Expected quite some noise, but found nearly none.

[eTobey]

Hi,

i have a coil, that i send current through with a half bridge. After the coil i have a shunt resistor, and there i measure the current (green trace in ampere), that swings (seemingly) about 2.6A pk-pk with quite a steep slope. It is the second ringing on the screenshots, that i refer to.

Blue trace:    Gate of HS
Red trace:     Voltage of shunt (0.104Ohm)
Yellow trace:  Measured noise with a 15mm shorted loop.
Green trace:  Ampereof coil

Full bandwith with peakdetect.

The loop was help right on the cable, but there was only 20mVpk noise. I would have expected about 10 times more. Other chips on the PCB generate higher peak noise.

Do i measure a signal that is not there with that amplidtude, or do i have to measure this noise differently?

[temperance]

The induced voltage depends on the rate of change of current.

[eTobey]

The induced voltage depends on the rate of change of current.

227A/us would not be quite a rate of change?

[jonpaul]

Ringing due to poor probe technique and bad breadboard setup.


Read about power supply measurements and probe measurements first.

j

[mtwieg]

Your descriptions of the scope traces don't make much sense:

Blue trace:    Gate of HS

I'm guessing we see two transitions because of dead time in the half bridge? Is the y scaling right, drive voltage is only around 5V?

Red trace:     Voltage of shunt (0.104Ohm)

So it looks like the current shoots up to >30A (3V/0.1ohm), at which point the trace is off the screen... is that right?

Yellow trace:  Measured noise with a 15mm shorted loop.

.... what?

Green trace:  Ampereof coil

So you're measuring current with something else besides the shunt resistor (red trace)? What's the sensitivity of this sensor?

[temperance]

Ringing due to poor probe technique and bad breadboard setup.


Read about power supply measurements and probe measurements first.

j

That's not what the question is about? The question is about induced voltage in a loop created by shorting the probe to the tip.

[eTobey]

Your descriptions of the scope traces don't make much sense:

Blue trace:    Gate of HS

I'm guessing we see two transitions because of dead time in the half bridge? Is the y scaling right, drive voltage is only around 5V?

Red trace:     Voltage of shunt (0.104Ohm)

So it looks like the current shoots up to >30A (3V/0.1ohm), at which point the trace is off the screen... is that right?

Yellow trace:  Measured noise with a 15mm shorted loop.

.... what?

Green trace:  Ampereof coil

So you're measuring current with something else besides the shunt resistor (red trace)? What's the sensitivity of this sensor?

It is the gate on turn on. See the miller plateau. It goes up to 10V after that.

The red trace is a bit missleading, as i have F3 activated (which is of no interest). C2 ( the other red) is not leaving the screen.

Yellow trace is measured with a DIY field probe.

No, there is no other sensor, it is a math function.

[eTobey]

Ringing due to poor probe technique and bad breadboard setup.


Read about power supply measurements and probe measurements first.

j

What about you reading my post and looking at my pictures first, before you give an answer? Or dont do drugs or alcohol, since you see things, that  are just not there! There is no breadboard, and the probing technique i would consider not poor.

It is of no use whatsoever, to give answers just from ones guessings, when there are informations, so that one would not need to guess!

[temperance]

Ringing due to poor probe technique and bad breadboard setup.


Read about power supply measurements and probe measurements first.

j

What about you reading my post and looking at my pictures first, before you give an answer? Or dont do drugs or alcohol, since you see things, that  are just not there! There is no breadboard, and the probing technique i would consider not poor.

It is of no use whatsoever, to give answers just from ones guessings, when there are informations, so that one would not need to guess!

I think you have to admit that what you wrote isn't very clear and the picture doesn't show much either. It shows a probe attached to a power resistor and some wires. Where is the DIY field probe? Because that's what you're question is all about.

[mikerj]

Ringing due to poor probe technique and bad breadboard setup.


Read about power supply measurements and probe measurements first.

j

What about you reading my post and looking at my pictures first, before you give an answer? Or dont do drugs or alcohol, since you see things, that  are just not there! There is no breadboard, and the probing technique i would consider not poor.

It is of no use whatsoever, to give answers just from ones guessings, when there are informations, so that one would not need to guess!

Calm down, this kind of attitude isn't going to want people to help you.  Why would you think that a wire wound power resistor would be appropriate for measuring rapidly changing currents?

[mtwieg]

It is the gate on turn on. See the miller plateau. It goes up to 10V after that.

I highly doubt that whole thing is one miller plateau, unless you intentionally added a ton of Cgd for some reason. Can you show what Vds looks like at the same time?

The red trace is a bit missleading, as i have F3 activated (which is of no interest). C2 ( the other red) is not leaving the screen.

By C2 you're referring to the pink-ish trace in the third capture? Is that the voltage across the shunt resistor?

Yellow trace is measured with a DIY field probe.

I understand using sniffer coils/pickup loops for detecting HF ringing, but it's impossible to make heads or tails of its output without knowing more about it (like what is its self resonant frequency) and exactly how it's positioned.

No, there is no other sensor, it is a math function.

What is a math function? Of what?

the probing technique i would consider not poor

It's hard to say for sure without seeing the larger setup.

[eTobey]

I think you have to admit that what you wrote isn't very clear and the picture doesn't show much either. It shows a probe attached to a power resistor and some wires. Where is the DIY field probe? Because that's what you're question is all about.

Its a wireloop on the tip of a probe, square about 15-20mm. Not worth a picture IMHO.

Why would you think that a wire wound power resistor would be appropriate for measuring rapidly changing currents?

Because it does conduct current, and it creates a voltage drop without going up in flames. Why would it not be appropriate?

....

That Millerplateau looks distorted, since the gate current is controlled by the driver (6EDL7141).


Yes its the pinkish trace. Its a mess with these colors...
Yes its the voltage accross the shunt.


Would the selfresonant frequency be so much of a factor, when there are only a few swings, or even on edge?
The wireloop was positioned on the cable in the best way (tested on multiple places) to pick up the most noise/signal.


Its the math function of the scope.

[eTobey]

I conducted another test, as the method of current measurement was questioned:

What i found is, that the magnetic field, that is produced by the changing current, induces a bigger voltage on the signal, than the actual voltage that is produced by the current in the shunt itself.

I did it like this:
Used a resistor wire, and measured with a short standart ground spring (pink trace), and did another measure with a longer ground wire (red trace), that had about 2-3 times as much loop area. Those two measures were taken on the same point (pretty close at least) of the resistor wire The second measure had much bigger Vpp values.

Therefore i did not measure to little noise, i just measured too much current.