Can you help to explain this waveform of an idle motor?

[injb]

Update: I didn't ground the ground clip on the probe correctly.

Hi, I'm seeing something that confuses me and would really like to see if anyone can explain it. My knowledge of electronic is very basic. I recently bought a picoscope and I've used it to look at the signal on this device (see pics) which is the idle stabilizer valve from an 80's Bosch Motronic fuel injection system.

It has 3 pins - the middle one gets battery voltage, and then the computer grounds it via one of the other 2. If it grounds one side, the valve opens, allowing air to pass through. Ground the other side, it closes. I've also attached the circuit diagram of the driver and it looks straightforward enough.

Knowing all of the above, I expected to see a signal of 0 to battery voltage (around 14v with the engine running) with the probe tip on one of the 2 ground pins, and the probe's ground clip on chassis ground. I used 2 channels to watch both grounds at the same time.

But to my surprise, instead of 0 - 14v, what I got was -7 to +7 volts! Everything else was as I expected. Now, I don't think the cars electrical system has a negative supply. I double checked that I was using DC coupling on my scope, and I definitely was.

I'd like to understand why I see -7 to +7 instead of 0 to 14. I thought it might have to do with the inductance of the coil in the device, and/or back EMF when it starts turning. But that's only a vague idea for me - I don't really understand how it would happen.

Can anyone shed any light on it for me? Thanks!

[Mr. Scram]

Doesn't it simply depend on your reference or ground? Voltage is always relative, so if you declare 0 to between the two peaks you'll have 7V and -7V. If you declare the lowest point to be 0 volt you'll have 0V and 14V. The potential between the two points is the exact same, you just shift the convention of what you consider neutral.

[injb]

Doesn't it simply depend on your reference or ground? Voltage is always relative, so if you declare 0 to between the two peaks you'll have 7V and -7V. If you declare the lowest point to be 0 volt you'll have 0V and 14V. The potential between the two points is the exact same, you just shift the convention of what you consider neutral.

Well, yes, but since I set my reference to the chassis ground (shared by the battery and the computer) I thought that this would be the lowest value I would see. But somehow I'm seeing it go 7 volts below that. This is 7 volts below what the probe's ground clip is attached to.

[vk6zgo]

Doesn't it simply depend on your reference or ground? Voltage is always relative, so if you declare 0 to between the two peaks you'll have 7V and -7V. If you declare the lowest point to be 0 volt you'll have 0V and 14V. The potential between the two points is the exact same, you just shift the convention of what you consider neutral.

Well, yes, but since I set my reference to the chassis ground (shared by the battery and the computer) I thought that this would be the lowest value I would see. But somehow I'm seeing it go 7 volts below that. This is 7 volts below what the probe's ground clip is attached to.

What does your Oscilloscope show if you put tne probe on the positive battery terminal?

[injb]

What does your Oscilloscope show if you put tne probe on the positive battery terminal?

It shows approx 12v (the engine was off when I did that).

[cs.dk]

What engine/ECU is it? Maybe i have some data on it.

[KrudyZ]

with the probe tip on one of the 2 ground pins, and the probe's ground clip on chassis ground.

If you see a signal between circuit ground and chassis ground then the two are most likely isolated from each other and the signal you see is AC coupled because of that isolation.

[injb]

What engine/ECU is it? Maybe i have some data on it.

It's a Porsche 944 Turbo, Motronic 3 point something...not sure exactly!

with the probe tip on one of the 2 ground pins, and the probe's ground clip on chassis ground.

If you see a signal between circuit ground and chassis ground then the two are most likely isolated from each other and the signal you see is AC coupled because of that isolation.

I don't think it's supposed to be like that, given the ground points shown in the diagram.

I'm beginning to think I was just wrong about the "ground" point I used being grounded. I thought I checked, but I'll go back and check again. When I checked the battery voltage I was using a different physical ground point.

[EPTech]

Hi there

I am no car mechanic, but this looks like a proportional valve setup. The width of the pulse determines how much the valve opens. The valve itself gets pulled to GND on one side during the on-time of the pulse and pulled to GND during the off-time of the pulse on the other side. In that case the signals look OK. Concerning the amplitude, it may be a case of bad ground reference or bad supply to the coil of the valve.

As an extra measurement you could measure whether there is a short of low resistance path between either two connections coming from the ECU and GND and 12V.

Happy hunting.

[injb]

OK...it was my mistake (who saw that coming? )

I just checked the "ground" point I was using again, and it's not grounded. I was sure I had checked this before. It was the metal box of the ECU itself I was using. When I captured the waveforms in my post above, I saw weird nonsensical waveforms until I attached my ground clip to the metal, then it suddenly looked like what I expected (except for the missing 0v reference). So I just carried on assuming I was using a valid ground point. I checked it again just now with a multimeter and it's floating, so I suppose that explains what I was seeing. Problem solved. Illuminati not confirmed.

I'm curious to know why attaching my ground clip to a big metal box that's floating has a similar effect to grounding (in that it turned my signal from gibberish into the right waveform). But next time I'll use a ground pin in the wiring harness anyway.

Hi there

I am no car mechanic, but this looks like a proportional valve setup. The width of the pulse determines how much the valve opens. The valve itself gets pulled to GND on one side during the on-time of the pulse and pulled to GND during the off-time of the pulse on the other side. In that case the signals look OK. Concerning the amplitude, it may be a case of bad ground reference or bad supply to the coil of the valve.

As an extra measurement you could measure whether there is a short of low resistance path between either two connections coming from the ECU and GND and 12V.

Happy hunting.

Yep you're exactly right about the valve works, and the bad ground reference. Thanks!

Thanks to everyone for helping anyway!

[KrudyZ]

I'm curious to know why attaching my ground clip to a big metal box that's floating has a similar effect to grounding (in that it turned my signal from gibberish into the right waveform). But next time I'll use a ground pin in the wiring harness anyway.

The coil has a certain amount of capacitance to the housing and your scope probe has a high input impedance.
What is happening is that the probe / scope is charging the capacitance to the average value of your AC signal.
For a 50% duty cycle signal that would be half the signal swing and what you are seeing on the scope is relative to that voltage and results in positive and negative excursions on the scope.