Tdi Loss Of Power (again) - Resettable

[wytonpjs]

Help!

 

I have a 2001 (65k miles) 1.9TDi (115 BHP) that developed an interesting (and ALARMING) problem last week whilst towing a caravan. After about an hour of "hard" driving the car suddenly "lost power" and would struggle to hold 35-40 mph (VERY embarassing on the motorway) - up to that point it had quite happily towed at 55-60 mph. There were no warning indications or anything that alerted to a problem (engine temp OK, etc). I limped into a Service Station, waited 30 mins and set off - all was OK with no reoccurrence for the remainder of the journey and for the rest of that week (not towing). On the return journey towing the caravan it happened again after 45 mins - after a 10 minute wait on the hard shoulder I set off again and it was OK for another hour. This time when it went again, I came to a halt on the hard shoulder, switched off and immediately restarted - engine back to full power! No problems for the remainder of the journey (approx 1 hr).

 

Reading the TIS I note that the turbocharger's vanes have an emergency position (maximum opening cross-section) to obviate engine damage due to execessive boost in the event of a turbocharger control "concern" (eg "in the event of an electrical concern or leak in the vacuum system"). It seems most probable that this was what was happening with the PCM putting the turbocharger into a latched "safety mode". Question is, what is tripping the PCM?

 

I persuaded my local (and very friendly non-Ford) garage to have a play with their SNAP-ON diagnostics kit and it came up with code "17965 - Turbo Pressure Control - Limit Exceeded - Occurs Intermittent". This appears to validate the latching of the emergency limp home mode. Question is what is the probable cause? The SNAP-ON diagnostics do not give any further clues. Outwith of the actual turbocharger the control system appears to be very simple: the PCM controls the Vane Adjustment Solenoid Valve which regulates a vacuum across the vacuum diaphragm unit. Given the car drives OK normally, it is not a vacuum issue and the range of actuator movement appears OK. My thoughts are turning to inputs to the PCM that would cause it to trip the turbocharger. I'm aware that operation of the brake lights will trip out the turbocharger (boost not required during braking) but this is not the problem (this is not even mentioned in the turbocharger section of the TIS). Has anyone had a similar fault and satisfactorily cured it? Does anyone know what other "sensors" might persuade the PCM to "trip" the turbocharger?

 

I am very hesitant in taking this to my local Ford garage who have a poor track record in my eyes - the options here (turbocharger change, PCM change, etc) are all horrendous and with WDS hook-up starting at

[golfmadman]

Have not sussed out how to add a thread yet but if you go to General questions and look at No Go from Golfmadman you will find a load of replies surrounding this sort of problem.

 

Ps, how do you add a thread to another post? :)

[seatkid]

Most likely vacuum pipes that control turbo. See here

 

you can insert a link using the "http://" button on the form.

[wytonpjs]

Thanks SeatKid for comprehensive post. So far I've had a general look at the boost valve, that hemispherical unit on the front of the engine and the small air filter. All hoses, filters, connections appear as new and no end on pulling, flexing, movement throws up any cracks, etc.

 

I'm at a loss anyway to understand what happens vacuum-wise when the PCM trips the turbocharger anyway - the load on the engine usually has not changed very much, it just sort of happens. This makes me think it cannot be vacuum related. Prior to the point this happens the turbo works completely as it is should - its just like someone throwing a switch. A vacuum related problem would surely be present all the time and would affect the normal day-to-day driving. This snag only happens under heavy towing and after some time.

 

Are there any temperature sensors on the turbo or something similar?

[GSMGuy]

I still think it' the T-MAP sensor!!

 

Mike

[seatkid]

My opinion....

 

If you look at the control system it is quite simple. Basically the ECU is trying to maintain a constant boost pressure. It measures the Boost pressure directly from the TMAP sensor, compares it to the desired value then increases or decreases it using the Boost Control solenoid to increase or decrease the control vacuum. The vacuum reservoir is present in the PD models to maintain control stability. The ECU software will be set to trip whenever the Boost Pressure is outside the range expected for too long (a few hundred milliseconds). Other sensors MAF, temperature etc influence the fuelling and not the Boost.

 

If its flagging "positive deviation" - this indicates an overpressure situation. This would eliminate main boost pipework/intercooler/air filter and TMap sensor. I would think its either a control vacuum problem which is either blocked pipework (but not leaking), a sticking boost control valve or a sticky vane actuator mechanism

 

A "negative deviation" on the other hand is an underpressure situation, which could be leaking pipework both main and control, leaking intercooler, blocked air filter, a faulty Tmap sensor, a sticking boost control valve or again a sticky vane actuator.

[wytonpjs]

Hi - Update!

 

Good root around the vacuum pipes, etc - no apparent snags. Further internet trawl revealed sheer number of people experiencing the 17965 problem - big discussions on Golf/Jetta/Bora websites about exactly the same problem. In parallel last week I ordered a KEY-USB from SK Pang. Arrived Sat - got time on Sunday to play. On the engine controller one fault - surprise, surprise, 17965 - Charge Pressure Control: Positive Deviation P1557 - 35-10 --- Intermittent. Interesting snag on the Auto Trans as well 0030 - Transmission Fluid Temp. Sensor (G93) 30-10 - Open or Short to Plus - Intermittent (which I take to mean that I have overheated the auto box on several occasions) - possible link? Anyway, I used Vag-Scope on the Engine Block 11 to measure requested versus delivered boost. After an initial lag, the turbo increasingly delivered in excess of that required by quite a margin - never caused the ECU to trip but there was a significant and rpm-dependant increasing gap between the lines. Given the gap was increasing and that it was a positive overpressure it is suggesting to me that the fault is the boost control solenoid valve - the one at the back of the engine.

 

Now to the buggeration. Tried this morning to log the outputs of block 11 so I could post the plots. Had great difficulty in detecting the controller, then, when I switched into basic settings strange things started to happen. The desired boost window switched to off, the car's rpm started to increase and I got all manner of warnings flashing in the inst panel together with a flashing ENGINE WARNING message on the red display. Now some history; after running the VAG-COM on the Galaxy on Sunday I hooked it up to my wife's S-reg Ford Ka to see if it could read generic OBD codes - no joy. Coincedence, or have I buggered it up in doing this? Interestingly, when I initiated VAG-COM this morning I did a test from the options page and whilst it found the interface and said everything was OK, there was only OK against K1 and K2 - there was no CAN flag. I intend to reload the VAG-COM programme today (regrettably, the wife has the car today so I cannot get at it again until tonight).

 

Thoughts/ideas on original and VAG-COM snags?

 

Peter

[seatkid]

OK, so you've identified too much boost.....

 

Step 1. Eliminate vacuum pipework (scientifically)... 2 possibilites

 

the vacuum isnt being released quickly enough when the correct boost is reached - the pipework from the ATM side of the boost control valve is blocked. This is easy to check, just pull the pipe from the ATM connection on the boost control valve (one nearest the middle of the solenoid) and see if it makes any difference.

 

the vacuum is buiding too quickly because there is a blockage in the pipework to the vacuum reservoir - collapsed/squashed/kinked pipe - to check this, with ATM pipe still off as above, (engine off) pull the pipe from the vacuum reservoir (not sure if this can be done) and blow into that end and check if its blocked.....

 

Step 2. Either the boost control solenoid or the vane actuator mechanism is "sticky"

 

First I would look at the vane actuator mechanism (as the most likely)......this needs to be checked for free and easy operation....

 

post to be completed after some further research........ :(

[GSMGuy]

Peter / Seatkid, have a look at my last post in the thread SK quoted above http://www.fordgalaxy.org.uk/ford/index.ph...ST&f=62&t=5921# It shows how to check if the actuator is working ... Mine seems to be ok... I am tempted just to go and buy a T-MAP for the hell of it - Anyone got a price on it?

 

Mike

[seatkid]

Peter / Seatkid, have a look at my last post in the thread SK quoted above http://www.fordgalaxy.org.uk/ford/index.ph...ST&f=62&t=5921# It shows how to check if the actuator is working ... Mine seems to be ok... 

 

Excellent post! :blink:

 

I am tempted just to go and buy a T-MAP for the hell of it - Anyone got a price on it?

 

Time to put your money where your mouth is?.... :angry:

 

Surely you could persuade someone for a quick swap?

[wytonpjs]

seatkid/GSMGuy - many thanks for your excellent posts and thoughts so far :blink:

 

Managed to get VAG-COM working again last night and to log the required boost against delivered boost against speed and time. I am not sure how to display this graph directly in the post so I have attached a screenshot of the results (Excel file is available if required) :lol:

 

Given that the delivered boost (red) follows the desired level (green) I believe this proves that the fault is not vacuum related - there is no lag that a blocked pipe would cause. Thinking about the T-Map sensor, the T-Map is capturing the pressure and is triggering the ECU to shutdown when the maximum positive limit is exceeded - I think this is working OK. My simplest thoughts say to me that this must be the boost solenoid control valve delivering too much output in response to ECU request :angry:

 

I standby to be shot down but other than finding that this car has a "funny" boost chip fitted (interesting and dangerous way to get extra power) I'm tempted to change the solenoid on spec :D

 

There is a dawning realisation that if I can correct this snag, the Galaxy going to be a lot poorer in performance terms given the 20% extra boost it is now getting at 3000 rpm :blink:

[seatkid]

The system is a closed loop system. The Tmap is the feedback signal that the ECU uses. That is to say, if the boost is too high, the ECU will automatically back it off and vice versa.

 

But now you mention the previously unmentioned.....funny boost chip

 

THIS is why your TMAP signal is too high! And probably why the ECU keeps tripping.

 

Your plot shows that "control" of boost is present, but very obviously the boost chip is messing about with the TMAP or other parameter to give you increased boost.

 

Either ditch the chip, or get it tweaked down.

 

GSMguy. I think possibly the same applies to you (chip ), but also from a previous post you obviously had some restriction in your ATM pipe also, amplifying the problem.

 

OK guys, change the TMAP sensor if you must, but don't say I didn't warn you......

[wytonpjs]

Sorry seatkid if I didn't make myself clear :( - to my knowledge I don't have a boost chip fitted - it is just a possible explanation that has occurred to both me and my local garage crew. I'm going to have a ferret around the ECU to positively prove that it hasn't previously been tampered with but to my knowledge, this is a straight car :blink: I'm not sure that such a chip would work as the fuelling is being controlled by the MAF and it won't unless this was tweaked as well in the ECU get enough fuel to cope with the boosted input :(

 

I'm also positive that it is not the T-MAP sensor - as you say, it is a closed loop sensor that is controlling. This leaves me with the boost solenoid valve that is, I believe, responding to commands from the ECU but progressively giving the wrong amount of vacuum :angry:

 

Tonight I'm going to go over its connections again but it looks like I'm off to VAG tomorrow to get a new one. Is the part number 1J0906627?

 

Many thanks

 

Peter

[wytonpjs]

Oh dear :( Have just had a coffee with a colleague and run through the system once again.

 

The one thing I had not thought about :angry: was that the actual boost line in the above graph comes from the TMAP sensor. Now if this is faulty and erroneously reporting the boost you get the situation where the car gives less power than it should do and trips out under a lighter load because it is telling the ECU that the pressure is higher than it is actually :(

 

Consequently, and after thinking through what would happen if the vacuum lines were leaking/blocked, I have allowed myself to be persuaded that it could be a faulty TMAP sensor.

 

Talk about making a decision :( Next thing I'll be wearing a skirt :blink:

 

I have no idea whether the TMAP can become obstructed, fouled, etc but tonight I will have a look at it - at least with VAG-COM I can get a direct result as to whether the indicated boost is following the desired boost.

 

The saga continues but I am coming round to GSMGuy's corner :(

[tim-spam]

You could try measuring the manifold pressure independently with a manometer, and comparinf this to what the TMAP sensor is telling you. I don't know if there's a suitable tapping point already there, or if you would have to drill one - you would then of course have to plug the hole afterwards. Anyway, it would seem to me that after having read several posts on this topic, independently measuring the boost pressure would confirm whether or not an overboost is happening or whether the sensors are misreading. This would help to eliminate / confirm the diagnosis.

[seatkid]

Just thinking out aloud....:(

 

Q1. How does the ECU know whether boost pressure is too much or too little?

 

A1. TMAP sensor.

 

Q2. So does the ECU have another way or recognising "real" boost is out of tolerance?

 

A2. Only if it compares the boost control solenoid demand is above or below preset limits for the selected condition....but surely VAGCOM could tell you that?:angry:

 

Q. How can a TMAP sensor go faulty and the boost control still work?

 

A. Well, it could drift (gain or offset) leading to possible Q2/A2, or the inlet port could become gunged up (reports of inlet manifold gunge common....)

 

Hmmm.....maybe.....have you taken it off and checked its inlet and the manifold?

 

What would be useful would be typical voltages at TMAP....anyone?

[seatkid]

Can't believe it, I had exactly the same problem in exactly the same situation, in exactly the same car, bhp and year, on holiday

 

Must be an attack of schizophrenia

 

:( :(

 

:angry: :)

 

:( :blink:

 

anyone seen Harvey anywhere?:( :(

[GSMGuy]

Peter, I have replaced the boost control solenoid on mine, and it made no difference whatsoever... Whereabouts are you as you can borrow my old one which is working fine to test...

 

I really am thinking Tmap on this, as SK suggested, if it has drifted in a + way it could be telling the ecu "too much boost" and shutting down the turbo via the Boost Control solenoid...

 

I am tempted to go and get a boost gauge to check actual readings, anyone got any idea what max boost should be? I have heard around 18psi??

 

Mike

[Guest marcusheawood]

Hello, unless I'm very much mistaken Gunge could well be the key here.

 

Sounds to me like de-gunging the sensor, sensor drilling, and plenum with carb cleaner sounds like the way forward.

 

A nasty job (could that be why all the prevaricating around the bush?) that usually yields good results in pretty much any fuel-injected engine (diesel or otherwise).

 

Until you get rid of that damn engine breather it will continue to gunk up until kingdom come.

[pacificgreen]

Thanks for all your help, didn't realise it was an exclusive club.Thought everyone was welcome! Shall not bother you all again.

[GSMGuy]

Hey Pacific Green... What's up??? Didn't see you posting in this thread... But no, it's not a private club, nd allthe info is there for all to see...

 

Take a chill pill!!!! :angry:

 

Mike

[wytonpjs]

pacificgreen - stop be such a WOOS :( Everyone is welcome - it is how we learn :( Please repost your original message as I did not have time to read it properly as I left the office - pm me it if you want.

 

Sorry for the delay in tonight's update but I've been busy - also, damm PC lost one of my drafts of this message so I'm keeping my fingers crossed :angry:

 

Went for TMAP as promised tonight. Totally gunked up with oil - gave it repeated wash in trich (no carb clean to hand) but it did seem to clean up. Then ran VAG-COM again and logged/plotted results. In general terms the plot remains of the same form - increasing excessive boost vs demanded boost with speed. HOWEVER, max boost is lower at top revs and if anything, this time it blips almost perfectly in time with the demand - a much more faithful and responsive picture :blink:

 

On the strength of this, I'm going for a TMAP change tomorrow - I note that as well as the VAG part no (038 906 051B) there is also a Bosch Pt No (0 281 002 399). My local crew will try the factors and if they have stock, will let me have it on a trial or return basis :( :( :(

 

If this works, here's some pictures of the VAG-COM plot and the oiled-up TMAP sensor:

[seatkid]

Thanks for all your help, didn't realise it was an exclusive club.Thought everyone was welcome! Shall not bother you all again.

Hey? What happened to his post(s)? :(

 

Must have been expecting an instant response......He can get one of THOSE from his dealer :angry: ........we take a little time longer... we like to think about the question(s)....and like I do have to work and eat occasionally... :blink:

[wytonpjs]

Bugger :angry: ONLY one picture uplifted!

 

Here's the TMAP:

[wytonpjs]

TMAP again :angry: