voltage hunching at certain RPM

[Shiggy]

Hi there,

I'm scratching my head to sort out the problem.
M800 voltage is hunching at certain RPM.

I've experienced on 3UZ-FE and SR20DE so far, not only one Motec ECU.
So it's not specific problem on certain engine or unit.

When certain RPM (mostly at just 3000RPM and 6000RPM), Batt V starts hunching.
(between 11V - 13.5V, not swinging fast, slowly like 30sec cycle)
It only happens at very narrow RPM range. 50-70RPM more or less, the issue will disappear.
I checked alternator and battery voltage by using voltage tester. voltage was stable like 14V.
However, ECU manager shows unstable voltage.
When I check voltage at A26 pin, the voltage actually changes as ECU manager shows.
When the voltage is changing in ECU, sensor readings (like MAP sensor) will also be moving.
(MAP from other datalogger showed stable pressure while this)

I thought it is due to high current of injector or ignition coil and A26 pin is too small for that current.
So I separated the voltage source of those from ECU power source. It didn't help sorting out.

Is it maybe due to certain frequency resonance in internal PCB?
Any idea to solve this? Please let me know.
This issue might not be happen on actual car, I believe it only happens on engine bench testing.
We are doing engine bench testing, and we need to make it possible to run exactly at 3000RPM and 6000RPM.

[RossB]

There must be a problem with the ground circuit on your installation. The MAP sensor is supplied by a regulated 5 volt supply in the ECU which should not be affected by variations in ECU supply voltage. If the MAP signal is changing as a function of ECU battery supply it is probably because there is a problem with the 0 volt supply. You should not be seeing a difference between ECU voltage and your actual battery voltage. I am not sure why it would only happen at a certain RPM but it may be that there is a loose connection which is affected by vibration.

[Shiggy]

Just checked all ground connection, MAP sensor is using original wires of M800 loom.
Ground wire is connected engine block, battery - and bench ground.
I made sure they are not loose too.

[SprinterTRD]

Your problem may be a blown diode in the alternator generating voltage spikes.

[Shiggy]

SprinterTRD, thanks for your advise;-)
I checked the voltage at alternator to ensure voltage doesn't spike,
then also tried running without alternator.
(supply voltage to battery by using battery charger while this test)

Unfortunately it didn't make difference...

[Mazi]

Battery charger could be a problem...

[SprinterTRD]

You cannot check a blown diode with a voltmeter you need to check with a scope on AC setting. You will also see the generated noise.

[Scott@FP]

You would need to disconnect ALL B+ sources from alternator, if one of the diodes is grounding internally it will have a unique scope pattern, and the period of the ground events can be very short, in sub-uS or less, and vibration frequency can be the trigger, the 3000 RPM and 6000 RPM points indicate a vibration trigger of the event.

Just removing the drive belt won't isolate a possible diode issue, must remove B+ from the alternator.