[keithsketchley]

Does anyone know the sequence and timing of the ECM’s use of the cam and crank shaft sensors during start? I am troubleshooting a no-start.

The description in the “Engine Startup Mode” section on page 14-90 of the 1994 service manual repeats itself without clearly explaining the sequence of using the cam/crank shaft sensor information. It says both energizing and de-energizing the ASD and fuel pump depend on “receives the sensor signals”, the de-energize action occurring if they are not received within one second of something it does not identify.

(The ASD relay provides common power to the injectors and ignition, from the ignition switch, the ECM switches ground to each injector.
The manual says “The PCM energizes all six injectors until it determines crankshaft position...”, which it does within one second.)

So, my question:
Does it open all injectors on receiving some signal (pulses), then start sequencing injectors on having enough information to be sure of piston position and cylinder ID? (But stopping fuel delivery if not getting enough information.)

 

[keithsketchley]

(More info:
“Signal” could in theory be that sensor is alive (which could be detected before cranking) or is detecting slots or ECM has detected meaningful sequence of slots sufficient to determine positions.

The only knowledge the ECM has to know the engine is being cranked comes from the
cam/crank shaft sensors. It does not provide fuel until the engine is being cranked. But it has to know the engine is being cranked, which can only come from the position sensors. Guessing cranking speed of 200rpm, or 3.33 rps, it is conceivable it can determine position within one second which is about 3 revolutions. So the period of opening all injectors is short. Recall these computer-controlled engines start quite quickly compared to carburetted engines.

Somewhere in there is ignition, which also depends on knowing crankshaft position. Ignition timing is more critical than fuel timing, since it determines when the push of combustion occurs thus which way the engine turns.

The ECM needs both camshaft and crankshaft position as this is a four-cycle engine with two TDC physical events per crank revolution for each cylinder, so camshaft position relates to timing of the power stroke that the ignition and fuel are timed for (camshaft rotates at half crankshaft speed).

The ECM uses the camshaft position to know which cylinder is near TDC for injection and ignition, and the crankshaft sensor for precise location.
(The sophisticated critter even uses the number of revolutions since cranking was initiated as one of several parameters used to vary fuel delivery.)

(Details of the no-start:
The vehicle had begun to stumble on occasion, not taking load unless the rpm was above 2000 rpm. It would suddenly start running normally.
Finally it would not start.

Then when start attempted it would fire and appear to be accelerating to idling on its own, but die. Immediate repeat attempts to start, including continued cranking, would give only sporadic firing.
But if left for a while the same sequence would occur on attempting to start.

This seems like lack of fuel.
Despite normal fuel px and indications that the ASD relay is working (and obviously the fuel pump relay is working – they are closed by the ECM using IGN ON power).
Fuel for the initial response might be from the initial opening of all injectors, or from a slow leak from an injector (px reduces very slowly, which could be leak through check valve in pump).

My current suspicions are the crankshaft sensor (given its location that could result in wiring damage or metal sticking on its magnet, and that the transmission was re and re several years ago), and the MAP (which is used by the ECM to set initial fuel strategy, along with coolant temp sensor and TPS – both replaced due condition).

Nevertheless I am crawling through all wiring, doing what sensor tests I can, checking relevant fuses and relays, and checking other parameters.

 

[KOG]

Cam sensor determines which cylinder to fire and inject, crank sensor tells it when to fire. All injectors are never open at the same time, they are always triggered in sequence.

I once had a no start due to cam sensor plug having corroded, replugging it solved the problem.

 

[ImperialCrown]

CRANKSHAFT POSITION SENSOR

The crankshaft position sensor detects slots cut into the transmission driveplate extension Crankshaft Position Sensor There are 3 sets of slots. Each set contains 4 slots, for a total of 12 slots Timing Slots Basic timing is set by the position of the last slot in each group. Once the Powertrain Control Module (PCM) senses the last slot, it determines crankshaft position (which piston will next be at TDC) from the camshaft position sensor input. The 4 pulses generated by the crankshaft position sensor represent the 69°, 49°, 29°, and 9° BTDC marks. It may take the PCM one engine revolution to determine crankshaft position.
The PCM uses crankshaft position reference to determine injector sequence, ignition timing and the presence of misfire. Once the PCM determines crankshaft position, it begins energizing the injectors in sequence.
The crankshaft sensor is located on the passengers side of the transmission housing, above the differential housing Crankshaft Position Sensor Location The bottom of the sensor is positioned next to the drive plate.

CAMSHAFT POSITION SENSOR

The camshaft position sensor provides cylinder identification to the Powertrain Control Module (PCM) Camshaft Position Sensor The sensor generates pulses as groups of notches on the camshaft sprocket pass underneath it Camshaft Sprocket The PCM keeps track of crankshaft rotation and identifies each cylinder by the pulses generated by the notches on the camshaft sprocket. Four crankshaft pulses follow each group of camshaft pulses.
When the PCM receives 2 cam pulses followed by the long flat spot on the camshaft sprocket, it knows that the crankshaft timing marks for cylinder 1 are next (on driveplate) When the PCM receives one camshaft pulse after the long flat spot on the sprocket, cylinder number 2 crankshaft timing marks are next. After 3 camshaft pulses, the PCM knows cylinder 4 crankshaft timing marks follow. One camshaft pulse after the 3 pulses indicates cylinder 5. The 2 camshaft pulses after cylinder 5 signals cylinder 6 Camshaft Sprocket The PCM can synchronize on cylinders 1 or 4.
When metal aligns with the sensor, voltage goes low (less than 0.3 volts) When a notch aligns with the sensor, voltage switches high (5.0 volts) As a group of notches pass under the sensor, the voltage switches from low (metal) to high (notch) then back to low. The number of notches determine the amount of pulses. If available, an oscilloscope can display the square wave patterns of each timing event.
Top Dead Center (TDC) does not occur when notches on the camshaft sprocket pass below the cylinder. TDC occurs after the camshaft pulse (or pulses) and after the 4 crankshaft pulses associated with the particular cylinder. The arrows and cylinder call outs on Figure 4 represent which cylinder the flat spot and notches identify, they do not indicate TDC position.
The camshaft position sensor is mounted in the front of the timing case cover Camshaft Position Sensor Location

 

[keithsketchley]

Thanks, that describes encoding but not full sequence of how ECM decides when to start providing fuel and when to stop if cam+crank data not good.

The service manual, which your graphics is from, does not show the actual position of the camshaft sensor relative to the junk at that end of the engine - engine mount, bracket on engine, braces for them, alternator, etc. It is just above the camshaft pointing down toward it.

I'll hook up a mini video camera to look in the area.

Remember that is the end of the 3.3L stuff job that is tight against the body sheet metal, with the alternator taking up a lot of space at the back and various hoses routed closer to the engine that they should be.

 

[keithsketchley]

Thanks KOG

Do you recall how you got at the camshaft sensor?

From various illustrations in the service manual and my looking including with cheap borescope and feeling I think it is inboard of the aft brace that runs from from engine mount to block.

Figure 5 in the fuel injection section of the service manual omits the brace and some other obscuring parts. It shows what appears to be the aft end of the bolts holding the engine mount to the aft flange of the “engine bracket” (mislabeled as the mount).

Unclear if parts other than the brace have to be removed to get at it, such as the engine “engine bracket” and alternator bracket. I can see wiring and what probably is the connector pair in a groove inboard of the “engine bracket”, partly under the coolant thermostat housing which is cantilevered off something that projects outboard (from the aft head or intake manifold).

Elsewhere there is an illustration showing it is near the front corner of the aft head-block joint (well above the camshaft as the sprocket it reads is sizeable, points down so wiring pigtail comes from the top of the sensor).

Before pulling pieces off I want to have better knowledge. Hey, I do want to remove the manifold to replace the aft valve cover seal. ;-) Nooo, I do not want to remove mount and engine bracket unless really necessary.

 

[keithsketchley]

The camshaft sensor is hidden in the right end of the engine, inboard of the aft diagonal brace to the engine mount, aft of the thermostat housing but lower.​

By removing the brace you can see it and examine the connector pair including disconnect, spray contact cleaner, and reconnect. ​

(See http://www.allpar.com/forums/topic/146988-sensor-connections-locations-33l-v6-1994-grand-caravan/​

for details including information on removing the brace and other access tasks.)​

 

[keithsketchley]

I now have a vehicle that seems to run.​

I am not sure if it is the new MAP sensor or bending/wiggling wiring of MAP, TPS, cam, and crank sensors while I was testing them that made the difference. I’ll be wary of it for a few weeks.​

I’ll post a bit more of my understanding of the system’s operation later.​

 

[Bob Lincoln]

Fuel injectors all have a common +12V supply to the + pin, and the computer switches the - pin from +12V to ground for each individual injector to fire them in sequence. If you have any knowledge of electronics, the computer turns on the gate terminal of a MOSFET on the 'low' side of the injector in order to switch the - pin of the injector to ground. All injectors have +12V to their positive pins as long as the ASD relay is closed.