JonathanK
Mechanical
- Jun 19, 2007
- 15
Hi all,
Recently my car developed an issue. Recent as in a couple or few months but I finally have the time to sit down and look at it. I had issues with my hall effect crank position sensor sync'ing up according to the DTA ECU I run. It would take longer than normal to crank over. I hooked a scope up to the sensor and was receiving no output - replaced it with an identical sensor (Cherry GS100502).
However, it is still taking a long to sync up. I looked at the ECU and it will display "engine not turning" then I hold the starter on and it shows "engine turning, sync'ing, sync'd" then it will (with me still holding the starter) show 0 RPM/not turning for about 200 - 250ms, and then come back and start. The weird part is, when I scope the output of the sensor (ECU has an internal pull-up resistor 5o 5v of 1K ohm) I see the square wave pulses for the teeth, then the signal goes high (due to pull-up) like the ground drop out and it can't pull down! Then it comes back.
So, then I scoped this ground. The ECU has two pins (Pin 21 and 33) dedicated to "sensor grounds". I had the hall sensor grounding here but realized its likely not ideal given I provide the hall with 12v from car but ground to the sensor ground, which has sensors with 5v conditioned from the ECU. So, its possible the ground plane there is not right for the 12vdc coming from the vehicle. When I move the ground for the hall sensor from this sensor ground to the chassis directly in the engine bay - no noise on the hall signal what so ever. However, if I put the scope probe on the hall sensor output and chassis ground I get a clean square wave signal with no drops/noise/false triggers, car idles with 0 cam/crank errors for 16 minutes of running - but, if I hook the CH2 probe up to the sensor ground wire Pin 21/33 and chassis ground I see the ignition coil noise again.
So, to test whether it was radiated noise or through the ground/system I unplugged Pin21/33 and ran a wire out of the ECU completely separate of the wire bundle and into the engine bay hooking up to nothing - no noise on wire showing. I then made this new wire tie to the sensors as the sensor ground path and boom - noise is back. I then put the old wire back in the connector (Pin 21/33) and with it cut from the sensors in the engine bay, I scoped it - no noise. So it seems to not be picking it up in the wire bundle through the firewall but rather from the sensors themselves.
One thing worth noting is that even when the sensor ground wires were unhooked from the sensor ground pin of the ECU, they still report values - aka, they are getting a ground source from someplace else. This is likely what is feeding back into the ECU. A few of these sensors (Map sensor, oil pressure, fuel pressure) have a 5v, sensor ground, and output wire - the rest are 2-pin sensors that have a signal and sensor ground wire... but must be grounding through the body? But, if that's the case, that'd mean they are grounding to the 5vdc/sensor ground circuit AND the chassis ground - not ideal?
So - I really want to get this noise off this ground because it has the potential to mess with sensor readings and its keeping me from providing my hall sensors with 5vdc because its regulated and how the diagrams show it should be (sensor can run w/ 4.5 - 24vdc supply, so 12vdc works, but...) but the sensor ground path induces noise onto the hall sensor signals that causes sync issues. This whole issue didn't seem to present itself until recently (car has run for 1.5 - 2 years without having these issues).
Does this sound like a bad system ground (ECU) or bad engine ground? I am not very experienced with this sort of noise and thought its supposed to be dumped into the chassis ground... so I would think the 12vdc/chassis ground hall sensor setup would show the noise vs the sensor ground which is separate from the main ground connections.
Here is an image of the wiring diagram you can see the comment I added about 12vdc supply to hall sensors.
Thanks all for any advice!
Recently my car developed an issue. Recent as in a couple or few months but I finally have the time to sit down and look at it. I had issues with my hall effect crank position sensor sync'ing up according to the DTA ECU I run. It would take longer than normal to crank over. I hooked a scope up to the sensor and was receiving no output - replaced it with an identical sensor (Cherry GS100502).
However, it is still taking a long to sync up. I looked at the ECU and it will display "engine not turning" then I hold the starter on and it shows "engine turning, sync'ing, sync'd" then it will (with me still holding the starter) show 0 RPM/not turning for about 200 - 250ms, and then come back and start. The weird part is, when I scope the output of the sensor (ECU has an internal pull-up resistor 5o 5v of 1K ohm) I see the square wave pulses for the teeth, then the signal goes high (due to pull-up) like the ground drop out and it can't pull down! Then it comes back.
So, then I scoped this ground. The ECU has two pins (Pin 21 and 33) dedicated to "sensor grounds". I had the hall sensor grounding here but realized its likely not ideal given I provide the hall with 12v from car but ground to the sensor ground, which has sensors with 5v conditioned from the ECU. So, its possible the ground plane there is not right for the 12vdc coming from the vehicle. When I move the ground for the hall sensor from this sensor ground to the chassis directly in the engine bay - no noise on the hall signal what so ever. However, if I put the scope probe on the hall sensor output and chassis ground I get a clean square wave signal with no drops/noise/false triggers, car idles with 0 cam/crank errors for 16 minutes of running - but, if I hook the CH2 probe up to the sensor ground wire Pin 21/33 and chassis ground I see the ignition coil noise again.
So, to test whether it was radiated noise or through the ground/system I unplugged Pin21/33 and ran a wire out of the ECU completely separate of the wire bundle and into the engine bay hooking up to nothing - no noise on wire showing. I then made this new wire tie to the sensors as the sensor ground path and boom - noise is back. I then put the old wire back in the connector (Pin 21/33) and with it cut from the sensors in the engine bay, I scoped it - no noise. So it seems to not be picking it up in the wire bundle through the firewall but rather from the sensors themselves.
One thing worth noting is that even when the sensor ground wires were unhooked from the sensor ground pin of the ECU, they still report values - aka, they are getting a ground source from someplace else. This is likely what is feeding back into the ECU. A few of these sensors (Map sensor, oil pressure, fuel pressure) have a 5v, sensor ground, and output wire - the rest are 2-pin sensors that have a signal and sensor ground wire... but must be grounding through the body? But, if that's the case, that'd mean they are grounding to the 5vdc/sensor ground circuit AND the chassis ground - not ideal?
So - I really want to get this noise off this ground because it has the potential to mess with sensor readings and its keeping me from providing my hall sensors with 5vdc because its regulated and how the diagrams show it should be (sensor can run w/ 4.5 - 24vdc supply, so 12vdc works, but...) but the sensor ground path induces noise onto the hall sensor signals that causes sync issues. This whole issue didn't seem to present itself until recently (car has run for 1.5 - 2 years without having these issues).
Does this sound like a bad system ground (ECU) or bad engine ground? I am not very experienced with this sort of noise and thought its supposed to be dumped into the chassis ground... so I would think the 12vdc/chassis ground hall sensor setup would show the noise vs the sensor ground which is separate from the main ground connections.
Here is an image of the wiring diagram you can see the comment I added about 12vdc supply to hall sensors.
Thanks all for any advice!
