An N43 or an N53 that runs on three cylinders or refuses to start at all often leads back to the MSD80, and the injector output inside that case is the first suspect. The unit sits on the intake side of the engine and it fires each piezo injector from its own driver stage on the board. The order below keeps you out of the cylinder head until the wiring has answered.
The report is usually a shake at idle that clears above two thousand revs, a check engine lamp that blinks under load, and a car that has lost its pull on one bank. Others arrive on a truck after a cold start that never caught, with the smell of fuel at the tailpipe. Ask whether the fault came on over a week or in one morning, since a stage that dies slowly leaves a different history from one that fails at a key on.
Write the fuel level and the last refuel on the sheet. Water in the tank gives a misfire that wanders between cylinders and it costs a day on the bench. Ask about work near the injectors or under the valve cover in the weeks before, because a pin pushed back in a plug during that job reads exactly like a dead output stage until the connector is opened.
Connect the tester, read the DME, then open the misfire counters per cylinder rather than the summary page. A single cylinder that counts up while the others stay at zero points at one driver stage or one injector. Counters that climb on two neighbours in the same bank point at the supply that the board shares, and that changes the measurement you make next.
Save the identification page and the injector classification numbers before anything is unplugged. Those numbers live in the DME and they are what a replacement unit is coded with. A car that arrives with the numbers already missing has had a unit swapped elsewhere, and the misfire may belong to the coding rather than to the hardware in front of you.
The output stage cannot fire an injector on a feed that sags, so the supply is measured first. Back probe the permanent feed and the ignition feed at the unit with the plug fitted and the ignition on. Each should sit at battery level. Load the permanent pin with a bulb and read again, since a joint that shows twelve volts open circuit drops away under a small load and the driver stage starves at every pulse.
Put the meter on millivolts and read from the ground pins of the unit to the battery negative post with the engine cranking. The number should stay near zero. A drop that rises as the starter turns moves the processor reference under load and the outputs behave in ways no fault code explains. Clean the earth point on the block and repeat the reading on the same pins.
This is where the fault is separated from the fuel system. Fit a current probe or a high ratio voltage probe at one injector connector and crank, then compare cylinder one with cylinder three on the same trace. A healthy driver gives a short charge pulse with a clean edge. A dead stage gives a flat line, and a stage that is shorted inside holds the line down and keeps the injector open.
Repeat the trace with the engine running at idle if it starts, because some stages fail only when they are warm. Watch the pulse width at the same time. A width that grows far beyond its neighbour while the counters climb on that cylinder is the board compensating for an output that no longer delivers, and it confirms the fault without opening the case.
To rule the injector out, move it to another hole and read the counters again. On these engines the injector carries a classification value that the DME needs, so the swap is coded before the test, otherwise the misfire follows the coding and the answer is worthless. A misfire that moves with the injector clears the board. One that stays on the hole points at the output stage.
Look at the tip of the injector you removed. A tip washed clean by fuel and a plug that comes out wet on the same hole mean the driver held that injector open, and the injector may be damaged as well. That case is written on the sheet before the unit travels, since a repaired board fitted over a flooded injector brings the complaint straight back.
The injector loom runs in the hottest part of the engine and its insulation hardens with age. Flex each branch with the engine idling and watch the misfire counters on the live page. A counter that jumps when the loom moves names the wiring. Open the plugs and look for a green pin, a pushed back terminal or a crack in the shell that lets oil into the contact.
Measure resistance end to end on each injector wire with both ends unplugged, and shake the branch while the meter is connected. A reading that steps while the loom moves is a broken strand inside the insulation and it will not show on a still test. Compare each wire with its neighbour in the same bundle, since these looms fail in the same place on every car.
On the bench the unit is fed at thirteen and a half volts and each output is loaded with a test injector or an equivalent load, then fired from a signal generator at a fixed width. A good stage holds its pulse and its current stays flat. A stage with a short draws current the moment the supply is applied, before any signal arrives, and it heats within a minute.
A stage that is open shows nothing on the load at any width, and the current meter stays at the idle figure for that channel. Record the channel that failed along with the load used, because a unit that was tested without a load proves very little about a driver. Keep the supply on for several minutes after the test, since a board that drops out when warm passes a cold test and fails in the car.
Fit the unit and connect the battery, then read the identification page before the first start. Write the injector classification numbers and any coding the car needs. Clear the adaptations and let the engine idle for a few minutes before you touch the throttle. Read the counters again at idle and after a short road test, since a repaired stage that is marginal shows at the first load and nowhere else.
Write the feed and ground figures on the sheet, with the pulse widths per cylinder and the bench result. That page makes the second visit short when the car returns with a different misfire. Check the fuel trims before it leaves, because a stage replaced over a weak high pressure pump hides the pump for another month.
Judge the repair on a cold start rather than on a warm idle in the workshop. Leave the car outside for a few hours, start it with the tester connected and watch the counters on cylinders one and three while it settles. Then load the engine on a short climb and read them again. A stage that is going to fail shows at one of those two moments.
| Symptom | Likely cause | What is measured |
|---|---|---|
| Misfire on one cylinder only | Injector driver stage or injector | Charge pulse at the injector plug, compared with a neighbour |
| Misfire moves when the injector is moved | Injector or its coding | Counters after the classification number is written |
| No start with a wet plug on one hole | Output held open inside the unit | Current at the injector connector before any signal |
| Counters climb on two neighbours | Shared supply on the board | Feed pins at the DME plug under load, ground drop while cranking |
Yes. A pulse trace at the injector plug and a counter comparison across cylinders show a dead or shorted output, and a swap of the injector confirms which side of the plug the fault sits on.
The output for that hole is not delivering, so the replacement injector is being fired by a stage that fails under load. The pulse at that plug is flat or held down, and the bench test confirms it.
It does, and the car runs badly without them. The classification value on each injector is written into the unit before the first start, and the counters are read again afterwards.
Yes. An output held open washes the bore and dilutes the oil, so the car is left standing until the unit is repaired and the oil is changed.