I have been helping my friend trouble-shoot a "crank but no start" problem on his 1999 stock Miata. We carpool so I driving until it's fixed ;-) It was running fine but after a week of setting it had a rough start and died.


The steps that followed:


1. Checked timing belt- Looked good, has not slipped a tooth and all the marks were in-line. Turned the motor by hand and it all looked good.


2. Checked fuel pump and filter. - Fuel pump running (jumpered @ diag port), replaced fuel filter for good measure. We charged the battery and checked the resting voltage. Changed the spark plugs, as they were old.


3. Placed an inductive timing light on the plug wires, spark present - It tricked us, intermittent spark, more on this later.


4. Tried starting fluid - Still a no-go


5. Checked for spark again with the timing light - Found intermittent spark, car would try to start when we got a flash on the timing light.


5 1/2. Checked resistance of the two primary windings of ignition coils - Looked good within manual limits. Coils had been changed many years ago. Was told they were prone to failure.


6. Found an OBDII error code reader - Two errors, p1345 and p0339


7. Replaced crankshaft sensor, gap set to .030" - No luck


8. Replaced Cam sensor - No luck, car throws a p1345 and p0339 every time a start is attempted.


At this point I am open to suggestions.


ECU?
Bad coils that ohm out good?
Broken wire to one of the sensors?


Any input would be appreciated. I have done a lot of internet searching but can't peg this one.

 

Found a good thread for you: HELP!!! P0300, P0138, P0339, P1345 - MX-5 Miata Forum

Here's bottom line: After all this time I knew there was something wrong other than bad sensors or the wiring to them as they had been checked and/or replaced/repaired. So I needed to look elsewhere. From the beginning I said to myself that the motor sounds badly out of timing as it sputtered trying to start so tonight I ventured to discover if the timing belt had skipped. I took off the valve cover so I could see the cam sprockets and then spun the crankshaft until the timing marks lined up with TDC. The cams were not close to the right position-(AHAH), they did however line up when the timing mark on the main pulley was about 45deg CCW to the TDC mark (double AHAH). I knew right away what I was going to find as I dismantled the front of the engine further and I proved myself right: a loose crank bolt and a broken woodruff key. The key sheared at the front of the timing belt pulley so thankfully the crankshaft is not toast. The valve timing remained in sync with the crank, just the outer pulley with the timing wheel shifted which through off the ignition so badly that the motor couldn't run.
I should have discovered this earlier as the engine had lost power from idle to 3500rpm. A timing check should have been done but I didn't because it's not adjustable on a '99 (without mods).
corrosion was a big factor, the bolt had rust covering all of the threads. I will be using the Loctite Fix as described in the Garage as the keyway is shot on the outer portion where the pulley boss sits.
Moral of the story: P1345 and P0339 together means that your timing is way off or both cam and crank sensors failed simultaneously.

 

Thank you UseYourNoggin,

I will verify TDC and give an update. We did look at #1 when we had the plugs out but it could be off.


The timing belt was changed along with the water pump about one year ago.


Don't think we will get back to it until this weekend. It's too dark to work on the weekdays.

 

Addition trouble-shooting


1. Took a look at the cam as the engine was being turned over. Its movement appeared normal. Later confirmed the #1 TDC was in sync with the timing belt marks. It doesn't look like it a broken key or keyway.


2. Read the thread UseYourNoggin suggested to the end and thought it might be possible we installed the NB crank wheel "dish side out" while inspecting the timing belt.


3. That turn out to be true, so we flipped it around. The car started and ran briefly on #3 and #4 cylinders. Or at least that is what we think after inspecting the plugs. Thought the engine might have been flooded so we dried out the plugs and tried again. No luck. I find it strange that when the motor is a TDC none of the four pips are anywhere near the crankshaft sensor. Makes me think crank wheel is still not install properly.


At no point have we ever been able to get consistent spark to the spark plugs.


Going to try:
1. Check voltage and ground to the crank and cam sensors.
2. Check signals coming out of the two sensors.
3. Confirm the crank wheel is install properly.
4. Try to figure out when the ECU decides its ok to fire a spark plug and what is interrupting the sequence.


The trouble-shooting continues!!

 

Have you removed all the pieces of the timing belt cover?

I suggest removing all the covers, wind the engine over by hand and make sure both the cam gears are aligned and the crank gear aligned all at the same time.

 

Yes, we did! It all looked good. Turned the engine over by hand a few times and everything lines up. Also verified #1 TDC when the marks are in line.

I'm sure we are missing something basic, just can't find it. Going to dig into the electronics next. Don't think that's the problem but going to try it anyways.

 

Ohm out the CKP wires from the CKP/CMP connector's to the PCM. These sensors produce AC voltage and will not have 12 volts applied. The wires nearest the CKP/CMP connectors get heated regularly and can cause the wires to degrade yet still have good coating. A wiggle test might find that problem. An ohm meter is probably all it will take to find the problem. The CKP shares a ground with the MAF so try disconnecting the MAF and cranking it. There is a thermo sensor right next to the coil that can cause a no start but generally sets a code. The ignition could very well be bad, it turns 12 volts into 40k volts so it works hot and hard, and is in a very hot spot wedged between the back of the cylinder head and firewall.

 

Rained out last weekend. Going to give it a try Saturday.

 

Summary: Problem caused by a bad crankshaft sensor.


This is easy to check, well kind of....


Monitored the #1 injector and the #1 plug wires to see if one or both fire. If neither trigger at all or intermittently it's most likely the crankshaft sensor OR the crankshaft timing wheel is installed backwards.


Debug was accomplished with with an inductive timing light and a probe inserted into the backside of the ECU connector to monitor the injector. A LED with a series current limiting resistor was inserted between the probe point and +12 VDC. When the ECU grounds the injector to fire it the LED will flash. This was the only place we could get at the injectors (on a '99) without tearing the engine apart.


Turns out both the injector and the spark plug were triggering intermittently. At this point we returned to the crankshaft sensor and found the old sensor was bad and the new sensor was working. This left only the crankshaft timing wheel.


Drum roll, we put it in backwards (head lowered and under my breath) two times. So, after we fixed the original problem (bad sensor) and the one we created (backwards timing wheel) it started and ran just fine.