parts (harness, or any of ECUs) suddenly malfunction overnight without anyone touching them.

[peniole]

I assume you mean the 4 13 jump? 30 min might not be enough, need to leave the key in position 2 for more than 30 minutes, Ive seen people say 35 and some even to be sure leave it for 45. Also, did you see and continue to see all lights on the dash flashing when you did that? How is the Battery state?

Also, you're right, if it had been successful it should have written the vin to the ECU.

One last thought, in case it has push button start, did you leave the key in the car the whole 30 min?

[Linas.P]

According to video it is 30min, but could last a bit longer. I think in my case it was literally like 30min +45s. Yes it is push button start and the key was in the car the whole time. 

Regarding the Battery, that was my concern as on IG on position car consumes 20A+, so my concern was that by the time this thing finishes the Battery will be dead and I could not complete the reset before taking it out and charging again (which may invalidate the result), so I fully charged the Battery before the start and as soon as 30min timer stopped I have pulled the jumper and switched the car off. But I had same issue upon restart. Battery still had enough juice to crank the car for a while. It would be ideal to keep the Battery on the tender, but sadly I have no place to do that. 

When you say "write VIN" where does the ECU take the VIN from? On my car the win was never present from beginning (not displayed in Techstream), both before and after the reset. And now when I have manually entered it, the car still have same issue, so I wasn't even sure what it suppose to be.