Car runs and shifts great,always has.For about a month,the CEL will be on .Then,it will remain off for about a month.Then it will reappear for about a month.I've noticed this since mid-November,when I got the car.

I'm told that it is likely static causing this problem (?).I'm also toldto disconnect the battery,and hold the 2 battery cables together for 30 seconds,then hook battery back up,that many dealers do this.

I haven't tried this yet,being hesitant thinking there's a chance of shorting a capaciter or something in the ECM.Any validity to this information I'm getting?

 

Might be a good idea to find out what codes are being stored before you erase them. Usually when your CEL comes on it's trying to tell you something. Holding the hot and ground together as long as they are unhooked from the battery should not hurt anything. You can also try just unhooking them and push on the brake pedal, but this wont keep the CEL from coming on in the future.

 

I did pull the codes a couple of months ago.It was pointing to drive train management.It's been brought to my attention that could be a bad ground somewhere (?).Also been suggested could be the ECM,due to the age and years of varying extreme temperatures it experiences.Suggestions I've heard somewhat make sense.So I have just let it come on for a month,then go away for a month.Really weird.Car runs and shifts fine,always has.I guess I could go ahead and try this "snake-oil (?) " remedy for the light.In fact (hahaha),it's due to go off in about a week,maybe sooner.I welcome any more input.

 

I agree that knowing what the code number is would be a good first step. There are just so many different things that can cause the CEL to come it would be impossible to guess. Disconnecting the negative battery cable and then touching it to the positive post/cable is just another way of clearing trouble codes and reseting the fuel trims to zero. Touching negative to positive simply drains off residual voltage in transistors and the back up battery of the air bag system to ensure that all voltage is eliminated in the electrical system that could keep the code from being cleared.

 

So,like I thought,the touching of the cables is just a "band-aid".I know I should (and will) go buy my own scanner (what's a good one without breaking the bank?)and check the codes myself.It's just that it's very puzzleing why the light comes and goes at monthly intervals,with the car running fine all the while.

 

I just bought an Auto Xray Code Scout 700.I've never used a scanner before (I'm old school/carbs.and no OBD).MIL was giving me evap.,catalyst,misfire,O2 sensr.,and O2 htr.Gave me a P0400 (EGR flow malfuntion).I looked aroung the area of the EGR and found a small vacuum hose coming out of the back of the intake manifold to a small solenoid/switch mounted on the fire wall.Hooked it back up and cleared all the codes,thus the CEL light is off now.I'll be driving about 25 miles tonight and see if the light comes back on.

 

Would be willing to bet that you just found the source of the P0400 Am4. That vacuum hose probably goes to the EGR boost sensor solenoid valve and then another vac line off that solenoid should go to the boost sensor. This is the system that checks to see if the EGR is actually working andis not clogged up. If that vac line was off or EGR passages are clogged up and no intake vacuum can reach these two items then will set the P0400 EGR flow malfunction.
See there, you are not so old school that you could not figure out a malfunction that did not seem to have anything to do with the code stored. Most would have been yanking the EGR valve and replacing.
The P1777, I think it was before the edit, has to do with TCM communication with the PCM concerning torque reduction signal when the coolant temp is below 140'F. Should not be a source of CEL turning on but only HOLD light on/flashing.
I just bought a quick code scanner from MAC tools for $70 that is awesome for quick code pulling in the service drive. The IDS from Mazda works great but this little guy can pull a code # in under 10 seconds which is very nice. Can also clear codes, get some freeze frame data (although in metric[:@]) and look to see if drive cycles have been completed although have not figured that part out entirely. Directions for use suck but think I will get used to it and use it more often once I tinker around with it.