Intermitted Power Loss, Please Help!

mine is at 14.5% with AC turned off. And 21.1% with the AC compressor engaged.

Higher idle, or load on the engine (Radiator fan, lights, AC, etc) all adds up and will affect calc load.

cap1.JPG

cap2.JPG
 
Thanks for posting those. Maybe it is fine then. I guess normal calculated engine load % will vary for different vehicles.

I am going to wait for warmer weather, and then try getting to my garage. I am probably going to get a timing belt kit and do that too, as I doubt it has ever been done on the JDM motor.

I might not post again for a few weeks until warmer weather.


Anyone have a recommendation for a repair guide for the P5 or Protg? Like Haynes repair manual or one like it?
 
Hello All, I have an update.

I have finally had time to mess with my Mazda, and I may have solved the issue. I know I said this before, but I am like 69% sure that I have found the problem.

Over the past few weeks, I been unable to get my Mazda started. It would crank, start for a half second sometimes, and die. I wasn't sure if my timing belt broke, due to it being able to kind of start, but I decided to check it anyways. I took the valve cover off, only to find the timing belt looking fine to me.

But, when moving the wire harness that goes over the valve cover, I followed the connector and wire Crankshaft Position Sensor down to the sensor and notice an odd route that it traveled. It seem like it was extremely close to a pulley.

I had another sensor to put on the P5, so I decided just to do it sitting at my house, instead of trying to get the car to my garage which is 25 minutes away. I got the sensor off, and I found that the pulley was rubbing on the wire, and shaved about half way through at the deepest point.

I installed the new sensor, routed the wire differently, and put everything back together. I reset the ECU, and the car instantly started, and ran flawlessly.

This was not the first time that I reset the ECU and the car ran better, so I took it for a few test drives, and it ran great, not one moment of hesitation, stall, nor surging. I am not convinced that this is definitely the solution, and I probably won't be until next fall when the temperature drops again, but the P5 was running really well today.

I am attaching a picture of the wire, I am not sure why the wire was routed the way it was, it actually looks like it is supposed to be routed that way, but if there is slack in the wire, it will rub on the pulley.

IMG_3353.jpg
 
I am not sure how to mark the thread as "solved", but I have 500 miles on the P5 since finding the damaged wiring on the Crank Shaft Positioning Sensor, and the P5 has been running great since then.

I have driven it almost every day, and it hasn't given me once second of issues.

It also has been so long since driving it, I forgot how much fun it is to drive, even at its old age.

Anyways, that you all for your help, and there is a lot of useful information on this thread for people with similar issues.
 
The car is only as old as people let it get.................good subliminally hidden catch on that wire issue. I'm gonna check that out on mine even tho there is no hint of an issue.........and don't want there to be.
 
good subliminally hidden catch on that wire issue. I'm gonna check that out on mine even tho there is no hint of an issue.........and don't want there to be.

I might take a picture and post it to show the old and new routing of the Crankshaft Positioning Sensor cable.

I think the issue may have more to do with the clip on the connector of the sensor than the routing, meaning if the clip that anchors it onto the motor is broken (or just not attached), then this would create slack on the cable, enabling it to hit the pulley like in my situation.

I think if the sensor cable connector is attached to the motor, then there would be tension on the cable, pulling it away from the pulley.

Either way, I re-routed mine, I will take some pictures to better illustrate what I am referring too.
 
Thats cool you were able to find and fix it. Thats kind of what I was refering to. Mine was chaffed right at the eyelet to lift the motor. Same thing was happening though - it would touch metal short out and kill the engine. And without any code being thrown. Replacing it allowed me to route it securely away from any moving parts and add some extra tape around the area near the eyelet to prevent it from happening again.
 
The part that bugs me is that there was no code thrown.

You weren't shorting out your ignition to kill the car, you were shorting out a signal going to your ECU so the ECU killed the car.

The ECU obviously knew there was a problem with the crank sensor circuit so why couldn't it throw a P0335 code to give you a heads up ??

Just think how much time, money and frustration could have been saved if the car would have thrown that code along with killing the engine.

Maybe I'm expecting to much from an ECU running Windows 95 ??
 
The part that bugs me is that there was no code thrown.

I'm not sure either.

I was getting a Camshaft Positioning sensor CEL sometimes, but only when the issue was at its worse. I might have seen it 3 times total.

You weren't shorting out your ignition to kill the car, you were shorting out a signal going to your ECU so the ECU killed the car.

The ECU obviously knew there was a problem with the crank sensor circuit so why couldn't it throw a P0335 code to give you a heads up ??
I am wondering if the ECU assumed the data interpreted from the sensor was correct, and it screwed up the timing.

Maybe that would explain why the motor would drop RPM so quickly, maybe the ECU was putting fuel and spark in cylinders that were not ready for it yet.

Not sure if that could happen, but it would explain why there was no sensor code, and why it felt like something mechanical stopping my motor from moving, and the quick, dramatic, and sudden drop in RPMs.

Just think how much time, money and frustration could have been saved if the car would have thrown that code along with killing the engine.

Well, luckily I didn't spend too much money on the P5 itself, but it was very frustrating, and I did waste plenty of time dealing with the issue.

I guess you could say it cost me about $800 in extra fuel due to having to use my gas guzzling truck instead of the P5 for about 7-8 months. My truck gets 11.7 MPG, and my P5 averages 23.8.
 
Thats cool you were able to find and fix it. Thats kind of what I was refering to. Mine was chaffed right at the eyelet to lift the motor. Same thing was happening though - it would touch metal short out and kill the engine. And without any code being thrown. Replacing it allowed me to route it securely away from any moving parts and add some extra tape around the area near the eyelet to prevent it from happening again.
Thanks, I should have just replaced it a while back when you suggested it. It was one of the things I suspected, and I purchased the sensor months before actually installing it, now I wished I would have taken a closer look sooner.

I am curious, was the clip that holds the sensor connector onto the motor broken? I am thinking that maybe the extra slack was allowing the wire to rub on things it should of been.
 
I am wondering if the ECU assumed the data interpreted from the sensor was correct, and it screwed up the timing...

I'm 95% sure that the car will run with a bad cam sensor (no data) but it won't run with bad crank sensor data and the ECU kills the spark.

The ECU needs to know where the pistons are but can run not knowing exactly where the valves are.

I do remember one guy that couldn't get his car to run. He would get one spark then nothing every time he tried to start the car.
It turned out to be a dead crank sensor.
The ECU would allow one revolution of the engine waiting for the crank pulse and when it didn't show up it would kill the spark.


Perhaps if the car is going down the road when your crank sensor wire shorts out just a couple of pulses it can recover because your engine is still rotating until you come to a stop. But if your crank sensor stays shorted you get no spark and no power.


Either way I'm 95% the ECU was killing your spark and if so, it should have thrown a code along with it.


Perhaps upgrading your ECU to Windows 98 might get rid of that software glitch. (lol)
 
Maybe I don't fully understand how the ECU interprets and uses the data from the Crankshaft Position Sensor, but I was thinking that maybe the ECU was recieveing data from the sensor, but the data itself was bad.

If there was no data, then wouldn't the ECU kill the spark? What would happen if the data was just bad?

What would cause a sudden drop in RPMs? It would behave like the motor seized for a second, I guess that is the best way to explain it.


Anyway, the P5 has been running great, now I have to decide if I want to repair some front end suspension and steering issues.
 
Last edited:
The crank sensor picks up a pulse from every tooth on the crank pulley and sends it to the ECU.
The ECU looks for the missing pulse to know where the crankshaft is in it's rotation.



The pulse stream is quite simple in that it is always there or not at all. If the crank sensor is too far from the teeth, you basically have no data.

If the ECU gets no data for one complete rotation it kills the spark, but if it comes back before the car rolls to a stop, it may very well give the spark back.
It all depends on how long your wire stays grounded/shorted (no data/bad data).

So I'm thinking the sudden drop in RPM's was because your sensor wire shorted out and stayed that way so the ECU killed the spark.
I'm pretty sure the drop in RPM's would be exactly the same as turning your car off while you're driving along.
No ignition, no power.

I'd like to know what the parameters are for throwing a P0335 code. There was ample opportunity for the ECU to notice the bad data/no data/missing data and throw a code, especially if it steps in and kills the spark.
 
The sudden drop in RPM was not the same as turning the ignition off, it was like the motor was being forced to stop moving. It was not like engine braking. It was like dropping from 4k to 1k in a split second.

Besides, when a modern car is engine breaking, does the spark and injectors turn off? Or just injectors?
 
The sudden drop in RPM was not the same as turning the ignition off, it was like the motor was being forced to stop moving. It was not like engine braking. It was like dropping from 4k to 1k in a split second....

OK,.. So how about this ???

The ECU was waiting for it's next missing tooth pulse but then your sensor wire grounded,.. say... 40 BTDC and the ECU trusted the info from the crank sensor and then even tried to blame the cam sensor two or three times.
It could have ran like this for 4.2 seconds before the ECU cut the spark.
(I'm really trying to over think this... &I have been drinking)

 
Last edited:
Besides, when a modern car is engine breaking, does the spark and injectors turn off? Or just injectors?

I don't know if our car cuts injectors but it will cut spark.

I don't know if our car is considered a modern car ??
 
Last edited:
I don't know if our car cuts injectors but it will cut spark.
I think it does cut the injectors. IIRC, the O2 sensors read zero volts when engine breaking. Wouldn't that mean no fuel?

I know it does this for my 02 Ram, and my 2015 Mazda5, but I am not 100% positive it does this for the P5. I just can't remember.

I don't know if our car is considered a modern car ??
In the US, I would consider a modern car as one with OBDII.

It has been a while since I read it, but I remember a thread on a MR2 forum with someone asking if the injectors were cut off during engine breaking, which I can't remember the answer, but it was debated that since it is EFI, it should turn off, and other saying since it was OBDI, maybe it doesn't turn off. IIRC, other people said there are OBDII EFI engines that do not fully cut fuel during engine braking.

Not that all that clears anything up, but I guess it might be something for you to over think next time that you have been drinking. :)
 
I think it does cut the injectors...

Yes, it does, I remember now.
The hyper-miler guys on the forum some years back were leaving the car in gear while slowing down but not pushing in the clutch.
The car would slow down faster because the fuel was cut. As soon as the clutch was pushed the IAC would kick in and give the engine enough fuel to idle the car.

I tested this idea by coasting in fifth gear without pushing the clutch to see if it would roll to a stop and die (yea I was kinda bogging it). But once the RPMS dropped to around idle RPMS (700 or so) the injectors would kick back in and try to not let the RPMS drop any lower.
You could actually feel the injectors kick back in on smooth level pavement.

It sounds like your car was braking way harder than that, so I think you did have spark and fuel but your ignition timing was way off (due to bad crank info) and it was way before TDC.
 
Back