Citroen engines that are still poor at starting from cold but OK at hot starting despite being free of any ignition or starter motor related faults could have compression problems, or problems with the petrol supply. (Petrol pump, automatic  choke, manual choke, or fuel injection enrichment devices.) Engines require  a minimum compression value to give satisfactory cold starting and   this value  increases as the temperature drops. Hence an engine with  borderline compressions can start OK all during the summer but can refuse to start as soon as winter  arrives. Over 150 PSI is OK for all engines though some types of engine design are OK with as little as 110 PSI.  A petrol pump with a poor output on cold cranking won't matter if  there is plenty of petrol in the carburettor float chamber as is normally the case if the  car is used every day, but if leaving the car for a few days results in a long time to start look at the petrol pump. Manual choke faults are easy to  spot – pulling the choke out at the dashboard should close the choke butterfly in the carburettor barrel, whilst still allowing the butterfly to open against a  light spring pressure. The throttle should open at the same time. Automatic chokes that control the butterfly should do the same with a cold engine.  Whether manual or automatic satisfactory operation to here will get  the engine started OK - though it might not run with the same enthusiasm if  the rest of the choke system has problems. Automatic choke systems that don't control a choke butterfly in the  carburettor barrel, are impossible to establish satisfactory operation from external examination - they need to  have the engine started to see if they work OK. These systems can be  converted to manual operation - but first make sure that the choke unit itself is good since manual control over a bad unit may be just as bad as automatic control over a bad unit.  Fuel Injection engines increase the petrol supply to the injectors for a cold engine - computer controlled models use a temperature sensor to  alter the supply via the computer, mechanical injection systems use a temperature sensor to  alter mixture strength via a hydraulic transducer.

     For whatever system if it is suspected that not enough petrol is the  cause of the problem then add about a desert spoonfull of petrol down the inlet tract and if starting is then successfull then your suspiscion is confirmed -  conversly if flooding is suspected then disconnect the petrol pump outlet after flooding is suspected to have occurred and starting will again confirm  your suspiscion. 

Emissions failure.  The most common cause of rich running faults is the lamna sensor or the ECU (engine control unit) both of these parts are relatively simple to check,  to check the ECU strip back the insulation onthe signal wire from the lamna sensor, this wire is usually black. Then hold the bare wire between you finger and thumb then touch your other hand on to the battery positive if the emissions come down then you have a faulty lamna sensor but if the emissions remain the same then the ECU is most likely at fault. For more information on ECU testing click on ECU

The Citroen BX had a fault that had defied all attempts to cure in the last few weeks before it came to us - If accelerated gently from a standing start it would cut out completely but would instantly restart and idle perfectly. But if when the engine was felt to be stalling the accelerator pedal was floored then the cutting out was averted. From the symptoms experienced we were fairly certain that this was a carburettor fault but since it was possible that the fault could be due to a sudden loss of ignition also we first had to isolate the fault to either fuel or sparks before proceeding. By installing a Portable Datalogger onto the Ignition side while roadtesting we found that there was no interruption in sparks when the fault occurred so after tuning the engine and finding no fault we stripped and cleaned the carburettor suspecting an intermittent blockage in the Primary idle circuit. Still the fault remained even after cleaning the carburettor in an ultrasonic cleaning bath. Then while adjusting the carburettor after it's fourth strip down the wiring harness going to the antirun on valve was disturbed and the engine cut out. The fault was caused by the engine rocking backwards on acceleration from a standing start in such a manner as to pull on the harness and cut the engine. If the accelerator pedal was floored the idle circuit was no longer needed because the main jet circuit took over.