| Author | Message | ||
     Udo Hoffmüller New User Username: udo Post Number: 3 Registered: 2-2008 |
Dear fellows! On this forum I found already a lot of themes and things of great assistance for me. But one theme I did not find: Faulty boost control at the TurboR (mine is MCX 35566) - and there is my problem, easy to explain: At exactly 2.000 rpm the voltage at the solenoid valve (signal pressure pipe) drops from about 10 V to 0 V whatever maybe the conditions. If the engine revs at less than 2.000 rpm there is about 10 V from the boost control ECU and the solenoid is open, but at the moment the engine passes the limit of 2.000 rpm - 0 V and the solenoid closes: no more boost pressure. Why always at exactly 2.000 rpm? I would have thought that this had the reason in an easy and very concrete failure!?!?!? The easy possible causes: Brake light is okay and does not cut off the Turbo, and the cable from the cruise control has been disconnected from the plug at the ECU. Every cable that goes to the plug at the ECU has been tested, there is no cable broken or too high the resistance. With the signal pressure pipe disconnected from the solenoid acceleration is as expected, so I know that the system is mechanically without faults. I connected a voltmeter to the air pressure transducer and observed the voltage while driving (voltmeter put on the center console): Engine off/ignition on gives about 2,4 V depending on weather conditions. Engine running, one can know from the manual that 2,4 V (or whatever maybe the voltage for the given outside air pressure close to 2,4 V) means 0 mbar for manifold pressure, less than 2,4 V means depression in the induction manifold while more than 2,4 V means boost pressure. Every 0,25 V increasing Voltage means 100 mbar of boost pressure. I am a gentle driver observing the voltage while accelarating with the hose off the solenoid, at 3,2 V or about 300 mbar boost pressure I take off the gas to avoid too much pressure. With the signal pessure hose fitted to the solenoid the voltage at the air pressure transducer does not exceed 2,9 V which is to be expected: This corresponds to the basic pressure of 200 mbar, admitted under every condition by the boost control ECU (except brake light on and cruise control on). The car runs great with the signal pressure hose off the solenoid, but the boost control ECU cuts off the voltage to the solenoid at 2.000 revs, that means that the solenoid closes and the signal pressure is added to the wastegate pressure which opens therefor and no boost pressure is available (better: not more than 200 mbar) - well, if the hose was fitted. So far, so good. I guess you see that I do understand how the system works - but maybe I am committing mistakes. I tested the remaining units/systems: knock sensors, air pressure transducer. The solenoid itself is obviously working, it is causing the problem without being the guilty part of the system. The air pressure transducer, too, is working fine because I get all the information that is sent to the ECU - and all the data do arrive at the plug of the ECU. The knock sensors have been renewed. They do react at slight hammering on them giving a sudden alternating mV-signal, received at the plug to the ECU. And last but not least: The ECU itself cannot be fautly, I have a second one fitted with the same behaviour. So everything seems to be okay, but there must be a reason that causes the boost control ECU to cut off the voltage to the solenoid at always exactly 2.000 rpm. But where???? It must be a reason within the electrical/electronical system because the ECU decides on the data it gets electronically. Or am I wrong???? Because every concerned indivudial unit (APT, solenoid, knock sensors, brake light, cruise control, ECU itfelf) is working fine and correctly, I am suspecting two remaining things: - the engine speed signal is wrong. But the cable is okay, and the engine runs fine, why should there be a wrong engine speed signal from the engine ECU to boost control ECU? - the knock sensor isolation is faulty so that the ECU receives data that could be interpreted as knocking, from the ignition leads for instance. One more thing: I do not believe that the engine is really knocking. I changed the fuel station many times, and I even changed the lambda-sonde because an old one tends to let the engine run too lean which may result in higher engine temperature and therefor higher sensitivity for knocking. So I do not believe in a really knocking engine, new lambda-sonde and several different gas station did not change anything. But nevertheless the most probable unit causing the problem is - for me - the knock sensors. They do work, they do send there data to the ECU, but what about there isolation against high voltage pulses from the ignition leads? The strange thing with those isolation leads is that they are good from the plug at the ECU to the plug at the side of the engine oil filling lid, but from there to the knock sensors the isolation lead does not arrive!? Where does it start? The knock sensor itself has only two leads, the third one for isolation seems to start "later". But where? How is it fixed or connected? I cannot find where it starts. It does not start at the knock sensor itself, but it must start before reaching the plug/connector a the oil filler lid because there it is! I hope this is an interesting problem and you do understand what I try do describe - and I hope to find assistance because I am "at the end of my Latin". Perhaps one of you has ideas where to go on searching - or even had this problem and found the solution. Thank you in advance, regards Udo | ||
| Patrick Lockyer. Grand Master Username: pat_lockyer Post Number: 752 Registered: 9-2004 |
Sorry if i have come in on this rather on the quick side. Could the car have been used with 4 star fuel or has any work been carried out on the exhaust up front of the cats. Could be that there is to much pressure building up in the exhaust system. Cats or silencers [mufflers] blocking. I will read up some notes that may give other hints with the managment system as there are many other causes. Have you had a code check. What are the emissions reading. | ||
| Udo Hoffmüller New User Username: udo Post Number: 4 Registered: 2-2008 |
Dear Patrick, I guess you are looking in the wrong direction: The Turbo managment has nothing to do with the engine management. There is no way to read out any codes of the boost control ECU. The only connection between engine ECU and boost ECU is via the engine speed signal that is leaving the engine ECU at pin 25 and entering the boost ECU at pin 25. The Turbo ECU is a completely independent unit working with the data from: brake pedal, cruise control, engine speed (sent by the engine ECU), air pressure transducer and knock sensors. From this data the ECU decides the voltage that is given to the solenoid (two different programs depending on kat or non-kat car), this solenoid controls if the signal pressure is lead out (solenoid open) or added to the wastegate pressure (solenoid closed). That's it. Very easy. Of course, if it would have been developed by RR it would be much more complicated, but it is SAAB patented (the Turbo ECU shows the SAAB parts number). My problem is that the solenoid is fed correctly with about 10 V up to 2.000 rpm, but passing the 2.000 rpm it cuts off the voltage to 0 V. What I tested, I described before. The reason for the failure can only be found within one of the components sending data to the boost control ECU. All the components work correctly - only for the engine speed signal I cannot be sure because I don't know how to check the correct value given to the boost control ECU. But missing or wrong engine speed signal does create other faults (following the the manal), not the one that in my case appears. Therefor it is improbable that the engine speed signal is wrong or missing, but who knows ... So the knock sensors, although correctly working, are for me the most probable source of wrong data, i.e. sending data which are interpreted as knocking but without the engine really knocking. This is the region where I am searching - but I don't know what more to do. If I knew more about the knock sensor isolation lead ... But let me see, what you and others think. Regards, Udo | ||
| Udo Hoffmüller New User Username: udo Post Number: 5 Registered: 2-2008 |
Patrick, I forgot to answer your questions: The car uses 98 octan. No work done to the exhaust ever Too much pressure - how should the boost control ECU get those data????? The ECU is actuating because of the data it gets!? The data given by the APT to the boost control ECU is what I see on the voltmeter, nothing strange, working wonderful - with the signal hose off. No engine ECU faults stored - but as already said, the boost ECU has no connection to the engine ECU (except for the engine speed signal) The car just passed the ITV/MOT/TÜV with lambda = 1 - so no unallowed emission Just to answer your questions - regards Udo | ||
| Paul Yorke Prolific User Username: paul_yorke Post Number: 127 Registered: 6-2006 |
Udo, Have you tried without the knock sensors connected / fitted? How about shielding the area you are worried about with aluminium foil? Also maybe shield the distributor ? | ||
| Patrick Lockyer. Grand Master Username: pat_lockyer Post Number: 753 Registered: 9-2004 |
Udo,My quick thought on this was the exhaust back pressure building up [Cats or silencer blocking] after 2000rpm causing the engines knock sensors to digitly back of the turbo boost. Stopping any engine,turbo damage! 4 star leaded will do the cats in as will debri such as exhaust paste, not good for back pressure. Other engine related knocks that are detected other than pinking through low octane will get the sensors working the lower boost. Engine managment ECU may store codes pointing to a certain failing that somtimes help. As do the four gas emission results.CO on fast idle, HC ppm, CO on idle, Your lamda is ok it seems. Richard T may well have pointers to the problem. He is in my opinion an expert on these cars. | ||
| Udo Hoffmüller New User Username: udo Post Number: 6 Registered: 2-2008 |
Thank you Paul, the point, where knock sensor leads and ignition leads seem to be very close to each other is at the back side of A- and B-bank, just where it is impossible to put the fingers in order to handle the leads and aluminium foil. I guess I will have to put my fingers there - but I don't know how!? I tried already to find the knock sensor lead there - and gave up :-( And Paul, you cannot drive with a knock sensor disconnected - for the ECU this would mean missing or wrong data from the knock sensor and the ECU would limit the pressure to the basic figure, i.e. 200 mbar - just what I already have. Thank you Patrick, exhaust back pressure is an interesting idea - but how can I find out if there is too much pressure between exhaust/catalyst and exhaust manifold? And wouldn't this mean that the point at which the voltage drops down to 0 V would vary - a least a little bit? But there is no room left for variation: At always exactly 2.000 rpm the voltage drops down to 0 V. But I am still nearly convinced that I have to search for an electronical solution: Once I felt the car (with the signal pressure hose fitted) accelerating better than before, so I stopped and fitted the voltmeter (in the meantime always preintalled to either the solenoid or the APT) and really: The voltage did not fall down to 0 V at 2.000 rmp, but the behaviour was completely normal over the next about 100 km. But then it started again on the motorway, the voltage dropped to 0 V for seconds, jumped up to about 10 V again, so up and down irregularly for some kilometers and finally was as before: 0 V at more than 2.000 rpm. Isn't that strange? Doesn't this point to a wiring problem? But where? Which lead would be the most suspicious? And nevertheless, when I take off the plugs to check connections from here to there to everywhere - nothing to be found, the current goes where it has to go. I am dealing with this problem already for more than one year and have been talking to Richard several times via Skype. We did not find any solution. But perhaps he has a new forgotten idea. And perhaps am I loosen in my ideas and I am on the wrong way!? Therefor I posted my problem here to see if there is something I did not see. Best regards - Udo | ||
| Paul Yorke Prolific User Username: paul_yorke Post Number: 128 Registered: 6-2006 |
Remove the knock sensors from the head but leave the wires connected? | ||
| Patrick Lockyer. Grand Master Username: pat_lockyer Post Number: 754 Registered: 9-2004 |
Udo,you said " and I even changed the lambda-sonde because an old one tends to let the engine run too lean which may result in higher engine temperature and therefor higher sensitivity for knocking". Unless I am out of date with this type all the lamda's when failed or failing go rich not lean, they send the ECU into a default setting to protect the engine etc. Now to help you I really must have a four gas read out, the fast idle at say 2500rpm and the HC ppm with the others all read and noted. | ||
| Udo Hoffmüller New User Username: udo Post Number: 7 Registered: 2-2008 |
Dear Patrick, maybe I am wrong, but: From the engine ECU I got the message "mixture too lean". I surfed around in the internet and ask in our (german) workshop here in TF, and result was: An old lambda-sonde tends to send a current too low to the engine ECU, therefore the engine corrects this current within certain limits - as mine did. To avoid problems, I changed the sonde - and the error message did not reappear. Since this time there is no kind of error message - well, there is one, but without significance for the problem we are discussing here: "full throttle failure" I do get, I tested it and found that the throttle switch does not respond to "full throttle". I am a gentle driver, the car does not know what "full throttle" means - maybe the contact within the switch is oxidated. But I have already an new switch, one not so far day I will change it - and try to repair/clean the old one. - But as far to the lambda sond ... maybe you are wrong??  | ||
| Udo Hoffmüller New User Username: udo Post Number: 8 Registered: 2-2008 |
Oh - too early sent the post, but well ... I will have to try to get the data you are asking me. But why do you need them? Lambda was exactly 1 some weeks ago, and CO at 2500 rpm was within the limits - but I do not remember the exact value, I talked to the guy at the test station (we go there with three RR/Bentleys and they know the cars) asking him for anything wrong or close to the limits, he only said that there was no problem, everything nearly perfect. But please tell me exactly what you want me to ask for, I will go to a friend's workshop to find out the required data. Paul, today I was on the road about 120 km with the distributors and ignition leads isolated with aluminium foil as far as possible - no change. The voltage drops to 0 V every time the engine passes 2.000 rpm. But really, I guess it is impossible to isolate the leads as required because one cannot see where the knock sensor leads are passing. But while discussing the problems here I got two new ideas: - I will put (temporarly) new ignition leads which go directly from the distributors to the spark plugs without passing the back side of the engine block but crossing simply the rocker cover. I will isolate them with aluminium foil. Shouldn't this separate perfectly the ignition leads from the knock sensor leads? - I will disconnect the knock sensors at the point close to the oil filler cap. There I will connect a second pair of knock sensors (spares) to the Turbo ECU while the sensors fitted to the engine stay out of use, but leaving the "new/additional" sensors simply there protected with some cloth and aluminium foil. Should the Turbo ECU not get the required information to work correctly? At least the sensors will not send any kind of data that might be interpreted as "pinging". I hope, to get a step forward this way, and I will report. But if not ...??? Using the car with the signal pressure hose off, is a pleasure - and I am looking at the voltmeter to avoid more than about 3,4 V - whicht is much less than the security limit of 600 mbar! Is there any problem to do so? Best regards - Udo | ||
| Patrick Lockyer. Grand Master Username: pat_lockyer Post Number: 755 Registered: 9-2004 |
Udo if you could put up the emission results. First the CO% and the HC ppm at idle after the cat has stabilised. Next the CO and HC ppm with a fast idle say at 2500rpm. And the readings of the lamda if the value changes. | ||
| Udo Hoffmüller New User Username: udo Post Number: 9 Registered: 2-2008 |
Patrick, finally I got the data you asked me, but I guess you will be disappointed: - at idle CO is 0.0 %, HC is 0 ppm - at 2500 rpm CO is 0.0 %, HC is 0 ppm When I give one or two short hits to the gaspedal, HC goes suddenly up to 300 or 400 ppm and comes down slowly to 0 ppm, perhaps within 20 or 25 seconds. In the workshop I have been told this was perfect. What is your conclusion? Regards - Udo | ||
| Patrick Lockyer. Grand Master Username: pat_lockyer Post Number: 760 Registered: 9-2004 |
Hi Udo. Just seen the posting,sorry not enough live progressive details to draw a conclusion. Long shot,Rev counter! | ||
| Stefan Morley Yet to post message Username: myupctoys Post Number: 1 Registered: 7-2009 |
Hi Udo, Saw your post from over a year ago. Did you ever come up with a solution. I have a very similar problem only my issue happens at about 1400rpm. Followed a similar path to yourself in trying to resolve the issue. Car performs beautifully in every other regard, and whilst it feels great disconnecting the feed pipe from the solanoid to the wastegate, not ideal when there is probably a clear problem to be resolved. Leaving it disconnected is probably not good long term. My gut feel is the knock sensors are probably the root cause. But given how well the car is running I dont believe it beacuse of real knock events. As mentioned in your post, I'm wondering if its induced from maybe the ignition system. Makes sense given the knock sensors would be a high impedance sensor with a low signal level. Simply disconnecting the sensors doesnt work as this would still represent a high impedance source prone to induced noise from ignition systems. Did you find the source of your problem. Thanks Stefan | ||
| Udo Hoffmüller Experienced User Username: udo Post Number: 32 Registered: 2-2008 |
Hi Stefan, fine to find a fellow sufferer , but of course not fine that you have the problem, too. And: No, I did not find the solution. The car drives fine and the voltage at the solenoid drops to 0 V at exactly 2.000 rpm. In the meantime the engineECU failed and had to be changed, so that one more possible reason is sorted out: The only relation between turboECU and engineECU is the engine speed data sent from/by the engineECU. As the engineECU is new, this possibility can be excluded, I think. You see me helpless. I have been discussing the problem with an automotive engineer over several weeks via emails, but he also gave up. He told me to check meticulously the earth points. As one cannot see where the black wire (earth) of the turboECU is connected to the body, I cutted it off from the plug at the engine oil filler. There are two plugs (hidden/protected) to connect the wires from the ECU to the main harness. I made a direct connection for the black wire to earth to be sure that the ECU has perfect earth. But nothing changed. One strange thing happened when I had changed the engineECU: The very first trip with the new engineECU and the voltmeter connected to the solenoid (positioned on the center console) showed that the turboECU worked correctly! The voltage did not drop to 0 V at 2.000 rpm! It stayed okay all over the trip of 60 km to the north of the island. I was happy and thought that wrong engine speed data had been the reason! And driving back in the afternoon everything was fine again - but then on the motorway at maybe 2.400 rpm the voltage suddenly dropped to 0 V, then back to about 10 V, stayed there for some kilometers, than dropped to 0 again ... some toggling from 10 to 0 to 10 ... and then the same as before: 0 V from 2.000 rpm upwards. I checked again the wire that transmits the engine speed data, but did not find anything wrong. And finally I gave up - although reluctantly, and surely not for ever. I am still suspecting the engine speed data (or the wire) or the isolation of the knock sensors, the latter one especially because I do not see where the isolation cable starts: At the knock sensor itself there are only two wires. But at the plug (fixed to the oil filler) arrive three wires. The upper of the two plugs contains the three wires of each knock sensor and earth of the ECU, seven wires in total. And the wiring diagram does not show where the isolation cable starts. What sounds strange to me is that in your case the voltage at the solenoid drops at 1.400 rpm. Did you check this while driving? How do you know that this happens exactly and always at 1.400 rpm? I never found a hint in the manual that this could happen at 1.400 rpm. 2.000 rpm (or 2.100) sounds reasonable because it corresponds exactly to the basic pressure (200 mbar). But at 1.400 rmp? I will test the pressure at 1.400 rpm. Maybe it is 0 bar in the manifold? But why that? - I will let you know what happens in my car at 1.400 rpm as far to the pressure. - By the way, is your car fitted with catalyst? Did you find out where the isolation cable of the knock sensors starts? One more thing is there to be tested again on my car: I got the impression that the pressure suddenly goes up at about 3.500 rpm. But as my car has the longer rear axle fitted (although the car is with catalyst), I normally don't arrive 3.500 rpm. I will test it with the voltmeter connected to he ATP and driving in second at 100 km/h and then accelerate. Maybe the voltage jumps suddenly from 2.8 or 2.9 V to 3.3 or something about that. - I will tell you, what happens. Best regards - Udo | ||
| Stefan Morley New User Username: myupctoys Post Number: 2 Registered: 7-2009 |
Hi Udo, Sorry to hear you havn't sorted it, also commiserations to hear the ECU needed replacing. Touch wood....... it doesnt happen to me. Initially I went through the process following the manual to fault find the issue. Everything mentioned in the manual tests exactly as advertised within margins, other than the solanoid voltage dropping to 0V in the wrong circumstances. Need to get a bit more analytical in finding the fault. Really the system is reasonably simple with effectively only one closed loop in the APT/solanoid with a couple of extra inputs from the engine speed and knock sensors, plus brake. Guess thats what makes it frustrating. One thing that puzzles me; the APT, solanoid, one of the knock sensors are on the right side of the engine bay. With only one knock sensor on the left side and yet they place the boost controller as far away as possible on the left. Cant quite figure that. Not that it has any moment other than it opens up greater possibilities of induced noise in the system than if thay had placed it closer to the sensors. Yes mine has a catlytic converter. Assume it works as my emissions where tested in March and passed. If it where blocked or burnt out I assume the results would have been different. Chassis number is 1989, the car build started in 1987 but has most of the running gear of an 89 model. Think its the same with most, gestation period is quite long. See your point about the 1400 as opposed to the 2000rpm pressure wise. I was only using the tacho in the car. Pressure wise I opened the close loop of the circuit by placing fixed resisters on the APT cable (apt cable removed). The system still doesnt work. I dont beleive its from the monitored pressure for this reason. Over the next couple of days I will get a CRO onto the signal lines to the Turbo Control unit and see what that reveals. No I didnt look for the shield on the Knock sensors. Initial tests I simply disconnected them. But thats not a very good test if there are induced problems from the ignition system. Once I have a look at the wavefroms to the control unit we may have something of substance to discuss. Will monitor frequency of the ignition to determine the RPM exactly. Dont know the error there is in the tacho. In relation to your unit sometimes working..... I noticed once when at quite a speed, no I wont incriminate myself, that I got a pulseing of power, almost like the turbo was cutting in and out. Tried the same thing statically in the garage and the solanoid voltage flickered at about 3500rpm. Dont quote me on that rpm, but it was high. Tend to get a little distracted when the engine is running at those sort of RPM. Will get back to you if anything reveals itself on the CRO. Regards Stefan | ||
| Stefan Morley New User Username: myupctoys Post Number: 3 Registered: 7-2009 |
Udo, Didnt answer one of your questions, yes I saw isssue at 1400rpm whilst driving. But the issue happens statically in the garage as well. Again not sure on the accuracy of the RPM. One extra thing I noticed, when I get to the RPM I have an issue with, I noticed the voltage on a DMM doesnt simply drop to 0V. Within a fine RPM range the voltage seems to change proportionally to the RPM. (This happens with or without the APT defeated). Almost seems like the solanoid is controlled proportionally by the control unit with a PWM signal. This range appears to be less than 100rpm. Will look closer with the CRO to see if this is the case or what exactly is going on. Stefan | ||
| Stefan Unregistered guest Posted From: 219.90.162.102 |
Udo Did you ever end up figuring out your problem with the Turbo ECU issue. The problem I have is almost exactly the same and I've gone through all the same step you did. Only difference with mine is the solanoid drops to 0V at about 1500rpm. Its not a sharp drop, if you get the revs right it progressively goes to zero, like its regulating pressure but way too early. My APT tests fine, even tried to defeat it with fixed value resisters which prooved the change wasnt coming from this. Nor the knock sensors, nor the braks signal line. As I said almost exactly the same symptoms you saw, plus I got defected in Feb (for tinting), so went over the pits and emissions passed fine. Hopefully you found an issue. (Message approved by david_gore) | ||
| Stefan Morley New User Username: myupctoys Post Number: 4 Registered: 7-2009 |
Udo, I've done a few quick checks but havnt gone to the extent of digging into the wiring of the boost controller yet. However an interesting thing revealed itself. At this point in time not sure the ramifications but.... I monitored the voltage going to the solanoid with a CRO. It does have a form of Frequency Pulse modulation. Almost like the solanoid is controlled progressively more of an analogue action on the solanoid against the internal spring than a simple on and off. The interesting part is, with the solanoid connected, I see the frequency increase, then at the rpm range I have an issue with, the output drops to zero volts. We both know that happens. However when I electrically disconnected the solanoid and monitored the signal that would have been driving the solanoid, this frequency contiunued to get shorter well above the rpm range I have an issue with, ie didnt drop to zero. I tried this with both the feedpipe connected and disconnected just in case. Both circumstances the same. Seems to me either the output driver in the boost controller is faulty, ie not capable of driving the solanoid due to an internal fault in the driver. Or the solanoid is faulty and the driver is smart enough, or dumb enough to go to a fall back position, that being off. Smart enough in that maybe becasue they are driving the solanoid in frequency modulation domain (interesting its not simply pulse width modultion) that they actually monitor the average current to control the solanoid in the analogue domain. So if the solanoid had shorted turns the inductance would drop and the average current would go up meaning the smart driver turned the solanoid off. My solanoid meansures 27.2 ohms, struggle measuring the inductance, not enough resolution as it only comes to .01H or 10milli henries. Inductors with shorted turns are difficult to measure simply with resistance but comparing the inducance of a known good coil and a dud one certainly tells a story. Not sure where I would get a good one to simply replace and try. Anyway wonder if you couldnt try something similar, measure the voltage feeding the solanoid with the solanoid disconnected. Also anyone else reading this post who might have access to a good DMM and a working solanoid in a Bentley Turbo R or RL giving us an idea of the resistance and inductance. Cheers Stefan | ||
| Stefan Morley New User Username: myupctoys Post Number: 5 Registered: 7-2009 |
Udo, Just had a look in the manual. States the solanoid should be between 26 and 27 ohms, so my 27.2 is only just outside. Think inductance would give a better picture but not too many auto workshops would have a meter to check that with any sort of accuracy. Might find an inductor of equivalent resistance and higher inductance and see how the boost controller drives that from an electrical perspective. Stefan | ||
| Stefan Morley New User Username: myupctoys Post Number: 6 Registered: 7-2009 |
Udo, Spent a few hours this afternoon looking around with the CRO. The problem isnt related to the frequency modulation I origionally thought it was. Although it did lead me to looking at ramifications of this symptom. I have a brownout on the supply to the boost controller. Drops from nominal 13V to roughly 9V. Seems to coincide with powering the ignition coil. Unless I read it wrong the Boost controller and ECU share this same supply source. One or other is souring too much current or a poor connection somewhere causing the voltage drop. This brownout can be seen on the rpm input from the ECU to the boost controller. It may be that the boost controller sees this brownout and interprets it as a revolution. Meaking the apparent rpms twice what they really are. My RPM issue is 2180 (=/- 20rpm). Dont know where I got the earlier 1400rpm. BTW I looked at the knock sensor shield. Only one end is grounded and that is in the boost controller itself. The other end isnt grounded. I also shorted the two inputs, made no difference to the issue. I think the crux of the issue is to find the guilty part the ECU is supplying causing the problem. Or find the poor supply/ground connection to the ECU to fix the brown out. Stefan | ||
| Stefan Morley New User Username: myupctoys Post Number: 7 Registered: 7-2009 |
Udo, Found my problem. Easy fix, painful to find. The brownout is the key symptom, and it can be measured on a number of the fuses easily with a CRO. Even a DMM set to AC will show you something in the absense of a CRO. Fuse panel F1 B3 is where it can be easily measured in my case. The fix, the lower of the two ignition amplifer plugs (under the wash bottle) was badly corroded. Cleaned both for good measure. The Boost controller works fine and the brownout has dropped to about 0.5V (from the origional 4V). Corrosion of the socket happened I think becasue there is evidence of a repair to the wash bottle. Assume it split at some stage. BTW I tested all the relays of moment, even dissassemebled the dash and cleaned the ignition key contacts. None of these things had any impact. Cheers Stefan | ||
| Paul Yorke Grand Master Username: paul_yorke Post Number: 363 Registered: 6-2006 |
I Bet that induced a big sigh of relief Stefan! Good work. Happy motoring! | ||
| Udo Hoffmüller Experienced User Username: udo Post Number: 36 Registered: 2-2008 |
Stefan, that is very interesting. I will have a look at the amplifiers! And try to find any brownout at the fuses. In the meantime I detected that the system is working fine from 2.900 rpm upwards! Only between 2.000 and 2.900 rpm the voltage at the solenoid drops to 0V! What is that??? I will follow your way and see what I can find. Best regards and crossing fingers - Udo | ||
| Stefan Morley New User Username: myupctoys Post Number: 8 Registered: 7-2009 |
Udo, If you want to see if the rpm pulse is the issue. Disconnect the speed pulse Pin 25 of boost controller, maybe even cut it with enough lead for an inline crimp. The boost controller will do everything as intended except it wont cut out at nominal 4500, but it will regulate pressure and respond to knock events. If your issue is similar to mine the boost controller will continue powering the solanoid. Hassle is the rpm pulse goes to the Fuel sytem, Ignition system and the boost controller. I think the source of this signal is from the ignition system read from the crankshaft sensor but its not totally clear. Question is why if the car runs fine otherwise; why doesnt this stuffed up signal only affect the boost controller when three systems get the information. Should have a look again and make sure this issue has dissapeared on my system or I've just improved it slightly. The 4V I was talking about was the supply to the controller, not the rpm signal. Tomorrow...... Being fussy, but I dont want this resolution to mask another. Stefan | ||
| Stefan Morley New User Username: myupctoys Post Number: 9 Registered: 7-2009 |
Udo, Quick question. Have a look at your crank position sensor (below distributor next to firewall) how close is the wire to the spark plug leads? Stefan | ||
| Udo Hoffmüller Experienced User Username: udo Post Number: 39 Registered: 2-2008 |
Dear Stefan, in the meantime I opened the plug at the boostECU and isolated pin 25 (enginespeed signal). I connected the voltmeter to the APT and drove: Full pressure up to nearly 3,3 V which corresponds to nearly 400 mbar (car with catalyst) at already 3.200 rpm. Then I connected the voltmeter to the solenoid: No cut off at 2.000 rpm, the solednoid works correctly and starts to close from about 3.200 rpm, i.e. voltage goes down to 8 or 7 V. Conclusion: Without the enginespeed signal at the boostECU the system works fine except for the fact that engine knocking is not monitored. I tried to find out if there was a brown out anywhere and I found to points: The connectors of the thick brown cable that travels round the car (the main current supply). I cleaned the plugs and connectors meticulously, solded the cable to the connectors: no more brown out. But the turbo problem was the same as before. I am helpless. Without enginespeed signal the system works fine, with the signal at pin25 supplied the solenoid stops the pressure at 2.000 rpm which is exactly when boosting starts i.e. 0 mbar so that the system runs with basic pressure as maximum pressure i.e. 200 mbar, confirmed with the voltmeter connected to the APT which shows 2.8 V maximum (or very very close to this value). The only reasonable conclusion I can find is that the data sent by the knock sensors are wrong or interpreted wrong. But I changed them some time ago with no effect. And I have a second ECU which shows the same behaviour. How did you test the signal sent by the knock sensors? Did you test it? What was the result? I do not give up although I do not deal with this problem every day. Best regards - Udo |
