Mazda Protoge engine shuts off

I have a 2002 Mazda Protoge (I hate this little car - it drives like a go-cart) that my kids mostly drive. Lately it has developed a weird problem where the engine shuts off for a few seconds. It might only shut off for about a second, in which case the momentum of the engine starts it back up again, or it may shut off for 4 or 5 seconds, requiring an engine restart. The problem never lasts for more than a few seconds.

It’s fairly intermittent, but it’s getting more frequent. It started out happening once every several months or so, then about once a month. Now if I drive the car to work (about a 45 min drive one way) it might happen or it might not, but will usually happen on either the drive to or from work. In other words, it might make it one way (45 min) without a problem, but it usually won’t go all the way to work and back without the problem occurring at least once.

When it goes, all of the dash lights go out, the engine stalls, but the radio and blower motor still have power.

When it’s not happening, everything on the car works perfectly fine.

It’s a 4 cyl. automatic.

It’s thrown 3 codes so far:
P0705 Neutral safety switch circuit malfunction
P0117 Engine coolant temp sensor circuit low input
P0768 A/T shift solenoid ?D? circuit malfunction

My (possibly faulty) understanding of the neutral safety switch is that it shouldn’t cause the engine to shut off if it has a problem. There’s no coolant problem. The last code is possibly for the shift inhibit solenoid? (that’s a question). If so, the worst that should do is prevent the car from shifting out of park. It shouldn’t cause this either.

What the heck is actually broken on this thing?

correct. it only prevents you from starting the car if it’s not in “P” or “N.”

no. I know a little about the Protege; that solenoid is inside the transmission.

when you say the dash lights go out along with the engine shutoff, do you mean all of the backlighting, illumination, odometer, and warning lights? Do all of the gauges drop to 0 too? I’d WAG there’s a wiring problem somewhere. a circuit is faulting and cutting supply to the cluster and PCM. it’s possible that solenoid code is related, but w/o wiring diagrams and circuit tracing it’d be hard to tell.

edit: Or, there’s a circuit fault (short) in the communication bus somewhere. I remember on older Chrysler vehicles with the single-wire J1850 communication bus, it was fairly common for someone installing and aftermarket radio to inadvertently let the radio’s J1850 wire short to something, which would kill the cluster (often with a “n0 bu5” error in the odometer) and prevent the engine from starting.

I’ll try looking up tsb’s (technical service bulletins) on the car and glance at a wiring schematic, a few questions though first (in no specific order)-

when it acts up, is there anything you seemingly do to get it to come back around, or does it happen all by itself?

if it ever happens when you’re driving with your headlamps on, do they go out too, or just the engine and dash lights?

is the radio o.e.m. (or wired to function like the original)?

do you or anyone that drives the car have a lot of weight on your key ring?

what do your battery cables look like? (I’ll clean them regardless before starting to diagnose an electrical problem)

have you ever tried duplicating the problem by wiggling wiring harnesses, the ignition key, etc?

what engine, if there’s more than one option, does it have?

In the old days, the first notion might be the fuel filter or fuel system (pump?).

No idea how that may cause the error codes you describe to be initiated though… unless perhaps a dead motor situation while travelling down the road may initiate one or all of them.

dash lights don’t go out just because the fuel pump stops working.

it also wouldn’t trigger those error codes.

currently, i can’t say that it looks like those codes will have anything to do with the problem he’s describing

It would be helpful to know exactly what is meant here by “dash lights,” as there’s more than one possible interpretation. If it includes the gauges, my first thought is the ignition switch is worn and losing contact (mr horsepower alluded to this when asking about the key ring). If you can duplicate the symptom by maneuvering the ignition key, that’s almost certainly it.

Well, I can’t say they are similar vehicles, but FWIW when the Crank Sensor went on my Oldsmobile Cutlass Sierra, that’s what it acted like.

By “dash lights” I mean the check engine light, which is currently on due to the codes. Also, when it stalls, the battery light does not come on. A moment later, when it “recovers”, the battery light does come on. There’s an indicator for what gear the car is in, but the D for drive hasn’t worked in years, and of course that always happens to be the gear it’s in when the problem occurrs.

I think the speedometer and tach both drop to zero but I’m not 100 percent certain about that. I can only look at so many things when the problem happens.

It had a factory original radio in it, which hasn’t worked for years. The problem started long after the radio died. My son replaced the radio about a month ago, so the radio is no longer OEM, but that hasn’t affected the problem in any way, other than now I can tell that the radio stays on when the problem happens. The radio was installed with a plugin harness kit, so there was no wire cutting involved.

I initially thought that the entire electrical system was shutting off momentarily, but installing the radio proved that this was not what was happening. I was with my son when he was driving the car the other day and we intentionally cranked up the heater fan to high to see if the blower motor stayed running. It did.

There is nothing on the keychain. It’s just the remote key fob that came with the car. Jiggling the key has no effect.

I have not driven the car at night so I have no idea if the headlights stay on or not.

Battery terminals are good. Other than that, I haven’t jiggled a whole lot.

The engine is the 4 cyl. 2.0 L. I think that’s the only 4 cyl. engine available for that year.

The problem always (so far) fixes itself. There doesn’t seem to be any correlation to hitting potholes or anything else that might jiggle wiring.

I initially suspected that the error codes were really a function of the electrical system going screwy, but I’d like to hear what everyone else thinks about that.

Protege.

Carry on.

I’m an
[del]enginere[/del]
[del]engenere[/del]
[del]engeneer[/del]
I’m good at math.

Technically, it should have been Protegé anyway. Apparently Mazda did keep the diacritical.

More info.

We happened to need to go somewhere tonight, so I cleared all of the codes out of the Mazda and we drove it around to various places for about 3 hours total. After about 40 min. or so the problem occurred once, but it was very quick. The car was fine after that and never again showed the problem tonight. It did not throw any codes.

My son was driving. By the time I looked over at the dash, the speedometer and tachometer were both coming up from zero, so they definitely both went completely flat. The backlight for the dash was on when I looked. If it blinked off quickly during the incident I didn’t catch it, but I suppose it’s possible. The headlights dimmed a bit, but that could have been from the engine momentarily stalling. There was enough momentum in the engine that when the car recovered it started itself back up again, so no need to restart it with the key. We were on a flat and level road. No potholes, nothing that would have jiggled wiring.

I still see the ignition switch as the prime suspect. While sometimes jiggling can be part of it failing, that’s not always the case. It sounds like the ignition circuit is cutting out, and the switch is the most likely place for that to occur.

the format of my response is a little sloppy. I’m replying from my phone.
when i went to look up potential tsb’s and wiring diagrams,

-i was asked 1.6 or 2.0. the 2.0 looked to cover more models, you think it’s a 2.0, so i assumed it was.

-it then asked early or later production 2002, i didn’t see the month break point between early/late right there. it will somewhere. anyway, get the year & month of production off of the plate on the drivers door post (typically).

what’s the longest period of time the problem has occurred? (was your turning the blower on high preparing for when the problem next occurred or during ‘the problem’ happening?)

“battery terminals are good” - humor me. remove them from the posts and clean the posts and terminals anyway and make sure they’re tight (not lug nut tight, battery terminal tight) when you’re done… i can’t tell you how many cars, trucks, lawn tractors, big tractors, motorcycles, etc… that problems were fixed on despite the cables looking clean. look at any fusible link connections, verify ground wires are attached everywhere intended and that the connections are clean and tight, flex the cables & make sure they’re nice and flexible versus stiff from corrosion, etc… be especially mindful if the car has temporary terminal ends, aftermarket battery cables or has ever been in front end accident.
regarding the radio working. the first diagram i looked at showed it to be on a different circuit through the ignition switch than the engine controls. providing that schematic was the correct one, the radio working doesn’t tell us much on this particular car.

regarding the engine codes tying into this problem. educated guess, i kinda doubt they have anything to do with each other. especially if the codes weren’t triggered (after you’d cleared them) and experienced the problem again. (i haven’t looked at those codes and any potential shared circuit power sources or ground points though).

with the engine running, you need to wiggle any engine harness under the hood and under the dash you can get to without killing yourself. lightly at first. if you find an area that makes the bobble happen you can apply varying amounts of twisting/ moving the wiring to help pinpoint the problem at that point.

i agree with this. in particular the circuit that feeds a solid blue wire from the switch to a junction box at/ by the fuse panel.

I’d personally pull the column covers and look at the electrical part of the switch for anything that didn’t look right and wiggle the wires there.