P1166 fuel deviation is this a rare EMI case? PLEASE HELP

Ask for and share advice on using the PicoScope kit to fix vehicles here.
User avatar
RYM6746
TwoWaves
TwoWaves
Posts: 186
Joined: Tue Nov 01, 2016 8:58 am

P1166 fuel deviation is this a rare EMI case? PLEASE HELP

Post by RYM6746 »

Hi guys going to need some help with this one. The fault is very intermittent to say the least. I have however managed to catch it on the scope and I can see what is going on to a degree. Symptoms are rail pressure goes sky high and engine cuts out. Then difficult to restart. Can happen with in the first mile or can take anywhere up to 50 miles for fault to occur. Vehicle is a 2009 Citroen C3 Picasso with the 1.6 L diesel 9HX (DV6ATED4) engine. Chassis no is VF7SH9HXC9T537460. If you look at the end of the capture when the rail pressure goes to its max you can see in the voltage control waveform for the fuel pressure regulator that the voltage does not return to 12V when the ECU is NOT pulling it to ground, I am pretty sure that this is at exactly the same time that EVERY injector is firing because the slight (injector firing) dips that can be seen in the rail pressure waveform line up EXACTLY with the interference on the problem circuit. Unfortunately I didn't have an injector sync hooked up at this time. but if I unplug an injector and then start the vehicle every fourth dip on the pressure waveform is missing so they are dips from injectors. Now here is the puzzling bit. EVERY injector seems to cause the interference (chances of having four faulty injectors are slim) the ECU always manages to pull it to ground so ground is GOOD. So it looks like the interference is on the power supply which comes from the fuse box under the bonnet. I am guessing at this point that it is the high voltage created inside the engine ECU needed to fire the injectors that is causing the issue??? please correct me if I am wrong. How is that effecting the power supply that is coming from the fuse box?. Possibly in the wiring loom?? each injector has its own 2 wires so they can not all be intermittently shorted to the fuel pressure regulator power supply. Is it possible the fuel pressure regulator itself is some how being effected by all of the injectors??. I will try and take more measurements on the supply circuit and injectors but it is very difficult to re create the fault so this is all we have at the moment and this was not easy to get to say the least. Thank you in advance for your time and input :D
Attachments
Citroen_C3 Picasso_2009_Diesel_fault occured.psdata
(9.8 MiB) Downloaded 692 times
Last edited by RYM6746 on Sun Jul 15, 2018 8:52 am, edited 1 time in total.

ben.martins
Pico Staff Member
Pico Staff Member
Posts: 539
Joined: Tue May 16, 2017 1:02 pm

Re: P1166 fuel deviation is this a rare EMF case? PLEASE HEL

Post by ben.martins »

Hello RYM6746,

Don't we just love those intermittent faults?!

Great you managed to capture the rail pressure heading skyward and then obviously the engine will prevent too much pressure and cut the engine. Can I just confirm that in you capture we see the engine being cranked at 1m9secs on the second buffer? Looks like a nice relative compression capture but with an extended crank.

I'm fairly sure that with this system the fuel regulator is normally open so by disconnecting it will see it open to it's max and allow full fuel flow to the pump. This will of course provide max rail pressure. We can see when things start to go wrong that the ECU is still trying to control the regulator but the voltage is dropping out and I would assume no longer providing the solenoid with enough current to over come the spring and so holding it open. I think it would be good to monitor current in this instance to see what happens when this starts to play up, easy said than done I know!

From my old days at Peugeot we had a few issues with under bonnet fuse boards - PSF1. There is a lot more than fuses in this but also a number of relays for ECU supplies and it's here where the ECM gets it's power supply from. I think it would be good to also look power to ECM to see if voltage is dropping when the fault occurs. Sensors will remain ok as only 5V but some of the actuators will suffer. I think the ECM has an internal driver for the injectors but it might be worth checking the voltage at those as well to look for any drop out.

Just a few things from the depths of my mind RYM and I'm sure others will be able to add to it. Keep posting with your findings and I'm sure we can get to the bottom of it.

Kind regards

Ben

User avatar
RYM6746
TwoWaves
TwoWaves
Posts: 186
Joined: Tue Nov 01, 2016 8:58 am

Re: P1166 fuel deviation is this a rare EMF case? PLEASE HEL

Post by RYM6746 »

Hi Ben thanks for your input. On this particular fuel pressure regulator current has to be flowing through it for it to be open to allow the pressure to go up (I know this as I have grounded it and watched the pressure go through the roof). You are correct that the 12V is fed through a relay then a fuse inside the under bonnet fuse box (relay is controlled by the engine ecu) to the fuel pressure regulator this is one circuit on its own that does not feed any other component. The engine ECU has capacitors in it producing high voltage to feed the injectors. If you clean up the rail pressure sensor waveform with the filter you can see each pressure drop as each injector is fired. These pressure drops line up EXACTLY with the lack of voltage on the pressure regulator waveform(when fault is occurring ). Also as soon as the engine ECU cuts the engine/injectors the interference/voltage returns to normal on the pressure regulator waveform. This is too much of a coincidence. Here is the million dollar question. HOW does the engine ECU high voltage/injector waveforms effect the 12V supply of the fuel pressure regulator (enough to stop it working correctly as seen in the pressure waveform). The 12V supply is not coming from the ECU (if i pull correct fuse the 12V disappears) and cannot be shorted to it as each injector has its own two wires (eight in total) and EVERY injector firing event is effecting the 12V to pressure regulator. The ONLY link I can make is the high voltage injector wires are twisted for that exact reason to reduce noise on other circuits. The 12V pressure regulator wire runs in the same wiring loom. Is it plausible to think that 4 twisted pairs of wires are running around the 12V pressure regulator wire and in the right conditions causing some sort of EMI. I really need somebody who fully understands the cause and effect of EMI to guide me here.Thankyou all for your help

I fully agree I could do with measuring the 12V supply and injector waveforms when fault is occurring. I have set the scope up to measure these circuits and driven it for around 10 miles. They have behaved exactly as you would expect because the fault has not occured. So all I have to offer at this moment in time is the one waveform. As for your cranking question I am not at work and on my phone at the moment so I can't see the waveform but I am relatively sure that there was no cranking event in second capture.

It's getting to the point where it would be beneficial for me to temporarily run an external 12V supply wire from the under bonnet fuse box to the pressure regulator and just cable tie it externally to the wiring loom and see if it fixes the issue. If so then I would have to open up the entire wiring loom and re position the original wire.

User avatar
RYM6746
TwoWaves
TwoWaves
Posts: 186
Joined: Tue Nov 01, 2016 8:58 am

Re: P1166 fuel deviation is this a rare EMI case? PLEASE HEL

Post by RYM6746 »

Well due to a lack of a better thing to do the ECU got sent off to AC tonics for a full test. Much to my amazement they found a fault with it. They wernt very specific (you speak to a receptionist) but they did say it was down to a dry solder joint on a 3rd party repair, they said it was irreparable and requires a new ECU, that they can supply, which will be plug and play. Not sure how this effecting the supply to the pressure regulator that came from the fuse box?. The interference how ever was almost definitely being created by the capacitors high voltage which is mounted inside the ECU. We will fit the new ECU and I will keep you posted. We might even re route the pressure regulator supply wire for the belts and braces effect. If all is good after a few hundred miles/weeks I will be sure to confirm the fix on here :D

ben.martins
Pico Staff Member
Pico Staff Member
Posts: 539
Joined: Tue May 16, 2017 1:02 pm

Re: P1166 fuel deviation is this a rare EMI case? PLEASE HEL

Post by ben.martins »

Hi RYM6746,

My apologies for not getting back to you sooner.

Well, there's a turn up for the books! High voltage leakage across the ECM could affect other components and I guess the warmer it gets the worse it will become. I'm thinking high engine load and high RPM to make the switching effect of the injectors increase. Interesting that it's had a 3rd party repair. Does this imply work previously carried out?

Hindsight being a wonderful thing, we know that the ECU controls the ground for the regulator. What would have been telling I guess is if we had observed both sides of the regulator to observe both the supply and the control. That way we could see if the 'leakage' was from the ECM side or the power side. If you are fitting a new ECM to the vehicle is there a chance you'll get the old one back? Be interesting to do some back to back testing with it! As you have mentioned how would a voltage leakage cause the regulator to be constantly grounded and therefore remain open to cause max fuel pressure. I wonder if the transistor that is used to turn on the ground control on in the ECU could be effect by this and unintentionally being turned on, I guess it's a hard one to prove.

Would be great if you could keep us all posted. I for one am interested to see the results! Once fitted any chance you can post a capture up?

I've also added a picture of what I meant by a relative compression waveform in the trace....

Kind regards

Ben
Attachments
Citroen Forum.png

User avatar
RYM6746
TwoWaves
TwoWaves
Posts: 186
Joined: Tue Nov 01, 2016 8:58 am

Re: P1166 fuel deviation is this a rare EMI case? PLEASE HEL

Post by RYM6746 »

Hi Ben
Unfortunately I didn't get the old unit back as they keep them for refurbishment, I believe. So I am unable to do any back to back comparison. I probably wouldn't have time to do that anyway due to the time involved and with it being such an intermittent fault. It would appear that someone had been inside the ECU and performed some sort of repair. I can confirm that the car has been working absolutely fine for nearly two weeks now so it is safe to say that it is a confirmed fix. Unfortunately it has left a lot of unanswered questions for us but sometimes that's just the way it goes. The main trouble was the random intermittency, one fault code and that one waveform.
I hope that someone can look at that waveform and if it it holds enough information for them they can enlighten us all. Thanks again for all your help.

User avatar
RYM6746
TwoWaves
TwoWaves
Posts: 186
Joined: Tue Nov 01, 2016 8:58 am

Re: P1166 fuel deviation is this a rare EMI case? PLEASE HEL

Post by RYM6746 »

WELL!!! It looks like i spoke to soon. The customer has driven all this time with no fault. Then returned with the same symptoms and fault code. If anybody has any suggestions they will be greatly appreciated.

User avatar
RYM6746
TwoWaves
TwoWaves
Posts: 186
Joined: Tue Nov 01, 2016 8:58 am

Re: P1166 fuel deviation is this a rare EMI case? PLEASE HEL

Post by RYM6746 »

Still fighting this one. I can however confirm that what you are seeing on the waveform that I have uploaded is not EMI nor is it high resistance causing a voltage drop. I can confirm with no doubt that the control side of the fuel pressure regulator is partially being grounded (when it is not supposed to be) every time an injector fires. That is why the fuel pressure goes through the roof. Because current is flowing through the fuel pressure regulator (when it is not supposed to be). I have spoken to AC tronics and they confirmed that they had found a fault on the fuel pressure regulator circuit of the old ECU. They also said that we have a completely different ECU now fitted on this vehicle. They simply transferred the data from one unit to the other. I have exactly the same intermittent fault and exactly the same fault code. Can anyone think of any reason other than the ECU that can cause the grounding of that circuit every time an injector fires??? In my mind it has to be the ECU. But the chances of 2 faulty ECUs are slim to none. Any input is much appreciated

User avatar
Tronic
TwoWaves
TwoWaves
Posts: 90
Joined: Fri Oct 23, 2015 2:43 pm
Location: Sweden

Re: P1166 fuel deviation is this a rare EMI case? PLEASE HEL

Post by Tronic »

Hi RYM6746!

I was wondering about your statement: "I can confirm with no doubt that the control side of the fuel pressure regulator is partially being grounded (when it is not supposed to be) every time an injector fires." What made you fully convinced that it is like that? It is not that I question it, I am just curious how you found out that it actually are the injectors causing what I see as a DC supply voltage modulated by a low frequency which in fact, seems to coincide with a firing injector, but ...

If this mystery take part in the ECU or before the regulator would be interesting to know, but I believe and most probably, it is on the switching side of the regulator as you confirmed.

Firstly, I am, because I have not even thought about it before, not sure how the ECU react to a voltage drop that feeds the regulator. Would the ECU in that case increase the duty cycle because of current loss through the coil to accomplish an appropriate pressure change or just simply follow the pressure command and with help of data from the high pressure sensor. Let us say that this, as I call it, low frequency modulated DC which is pulse width modulated, drops to roughly 7 volts every 20 ms and the ECU is fast enough to compensate a fast drop like that, can it really make it happen? Well, I do not think so as the PWM frequency seems too low for the rapid changes in the supply voltage that actually is there, unless it tries, gets crazy and run havoc. If it tries, the ECU has the ability to measure coil current and/or voltage.

I can hardly believe that something can modulate the PWM driver in a healthy ECU like that, but who knows, I would certainly not bet my best oscilloscope on that! :lol:

So, what could create a drop every 20 ms? I am thinking of an intermittent bad connection causing a few ohms of resistance between B+ to the regulator or ECU, IE for example: fuses, relays, connectors or wiring problem. If there is, with the sentence above in mind, could one or more inductive loads, except for the regulator, somewhere between the power supply that share same feed, make a 20 ms drop because of the current draw from one of them? I am not at work and do not have any wiring diagram for your car at home, so I am just speculating here...

Well, that was lots of thoughts coming from a real tired brain at late evening. I may bark up the wrong tree, but as you wanted some input and no one have tried for a while ... Hopefully some of the text might be food for further discussions.

Regards
Bjorn

User avatar
RYM6746
TwoWaves
TwoWaves
Posts: 186
Joined: Tue Nov 01, 2016 8:58 am

Re: P1166 fuel deviation is this a rare EMI case? PLEASE HEL

Post by RYM6746 »

Thank You very much for your input.
I have since used 1 input on scope having amps clamp on the fuel pressure regulator circuit.
Another input monitoring 12V supply to regulator
Another input monitoring PWM ground to regulator
Another amps clamp monitoring cyl 1 injector current.
Unfortunately I am unable to get the fault to occur whilst hooked up like this.
But if you pay attention to the waveform I uploaded before the fault occurs you can see a small amount of voltage drop on the PWM side of the regulator which then turns into a large amount of voltage drop and the fault occurs.
When I have my scope hooked up as described above with cyl1 injector current as a sync that voltage drop occurs at exactly the same time and on every 4th event.
So no doubt the voltage drop is caused by injectors.
Now for the good bit.
In order for pressure to rise electrical current has to be travelling through the fuel pressure regulator. I tested this by grounding it.
So the voltage drop/short has to be on the PWM side of the circuit and unwanted current flow has to be occurring for the pressure to rise.
If you zoom right in on the fuel pressure you will see very small rises in pressure every time that small voltage drop/short occurs.
It CANNOT BE high resistance or poor connections or a voltage drop on the 12V supply or on the grounds of the ECU as all of that would NOT ALLOW current to flow through the regulator.
So I have no doubt we have a partial short to ground on the PWM side of the regulator every time an injector fires.
I have replaced both wires and the fuel pressure regulator with the customers permission due to the intermttency of the problem (not firing the parts cannon).
Which in my eyes only leaves the inside of the ECU being the only possible place thst this short can occur.
Unless some one would like to point out what I'm missing.
If you pay close attention to the exact time the pressure goes through the roof its when the voltage drops low enough to carry a significant amount of current to actually seriously effect the fuel pressure.
Lack of electrical current WOULD cause the regulator to shut and pressure to drop
Also looking at the waveform that I had with the amps clamp on the fuel pressure regulator circuit I could see very small rises in current every time the voltage/short occurs on the PWM side of the regulator every time an injector fires further securing my theory.
Using the waveform of when the fault occurs (only half of the information) and adding it to the waveform (all the information but no fault) I have come to that conclusion.
If I have missed any thing that anyone can think of please say

Post Reply