Investigating radiating magnetic fields of different forms of inductors on PCBs.

[David_]

Hello.

In the process of designing a SMPS circuit I have come to have a great interest in determining the ESL of capacitors and capacitance of coils as well as getting some grip of how inductors can interfere with surrounding circuits and I want to build a device with a 16x2 LCD display to be hold on one hand while probing with a connected probe with the other hand.

I have a couple DRV5053 Analog-Bipolar Hall Effect Sensor, datasheet at:
**broken link removed**

I don't remember and to me it is not crystal clear but I'm pretty sure that I have the OA variety which means that the sensitivity is -11mV/mT.

"The absence of a magnetic field will result in OUT = 1 V. A magnetic field will cause the output voltage to change linearly with the magnetic
field.
The field polarity is defined as follows: a south pole near the marked side of the package is a positive magnetic
field. A north pole near the marked side of the package is a negative magnetic field.
For devices with a negative sensitivity (that is, DRV5053OA: –11 mV/mT), a south pole will cause the output
voltage to drop below 1 V, and a north pole will cause the output to rise above 1 V."

The operation of the device is very clear but I have no prior experience with this kind of stuff and I have no idea what 11mV/mT entails in relation to the kind of magnetic fields produced by inductors, if I have a accurate reading of the strength of a field and the distance to the coil is it possible to calculate the current in the coil?

I need design ideas on what might be appropriate to include n a circuit to find out what kind of fields there are around PCBs.

I should have begun with asking: Is this device at all applicable for the purpose I have in mind?

This thread and my other thread about measuring parasitics is a strive for me to get a better understanding of electronics and to develop a more intuitive understanding for the subjects since I am a complete stranger to this area of electronics.
But it also have real use for my circuit design and optimization in many instances.

Regards

 

[schmitt trigger]

A simple approach that will visually tell you where to start probing is with powdered iron. It is fascinating to see how the iron powder clings to high current traces. And how power inductors develop "whiskers", all of which fall off when the circuit is unpowered

 

[David_]

First I thought to have a 16x2 LCd disply shoing mT in decimal numbers but might there be a "better" way to visualise the results...

Just a thought and this is not a question if I could build such a thing but do you think that it is teoreticly possible to somehave get magnetic fields to show up to be seen by the naked eye on a disply through a camera or somehing alike?
It would be cool.

Hey, what if I would use matlab to do scans of areas, just as a example to explain what I mean.
A toroidal inductor is laying flat down on a PCB with current running through it, and I take the prope and at a distance of 1cm I start some centimeters away from the coil and then swep back and forth each sweep moving closer to the coil, and keep doing that untill I have sweept like a square over the coil. And then have matlab do some approximation f the sweeps and field strength plotted out to see something...

what do you think of anything in that direction?
Maybe its just something I find to be fun to do...

As for the circuit, do you know anything that you could relate 10mT or 100mT to in form of components or currents?
Or maybe I shouldn't focus on the actual measured value of a field in different locations but rather the propagation of any field strength?

I might have some difficult or strange questions now before actual taking some readings but what stopping me from starting is that I am pondering if a should just hock the sensor up to a ADC or if processing the sensor output with some opamp circuits could enhance or benefit the device operation in any way...

Any thought anyone?

Eh, I will test both approaches anyway,but suggestions on circuits is greatly appreciated.
Or there might not be anything to do other than scale to ADC input range and maybe filter for some frequency.

I guess the probe handle should be of non-conductive material so as not to interfere with the fields?

 

[chuckey]

I would use an analogue meter to search for magnetic fields, with auto ranging if you want to be ambitous.
Frank

 

[FvM]

I presume your "toroidal inductor" example refers to an inductor with toroid core.

In this case, the magnetic field is mostly kept inside the core and only residual fields can be measured outside. There's no simple or unequivocal relation between inductor current and outside field because it strongly depends on inductor design deficiencies.

So the example gives a glimpse on the limitations of field probing methods.

You also didn't clarify if you want to probe DC or AC fields, and if the latter, in which frequency range. Automatic RF E and H field scanners for PCB boards are offered by several vendors for EMC measurements, but you seem to focus on DC or low frequent AC.

Semi-quantitative measurement of PCB trace current is possible with the TT I-prober https://www.tti-test.com/go/iprober/

 

[David_]

The reason I acquired the sensors was as a stab in the dark while trying to design a device o measure magnetic fields on PCBs, I cant tell you what specifications I need cous don't know and I started this project to learn.

I have found a much better sensor from analog devices, AD22151 which can be adjusted to pretty much any gain.

However I would like to ask you to help me deduce what I am looking for f a device.
Home-brew is my only option since I can't afford any real instrument, but the fisrt design in which I would like to investigate fields and the propagation of magnetic/electric fields is a SMPS power supply. Its a pre-regulator for a linear lab power supply and will output between 3 and 30 volt.

If I want to investigate magnetic fields on a PCB and to learn and get a understanding for how magnetic fields acts and behaves what kind of device would you say is what I need?
Is there any suitable device which I can design on my own home-brew style?

 

[schmitt trigger]

As I mentioned previously, if you require a "visual", the iron powder will do the trick first. You can then use your hall sensor to probe around.

The key with the magnetic fields, is that they are vectors, and the probes have to be aligned to magnetic field's direction to detect them properly. Rotate the probe 90 degrees with respect to the correct orientation, and your reading will go to zero.

 

[jiripolivka]

For PCB trace magnetic field shape and intensity, the best device is a small wire loop pickup connected to an oscilloscope. The pickup coil can be made of 0.1 mm dia.wire on one-cm coil, and attached to a wooden "pencil" to hold it.
Best used with AC signal generated by e.g. 0.1 A at 50 Hz, you can calibrate your coil by passing 0.1A DC current through CB line and use DC oscilloscope coupling to see coil voltage output. Search for a maximum and minimum passing the coil above PCB, and turn it by up to 90 degrees to "map" field distribution

- - - Updated - - -

According to the oscilloscope sensitivity, use 1 to 100 turns to make the pickup coil.

 

[David_]

What about taking a thin trasperant plastic sheet and placing it over a PCB and then dust with iron powder...

I will test the hall sensor I have as well as the coil idea, thanks for the tips.
So is it true that I am not to be concerned about what strength or value a field could be said to have?