Some like it hot

BX Tech talk
User avatar
Way2go
Over 2k
Posts: 7279
Joined: Fri Jan 13, 2006 3:15 pm
Location: RCoBerkshire
x 2

Re: Some like it hot

Post by Way2go » Sat Jun 10, 2017 7:08 pm

Gsmack wrote:
Sat Jun 10, 2017 12:33 pm
So fuel probably isn't vapourising as it's currently winter, plus the car had driven about 80km on this tank before the problem started. I have undone the petrol cap just to make sure I'm not getting a vapour lock. No change.

Today I've cleaned all the inlet side, so air filter, maf sensors, throttle body. No change.
But the problem does now turn up after about 8 mins, also the car cuts out when nearly 'hot' when the door is slammed, any door.

Tomorrow is fuel pump and filter, thanks for the ideas re volume of fuel.

Any knowledge here around the symptoms of a failing catalytic converter? Do you have them in the uk?

Thanks


OK, fuel vapourisation cannot come into the equation as liquid fuel is pumped under pressure around a recirculaling fuel rail and does not become vapour until it passes through to the other side of the injectors. The spray patterns of the injectors are modified by the ecu dependent upon tps and load. The only way vapour could form in the rail is if there is cavitation in the pump or the pressure valve on the rail has failed low, both situations are however very unlikely. As has been mentioned earlier 11v at the pump is unlikely to be a factor as the rail has to pressurise during cranking.

It sounds most likely that the cause is "tracks & hairline cracks" in the distributor which change with heat expansion. Door Slamming could exacerbate this? Alternatively, more seriously it could be wearing/dirt of the maf track which signals the ecu on which injector spray pattern to utilise.

I don't have any experience with Cats on a BX but if faulty or o2 sensor faulty I would have expected that to cause the ecu to enter "limp home mode" not a total cut out.
1991 BX19GTi Auto

Gsmack
New Member
Posts: 9
Joined: Wed May 17, 2017 1:34 pm
Location: Melbourne
My Cars: Bx 19 Tri
Bx 19 Tzi by 2
Bx 19 Tri estate
Bx 16v black

Re: Some like it hot

Post by Gsmack » Sun Jun 11, 2017 1:23 am

Thanks for the lesson on the fuel vapour process.
We have considered a failing distributor, just wanted to check everything else first.
We have used a timing light during the failing point and it still seems to be sparking appropriately, does this rule a failing dissy out?

Does the jetronic le3 have a limp home mode, I thought it was a bit early for that sort of cleverness.

User avatar
Way2go
Over 2k
Posts: 7279
Joined: Fri Jan 13, 2006 3:15 pm
Location: RCoBerkshire
x 2

Re: Some like it hot

Post by Way2go » Sun Jun 11, 2017 2:30 am

Gsmack wrote:
Sun Jun 11, 2017 1:23 am
Thanks for the lesson on the fuel vapour process.
We have considered a failing distributor, just wanted to check everything else first.
We have used a timing light during the failing point and it still seems to be sparking appropriately, does this rule a failing dissy out?

Does the jetronic le3 have a limp home mode, I thought it was a bit early for that sort of cleverness.


Mine uses the Motronic 3.1 with wasted spark (so no dizzy) which I seem to remember has limp home, cannot recall though whether the Jetronic le3 does. But I seem toremember that the Jetronic has an electronic ecu under the seat so may well have. Here's a write up on the efi for the BX courtesy of Citroen DIY that may help understanding:

EFI systems

The two main inputs of the EFI ECUs are the ignition (the signal coming from the distributor) and the actual engine load (represented by the amount of air sucked in by the engine). The ignition pulses, having entered the ECU, go through a shaping and halving circuit that forms regular square pulses whose frequency is half of the ignition frequency. The required amount of fuel will be injected in two installments. This signal serves as the time base for the injectors and the frequency remains unchanged throughout the whole calculation.

The width of the individual square pulses are calculated (or, as we already explained, looked up in a stored table) based on engine speed (the ignition frequency) and engine load (the amount of air sucked in by the engine). For the purpose of measuring this second, different sensors can be used. Earlier systems (Bosch Jetronic) used an air flow sensor (AFS): as air flows through the sensor, it deflects a flap, which is connected to a potentiometer. In consequence, the resistance of the meter is proportional to the amount of air passing through it. Later systems (both from Bosch and Magneti Marelli) used a manifold absolute pressure (MAP) sensor instead, which functions as a simple pressure sender (just like the oil pressure sensor in the engine).

Simpler EFI systems—the Bosch Mono-Jetronic fitted to later 1380 ccm engines—do not use any engine load measurement at all. To reduce the number of components and the costs, these systems rely on the position of the throttle pedal as a subsitute input. This is less accurate than actually measuring the quantity of air entering the engine, but it is much simpler.

Using these two main input signals the ECU determines the base pulse width (tp). This square pulse signal would be enough to control the engine under ideal conditions. However, the operating circumstances of an egine are seldom so favorable, hence, the ECU must carry out additional calculations to modify the base pulse width according to some special requirements.

The first major real life factor is the temperature of the air entering the engine. The cooler the air is, the denser it becomes. To compensate for this difference, the AFS housing incorporates an air temperature sensor (ATS) as well. MAP-based systems use a standalone sensor—depending on its location it might measure the temperature of either pure air or air-fuel mixture. Based on the values obtained from this sensor, the ECU might decide to lengthen the pulse width to allow more fuel, a richer mixture to enter the egine.

Similarly, extreme operating conditions such as idling or full load operation require a richer fuel mixture. The throttle position switch or potentiometer (TS/TP) informs the ECU whether the throttle pedal is fully depressed, fully released or somewhere in a middle position.

Starting the engine in cold weather is an even more special situation. Part of the fuel becomes condensed on the cold engine parts, consequently, a richer mixture is required to start the engine. The ECU monitors both the position of the ignition key switch and the coolant temperature sensor (CTS). If the CTS indicates that the coolant fluid is hot—in other words, a warm start—, there is no need for the longer injection periods.

As soon as the ignition key returns to the normal position, the ECU starts a 30-second warm-up period. In the first second, the ECU adds about 50% of the normal amount of fuel. Until the end of this initial warm-up period, this surplus drops to around 25%. From that point, the fuel surplus is determined by the temperature of the warming engine, as dictated by the CTS. To stabilize the idle speed in a Jetronic-controlled engine still cold, the throttle is bypassed through an auxiliary air valve (AAV). This valve is fully open when the engine is still cold but as the temperature rises it starts to close. On a warm engine it blocks the bypass completely. The air going through the bypass is measured by the AFS, thus it tricks the ECU into providing more fuel. This device is heated electrically, hence, it will close after some time no matter what temperature the coolant has.

Later systems have a similar bypass around the throttle butterfly but instead of a simple mechanical valve, they use an electrically operated solenoid valve controlled by the ECU to decide the amount of air which is allowed to bypass the butterfly.

All these additional correction factors—air temperature, idle or full load, starting, warming up—sum up into an additional pulse width (tm). But we've not done yet. The operation of the injector solenoids depend heavily on the battery voltage fed to them. To compensate for lower voltages, the injection time period must be lengthened by ts.

The total pluse width (also called injector duty cycle) is calculated by summing up the three values we received, the base width, the various correction factors and the voltage correction: ti = tp + tm + ts.

If the revolution is above a specified limit (around 1,200 per minute) and the throttle is closed—this is called deceleration—, the momentum of the car is sufficient to rotate the engine through the roadwheels. To save fuel, the injection is cut off. As soon as the engine speed drops below the limit or the throttle is opened, the injection is reintroduced.

Finally, to avoid prolonged operation at revolutions exceeding the specification of the engine, the injection is cut off above a maximum engine speed (6,000-7,000 rpm, depending on the engine).

Models equiped with a catalytic converter use an oxygen sensor (also called lambda sensor) to measure the oxygen content of the exhaust gas to adjust the fuel mixture to achieve the ideal lambda ratio.

Theoretically, the mixture burned in the engine should contain air and fuel in proportion of 14.7 parts to 1 to achieve ideal combustion. The lambda ratio is simply the proportion of the actual mixture to the ideal. A lambda value of 1 means ideal mixture, values below 1 mean rich, those above 1 lean. By measuring the oxygen content in the exhaust gas the computer can decide how to adjust the mixture to keep the lambda value around 1. The main reason for this is the catalyst, whose efficiency depends heavily on the lambda ratio of the gas fed to it. If the lambda is just a fraction below 1, the CO emission rises sharply, and a little over 1 skyrockets the NOx emission. During cold start, acceleration or full throttle the computer decides on a different mixture but under normal operating conditions it sticks to 1, provided that everything else in the system works well. The sensor does not start working until it reaches the temperature of 350 °C.

Actually, this is the reason of the higher fuel consumption of cars equipped with a converter. Lambda 1 is ideal for the converter and is also ideal from a scientific point of view, however, it is certainly not ideal when speaking about fuel economy or driving dynamics.

There is a final safety circuit in the system. During a crash, when the engine has already stopped, the fuel squirting from the injection system could easily cause fire. Hence, the relay of the injectors is controlled by the ECU, allowing fuel injection only when the ignition signal is present.

All systems—with the exception of Jetronic—have self diagnostic subsystems which constantly monitor the signals from the engine sensors and, should a fault be present, log an error code in an internal memory. More serious errors light up the warning lamp on the dashboard as well to inform the driver of the failure. The error codes can be extracted by either a special diagnostic tool or using a simple algorithm. Whenever the ECU senses the failure of a minor sensor, it omits the signals coming from that sensor and substitutes a fixed default value. This value is characteristic of a hot engine, therefore, cold starting and the warm-up period may be less than satisfactory, however, in normal operating conditions the engine may actually run quite well.

Most systems can also adapt themselves to changing operating characteristics or engine wear. If some of the components have been renewed, the ECU should be re-calibrated, that is, allowed to repeat its learning process. Disconnect the ECU multi-connector for approximately 15 minutes (with the ignition off, this is extremely important!), during which time the ECU resets to the default values. Reconnect the connector and run the engine until it reaches the normal operating temperature.
1991 BX19GTi Auto

User avatar
mat_fenwick
Moderator
Posts: 7279
Joined: Tue Sep 20, 2005 4:08 pm
Location: North Wales
x 13

Re: Some like it hot

Post by mat_fenwick » Tue Jun 20, 2017 11:13 pm

If you are getting the strobe from a timing light then this proves then ignition system; additionally if it had been the ignition module then the rev counter would flick to 0 when the cut-out occurred. So from that evidence we can safely say it's fuel. But why?

A fuel pressure gauge on the rail would be useful, and/or a 'noid' light. The latter shows the presence of injector pulses so will prove whether the injectors are still being pulsed during the fault condition.

You do have a 'limp-home' mode but as W2G states, that would be a safe (i.e. rich) fuel map, not complete cut out.
Image

1993 1.9 TZD Turbo Estate
1996 3.9 V8 Discovery
1993 VW LT35 campervan
1985 Hyundai Stellar V8
2016 Hyundai iLoad

Gsmack
New Member
Posts: 9
Joined: Wed May 17, 2017 1:34 pm
Location: Melbourne
My Cars: Bx 19 Tri
Bx 19 Tzi by 2
Bx 19 Tri estate
Bx 16v black

Re: Some like it hot

Post by Gsmack » Fri Jul 21, 2017 10:22 am

So an update :?
It is a slow process of identifying problems and rectifying, sadly I am not 100% there yet, but it is amazing how well ECUs compensate for 30 years worth of wear and tear.
So what has been not working, below par, or just plain old dirty.
  • alternator - as it got hot the rectifier was failing so was delivering less and less power to the battery and system
  • Connection to fuel pump was poor, now getting a constant 13v to the pump
  • MAF was very dirty and sticky, gave it a good clean, moves beautifully now
  • Throttle body very dirty, including all the little holes, gave that a good clean
  • SAD was investigated, dirty so cleaned it with MAF cleaner because of the plastic, then lubricated it with a dry lube to get it rotating more smoothly
  • coil replaced because the resistance was going up as it got hot
  • distributor cap and rotor reviewed and identified as not the problem
  • new spark plugs
  • ignition module swapped out with another one (old) problem didn't change
  • Coolant temperature sensor that is in the bottom of the radiator investigated because it has a switch mechanism that is connected to the power side of the fuel relay. No change to the problem and no apparent switching or continuity issues with the wiring
  • Coolant temperature sensor (injector) replaced with a new one and wiring investigated (may have created a change will review tomorrow
  • Reset the ECU after each of these main changes
What I have managed to achieve in all this, is identify the problem better. The problem is actually related to warm starting. For example:
  • Car has been turned off for 24hrs, battery disconnected, start car, it will run no problems for 1+hrs, I have even driven it around in circles for 40mins
  • car has been turned off for 24 hrs, battery disconnected, start car, let it get up to temperature. Turn off car, let sit for 10mins, restart car. Run will car for approximately 10mins and then die out
It appears that during the warm start the car stays running too rich and for want of a better word appears to flood the engine till it dies, ie lots of fuel getting dropped out the exhaust, one can tell when the car is running better because the garage smells a lot better.

I still have the issue where opening or closing the drivers door will get it to cut out earlier than waiting for it to self die.
The test for tomorrow is to see whether the car still runs 'properly' when cold since changing the coolant (injection) sensor.

Does anyone have any more ideas?
Also what voltage should one be reading across the primary windings of the coil, I would have expected something in the order of 12v, however I'm getting 1.4v on average, and if that is wrong, would indicate that the ignition module is cactus.
I think the long and short is that there were a bunch of things wrong with the car that just meant the ECU couldn't compensate anymore, I've got to be close to the last one. [-o<

User avatar
Jaba
1K Away
1K Away
Posts: 1454
Joined: Tue Jul 05, 2005 7:54 am
Location: Usually in the garage
x 16

Re: Some like it hot

Post by Jaba » Fri Jul 21, 2017 11:26 am

A few thoughts:
The HT coil is probably reading correctly as it is voltage spikes at each revolution that fire up the ht coil to provide the sparks. You need a scope to see what the actual voltages are.
If its a too rich mixture have you checked the injectors for leakage ?? You can get them cleaned, refurbished and tested .
Is the fuel return pipe clear with no kinks or perhaps internal rust blockages ?
I see you have a cat. The lambda sensor might be playing up and richening the mixture when good and hot.
The Joy of BX with just one Citroën to my name now. Will I sing Bye Bye to my GTI or will it be Till death us do part.

Gsmack
New Member
Posts: 9
Joined: Wed May 17, 2017 1:34 pm
Location: Melbourne
My Cars: Bx 19 Tri
Bx 19 Tzi by 2
Bx 19 Tri estate
Bx 16v black

Re: Some like it hot

Post by Gsmack » Fri Jul 21, 2017 2:14 pm

Hi Jaba
Fair point about needing a scope for the coil.
The injectors don't have any leaks and have had a kit of seals put on them in my ownership, also why would that make any difference between cold start and warm start.
I've disconnected the lambda sensor to see if that made any difference, it doesn't, also I would expect that to be an all the time issue not a warm start only. Also i have previously replaced the lambda sensor, so it's only about 3 years old.

Other thoughts?