Difficult starting 96 Mazda 626 2.0

Tiny
TOHEARYOU
  • 1996 MAZDA 626
  • 4 CYL
  • 2WD
  • AUTOMATIC

Do it yourselfer. Full rebuild 96 626 2.0. Also changed temperature sensor at rear of head which removed idle roughness. Engine runs smooth as silk. 30 miles on engine. When in 20 degree weather the system acts as if no fuel is being sent to injectors. Checked pressure at fuel filter, ok. 3 to 4 hours of cool down is when this happens. It takes 20 to 30 bumps to start the engine. Once it starts never fails to start again. However smooth idles about 6 to 700 rpm. Almost to low. After no hesitation, no fuel shortage at sustained high rate of speed with passing acceleration. Baffled?

Do you
have the same problem?
Yes
No
Sunday, January 9th, 2011 AT 2:20 AM

22 Replies

Tiny
KHLOW2008
  • EXPERT
  • 42,032 POSTS

Have you tried to retrieve for trouble codes?
Did you check the injectors for pulsing when it could not be started? Ever tried to test if starting fluid works?

Symptoms indicates a possible fault with the ECM or its grounding cirtuit.

Was this
answer
helpful?
Yes
No
Sunday, January 9th, 2011 AT 9:54 AM
Tiny
TOHEARYOU
  • MEMBER

Took it to one of the box stores to retrieve codes. My 96 honda had the plug under the ash tray. On this car we could not find the hookup in the cabin, but I did find the diagnostics port under the hood on the left fender well. Box store had no jack for it. Checking this on monday. I did not check the injectors for pulsing. I thought about the injectors but questioned woul all 4 malfunction? I will try the starting fluid today and see what happens.

Was this
answer
helpful?
Yes
No
Sunday, January 9th, 2011 AT 12:47 PM
Tiny
KHLOW2008
  • EXPERT
  • 42,032 POSTS

When the ECM fails, all 4 injectors would not work.

The DLC is located underdash next to center console, above right foot of driver and if it is not found, it is a pre-ODB-II system, meaning rhe vehcile could be earlier than 1996 model. Code retrieval can be done manually with the aid of an analog voltmeter, LED lamp or the MIL.

CHECK ENGINE Light/Malfunction Indicator Light (MIL)

Turn ignition on. Connect a jumper wire between Self-Test Input (STI) pigtail and signal return (SIG RTN) terminal of Data Link Connector (DLC). See Fig. 2 -3. Go to KOEO SELF-TEST.

KOEO SELF-TEST
Ensure engine is warmed to normal operating temperature. If engine does not start (or stalls after starting), continue KOEO SELF-TEST.
DO NOT depress throttle. Turn ignition off. Ensure test equipment is properly attached. Turn ignition on (engine off). Record all KOEO and Continuous Memory Codes.

If a Code 111 (pass code) is not retrieved, service KOEO codes. See
DIAGNOSTIC TROUBLE CODE REFERENCE. If PCM will not output codes, go to TEST QA. If diagnostic trouble codes are retrieved:

If MIL/CHECK ENGINE light is on, service diagnostic trouble codes in order retrieved. See the DIAGNOSTIC TROUBLE CODE REFERENCE table.

If vehicle has a no-start condition, go to TEST AA.

If vehicle displays a Code 111 (pass code) and does not have any symptoms described in previous steps, go to KOER SELF-TEST.

Was this
answer
helpful?
Yes
No
Sunday, January 9th, 2011 AT 8:59 PM
Tiny
KHLOW2008
  • EXPERT
  • 42,032 POSTS

Sorry for missing out on the diagrams.

Was this
answer
helpful?
Yes
No
Sunday, January 9th, 2011 AT 9:00 PM
Tiny
TOHEARYOU
  • MEMBER

Thank you for a starting point. Monday & Tuesday I will begin these tests. Charged the battery on Sunday. I will send my result.

Was this
answer
helpful?
Yes
No
Monday, January 10th, 2011 AT 12:25 AM
Tiny
TOHEARYOU
  • MEMBER

Monday 10th bumbed the engine 38 times before the car primed and started. Computer codes were (sensor 1) (p o 117). Researching the codes after typing this into the forum.

Was this
answer
helpful?
Yes
No
Monday, January 10th, 2011 AT 8:21 PM
Tiny
TOHEARYOU
  • MEMBER

P o 117 is a intermitting sensor for the temperature guage. This is a seperate problem. Thinking the system is some how loosing prime, but when I remove the hose from the filter the fuel comes up strong, which makes me wonder about the regulator.

Was this
answer
helpful?
Yes
No
Monday, January 10th, 2011 AT 8:56 PM
Tiny
KHLOW2008
  • EXPERT
  • 42,032 POSTS

A bad ECT can cause a non starting condition as the ECM could be sending the wrong signal to the injectors and cause the engine to flood.

Did you test if it is sparks or fuel that is not getting to the cylinders or too much fuel is flooding the cylinders?

Was this
answer
helpful?
Yes
No
Tuesday, January 11th, 2011 AT 2:26 PM
Tiny
TOHEARYOU
  • MEMBER

I will check both your suggestions. It snowed and iced yesterday and today. I have to work outside in a parking lot. Thursday or Friday before I can get back to it. I will first spray some starting fluid at the air intake and see if any ignition spark shows.

Was this
answer
helpful?
Yes
No
Tuesday, January 11th, 2011 AT 11:00 PM
Tiny
TOHEARYOU
  • MEMBER

As stated earlier no code indicators. Bumped the car 1/2 bumped the car to confirm the same pattern. It was the same. Opened the air filter box sprayed starting fulid into the throat and hit the ignition the car started. Primed up and ran. However it is ideling down to 600-700 rpm while cold. Similar to the the way when before changing the temperature sensor?

Was this
answer
helpful?
Yes
No
Wednesday, January 12th, 2011 AT 5:24 PM
Tiny
KHLOW2008
  • EXPERT
  • 42,032 POSTS

This is and indication of insufficient fuel as starting fluid worked.

Have the fuel pressure checked.
Try priming the fuel pump by turning the ignition ON for 2 seconds and OFF. Repeat 2 or 3 times before trying to start engine. Does it improve?

The MAF could also be related to the problem.

Was this
answer
helpful?
Yes
No
Thursday, January 13th, 2011 AT 12:01 AM
Tiny
TOHEARYOU
  • MEMBER

I was told that this 626 has a crank sensor that picks up the crank turning and when it does not see the crank turning it turns off the fuel to the injectors. I did not see this ensor when I did the job. Idid this rebuild above and below while still mounted. The block addition above the pan exposes te flywheel when doing bearings and rings. Possibly in their. Have not researched this. Your thoughts?

Was this
answer
helpful?
Yes
No
Thursday, January 13th, 2011 AT 1:11 AM
Tiny
TOHEARYOU
  • MEMBER

MAF usually gives a code.

Was this
answer
helpful?
Yes
No
Thursday, January 13th, 2011 AT 1:13 AM
Tiny
KHLOW2008
  • EXPERT
  • 42,032 POSTS

The MAF fuel pump circuit might not produce any codes.

Crankshaft sensor is located at front of engine.

CRANKSHAFT POSITION SENSOR
1. Unplug Crankshaft Position (CKP) sensor connector. Measure resistance between CKP sensor terminals "A" and "B". See Fig. 2. Resistance should be 520-580 ohms at 68 F (20 C). Replace sensor if not within specification. When installing sensor, adjust air gap to.020-.060" (.5-1.5 mm).

Was this
answer
helpful?
Yes
No
Thursday, January 13th, 2011 AT 12:50 PM
Tiny
TOHEARYOU
  • MEMBER

I have checked the crank shaft position sensor and it tested fine. Took it off and checked the harness then remounted it at 26k. It was clearly out of position and now is properly secured. Started the vehicle then unplugged it at the connector, the engine immediately died. I am presently waiting for the car to cool down as to see if it starts any differently. I still believe I have a timing issue. Scoring the distributor as well as the head for the original timing position I have moved above and below this mark and the car starts easier when 1/4 to 3/8 counter direction over the clock wise direction. Possible computer timing difference? When running fuel seems to be in no short supply

Was this
answer
helpful?
Yes
No
Saturday, January 15th, 2011 AT 9:44 PM
Tiny
KHLOW2008
  • EXPERT
  • 42,032 POSTS

Get a timing light to check the ignition timing to confirm if it is within specs.

If the cylinder head and/or block surface were skimmed during the rebuild, you could have a timing difference problem, as the camshafts timing would be slighly retarded.

Was this
answer
helpful?
Yes
No
Sunday, January 16th, 2011 AT 3:25 AM
Tiny
TOHEARYOU
  • MEMBER

Hello everyone. I am back with an update on this 626. I looked at each and every one of the above suggestions. After I exhausted all sensors, timing issues, and fuel issues I decided to schedule with the best of mazda tech's in my area after speaking with numerous professionals. I waited anxiously for 2.5 weeks for my turn in line only to hear his words I have no idea why this car is not running correctly. Your indiviual specs are set properly. Anything else I would be throwing stuff at the car to see if it makes a difference. You possibly have a ECU problem.
Now I have found the ecu model (EEC-B(FSB9-18881-E) placed it in the car from a PA. Junk yard only to find out the unit sent was defective. I have been looking for another to replace the original, but I can only find the (EEC-B(FSB9-18881) with the F extension letter. Again this is a 96 262 2.0 automatic. Does anyone know if the (F) verssion will work in this car.

Was this
answer
helpful?
Yes
No
Monday, February 7th, 2011 AT 12:50 AM
Tiny
KHLOW2008
  • EXPERT
  • 42,032 POSTS

The Part No. Indicated in our database is FSB-18881-R0D for your model.

The F could be a later production and I believe it should work.

Did you remove the ECU cover to check for leaking capacitors on the circuit board? Some are repairable by replacing the capacitors.

Was this
answer
helpful?
Yes
No
Monday, February 7th, 2011 AT 4:18 AM
Tiny
TOHEARYOU
  • MEMBER

No I did not check the capacitors. The ecu works fine except in the cold start. It takes significant time to start 20-30 times. But once the temperature is up the car runs fine. I am suspicious of the ecu but if the capacitors were breaking down would the car run as well as it does, or is it a circuit board condensation/circuit issue temp. Related? Also I have looked all over the net and ave not seen the fsb-18881-rod extension, thank you I will look. This car has been the car that keeps on giving.

Was this
answer
helpful?
Yes
No
Monday, February 7th, 2011 AT 11:40 AM
Tiny
KHLOW2008
  • EXPERT
  • 42,032 POSTS

When capacitors have problem, it can cause the symptoms described. I have not encountered it on a Mazda but have experienced such problems on various occasions with Honda ECU.

Was this
answer
helpful?
Yes
No
Monday, February 7th, 2011 AT 11:51 AM

Please login or register to post a reply.

Recommended Guides