The engine won’t start. What should you do?

One of problems you need to solve most often during engine diagnostics is search for reasons why the engine won’t start.

This question comes up regularly discussed on Internet forums, so I want to share my experience and dot my i's and cross my t's on this issue. The most important thing is to understand what the motor requires to operate normally and make that the basis for your investigation.

In other words, if you are looking for the reason why the motor won’t start, you shouldn’t begin by replacing the fuel filter or inspecting a connector of the crankshaft position sensor, but rather something completely different. Let’s make this issue absolutely clear.

What does the engine need to operate?

A long time ago, At the Wheel magazine described it with a simple phrase: “either there is nothing to burn or there is nothing to ignite.” Generally, this is correct. I would only add: either there is nothing to compress or the ignition timing is wrong.

You need to understand the essential thing: on a modern engine equipped with an electronic control system, fuel supply and sparking are what this system ultimately produces. Any inspection must start with these “global” factors: fuel supply and sparking. All defects ultimately come down to these two. Therefore, we check from the top down rather than the other way around.

Note that we will speak further about engines that lack direct injection into the cylinders. The logic guiding the investigation into causes for a direct injection engine not starting is significantly different and requires, above all, proper use of a scanner.

Check the “spark”.

It is not enough to simply unscrew the plug and to turn the motor after connecting the plug to a high-voltage wire and the ground. The spark gap is such that at atmospheric pressure it will fail even if the ignition coil is half-alive. Furthermore, the quality of the spark cannot be estimated by sight. We must conclude that the spark can only be fully checked with an arrester. There is even a better option, which is to connect a motor tester to the wire and evaluate the fault based on its oscillogram. But this should be done in very doubtful cases. In most cases, using an arrester is enough.

Check the fuel supply.

In order to check, the fuel supply, you must first connect a fuel pressure gauge and check the pressure in the system. Second, you should connect a 5 W lamp to the removed injector connector and check if there are pulses while the motor is turning. Why the lamp? Because an LED will show pulses even if there are serious problems in the injector circuit, e.g. if the voltage on the injectors is 6-8 V instead of 12 V.

Install the pressure gauge, the arrester, and the continuity test lamp. Crank the motor with the starter. This is what’s critical. This is what you should start with. What’s missing? Depending on the result, you know what to investigate next.


This is the simplest case. Make sure that there is petrol in the tank, check for blockage in the fuel pipes, and verify there is power being supplied to the pump. The most common cause is the lack of power. This can result from triggering of the inertia fuel cutoff switch, which is designed to block fuel supply in a car accident. The lack of power to the fuel pump can be caused by poor-quality components in a car alarm, the lack of a standard ground in the connecter, or simply a break in the wire from the fuel pump relay to the fuel pump connector.


This is a more curious case. Here I need to say the following. Most cars are equipped with a standard antitheft system known as an immobilizer. Depending of the design, the immobilizer blocks either fuel supply or fuel supply and ignition. In either case, the fuel supply is blocked.

This is because if ignition is blocked but fuel is supplied, then a failed attempt to start would result in, first, fuel being discharged into the atmosphere and, second, the possibility of an explosion in the exhaust system during the next successful start. This allows us to say with certainty that the case in question has nothing to do with the immobilizer and that the cause is only the ignition system. Depending on the system’s structure, we then investigate further.

Check the ignition module – use an oscilloscope or motor tester to check the power supply, grounding, control pulses from the ECU. In the same way, check the distributor – verify the presence of all the required signals, in accordance with the wiring diagram; also, make sure that the coil is in good condition by replacement or by using a motor tester. If this is a COP system, check power and grounding on all coil connectors. There may be none. The problem can be ECU-related, but you can see if the ECU failed only after a thorough inspection with a motor tester. This is known as pin control, were we check for the presence of all input and output signals.


This case requires painstaking work, primarily with the scanner. Connect the device. Check if there are immobilizer-related fault codes in the unit’s memory. If there are, there is no point in investigating further without solving the immobilizer issues. If there is a fault code that might affect starting in any way (an incorrect signal from the camshaft position sensor, an incorrect signal from the crankshaft sensor), the problem must also be eliminated. The matter may be related to an incorrect position of the entire belt or the valve train circuit. Many control systems simply block the engine from starting in these situations.

If there are no errors in the ECU’s memory, we should focus our attention on the data stream. Our first concern is whether the engine speed is displayed during rotation. If yes, then the ECU “sees the rotation”. If no, then it is likely there is a problem in the setting sensor circuit (most often the crankshaft or camshaft position sensor).

Let us assume that engine revolutions are okay. Next, check injection timing and the ignition advance angle during rotation. If they exist, then the ECU is trying to open the injectors and give a spark. In this case the problem is most likely related to electrical wiring from the ECU to the respective assembly. If not, the ECU is deliberately not opening the injectors, e.g. I’ve encountered a case in my practice when the unit didn’t open the injectors due to the incorrect installation of the valve train circuit on a Toyota 1NZ engine (and the resulting discrepancy in the crankshaft and camshaft position sensors’ signals).


It’s quite likely that the problem is immobilizer-related. Though, of course, we cannot be certain. There might be a wiring defect or a glitch in the installed alarm system. All of these problems can be detected using a continuity test lamp and a multimeter.


This is the most curious case. In this situation, you usually have to check the fuel by draining it into some vessel from a fuel rack. In the tank, you might find water or diesel fuel – anything really – instead of petrol. Try igniting the drained substance or just smell it.

If it is actually petrol, prepare yourself for further painstaking investigation with the motor tester. Using the pressure sensor, you need to check the actual ignition time position relative to the top dead centre (TDC) and verify the correctness of the valve train phase setting.

How to do this is a subject for a separate discussion, but you can read about this in general terms in the article Analysis of the In-Cylinder Pressure Oscillogram or see in the Petrol Engines Diagnostics training video.

It’s not uncommon for sparking to happen somewhere at the exhaust rather than prior to the TDC, for example. The cause is a rotated setting disk or an incorrectly installed distributor.


A plugged catalytic converter. Check this by unscrewing the lambda probe or one of the plugs. If you unscrew a plug, observe safety precautions – when you start the engine, exhaust gases will be discharged from the spark plug hole.

That’s all, it seems.

I will repeat the essential point: you should begin your investigation from “global” signs: presence/lack of fuel pressure, control pulses on injectors and sparking, and only after that should you drill down into specifics.