carmudgeons.com - View Single Post - What the Mazda RX-8 should have been (new autocross car build)

John V · 06-28-2019, 02:45 PM

I haven't posted on this in a while. I've been struggling with random misfires and have been systematically tracing everything out to find it. The logs were showing crankshaft sensor errors (meaning the pattern was not what the computer expected). I was really meticulous when doing the wiring so I didn't think that was it. I checked coils, grounds, supplies, everything with no luck. I ended up buying a $65 USB oscilloscope (I really needed one anyway) to scope the cam and crank signals and found the issue. Below is a good crank (yellow) and cam (green) signal out of the sensors, with the engine idling. The crank trigger wheel consists of 16 teeth followed by a two-tooth gap, followed by one tooth, another two tooth gap, 13 teeth, then a two tooth gap. You can see this in the trace. Sixteen pulses, a gap, a pulse, a gap, etc.

That's how it should look. Now here is the same signal with the engine at 5,000 RPM. Obviously, this is completely fucked. The hall effect sensor is mistriggering all over the place. I don't like just replacing parts, but it seemed likely that was the culprit.

And... today was a good day. I got the new crank sensor from Mazdacomp, put that in, and holy crap, my crank triggers look pristine. RPM traces in the logs look a thousand percent smoother. The ECU isn't logging a single crank error in several minutes of revving, so I think this one is licked.

I looked back on my purchases and it looks like the original crank sensor I bought was an aftermarket one. So, lesson learned, buy OEM parts. I did order two sensors in case this one shits the bed, but I think it was just a case of bad Chineseum parts.

Also, I finished my coil brackets and mounted the Nissan R35 GT-R coils. Probably didn't need them, but it seemed like a good idea. We did see some spark breakup on the dyno which limited us to 405 wheel horsepower at 6,200 RPM. I closed the plug gap which helped, but it still broke up. Now, this might have been an early sign the crank sensor was going bad, but the GT-R coils were something I had and it seems like good insurance. The car should easily make 450 at the wheels and rev to 7,500 RPM all day. I cannot wait to get back behind the wheel. The car was good before, but it's going to be unreal now.

06-28-2019, 02:45 PM	#270
John V No more BMWs Join Date: Apr 2005 Carmudgeonly Ride: Ram, MS3, CX-5, RX-8 Location: Glenwood, MD Posts: 14,753	I haven't posted on this in a while. I've been struggling with random misfires and have been systematically tracing everything out to find it. The logs were showing crankshaft sensor errors (meaning the pattern was not what the computer expected). I was really meticulous when doing the wiring so I didn't think that was it. I checked coils, grounds, supplies, everything with no luck. I ended up buying a $65 USB oscilloscope (I really needed one anyway) to scope the cam and crank signals and found the issue. Below is a good crank (yellow) and cam (green) signal out of the sensors, with the engine idling. The crank trigger wheel consists of 16 teeth followed by a two-tooth gap, followed by one tooth, another two tooth gap, 13 teeth, then a two tooth gap. You can see this in the trace. Sixteen pulses, a gap, a pulse, a gap, etc. That's how it should look. Now here is the same signal with the engine at 5,000 RPM. Obviously, this is completely fucked. The hall effect sensor is mistriggering all over the place. I don't like just replacing parts, but it seemed likely that was the culprit. And... today was a good day. I got the new crank sensor from Mazdacomp, put that in, and holy crap, my crank triggers look pristine. RPM traces in the logs look a thousand percent smoother. The ECU isn't logging a single crank error in several minutes of revving, so I think this one is licked. I looked back on my purchases and it looks like the original crank sensor I bought was an aftermarket one. So, lesson learned, buy OEM parts. I did order two sensors in case this one shits the bed, but I think it was just a case of bad Chineseum parts. Also, I finished my coil brackets and mounted the Nissan R35 GT-R coils. Probably didn't need them, but it seemed like a good idea. We did see some spark breakup on the dyno which limited us to 405 wheel horsepower at 6,200 RPM. I closed the plug gap which helped, but it still broke up. Now, this might have been an early sign the crank sensor was going bad, but the GT-R coils were something I had and it seems like good insurance. The car should easily make 450 at the wheels and rev to 7,500 RPM all day. I cannot wait to get back behind the wheel. The car was good before, but it's going to be unreal now.