How to measure current with a digital oscilloscope?

[emaq]

For a DC-DC converter how can I display and measure waveform of a current through an inductor and current through MOSFET switch using a digital storage oscilloscope?

For the inductor current can I measure voltage across the inductor and use some bulit-in math function of oscilloscope or use integration on an off-line data, say for example in Matlab?

For MOSFT I suppose it is difficult to measure drain-source voltage as R_DS(on) is very low and I cannot put a series resistor there. However I can measure the voltage across a capacitor and inductor through which the currents add up and flow through MOSFET, but then again I need to apply math functions to get the two currents and finally add those.

Please guide me in this matter.

 

[KlausST]

Hi,

How I´d do it: Resistor
I can´t see why you can´t use it. Tell us more about it.

alternatively:
use a current probe.

Klaus

 

[BradtheRad]

For the inductor current can I measure voltage across the inductor

Install a low-ohm resistor inline with the inductor. Measure voltage across it. Ohm's law tells you Ampere level. (Example, 0.4V across 0.2 ohm resistor means 2 A)

It's informative to observe voltage across the inductor, and to see the high and low peaks. However voltage waveform does not necessarily tell you Ampere level in the case of switched-coil converters.

For MOSFT I suppose it is difficult to measure drain-source voltage as R_DS(on) is very low and I cannot put a series resistor there.

Attach scope leads across the mosfet terminals. As you examine the waveform closely, measure voltage particularly when it conducts. Voltage is high when it's off and low when it conducts.
You cannot get an Ampere reading directly through the mosfet. Obtain the Ampere reading by other means. Then the mosfet ROn=V/A. (Example 0.4V / 3A=0.133 ohm)

To get usable readings on a component you must attach scope leads across it. These types of measurements are not always done with scope ground connected to circuit ground.

 

[asdf44]

Bottom line: if you don't have a good current probe it can be pretty annoying to accurately measure current.

Current probes for scopes start at about $500, if you can budget that then they'll make your life much easier.

**broken link removed**

 

[emaq]

How I´d do it: Resistor
I can´t see why you can´t use it. Tell us more about it.

Here is the design...

MOSFETS has R_DS(on) = 40 mΩ and adding a resistance between MOSFET source and ground will deteriorate the circuit performance. Similarly adding series resistance in the inductor path will introduce current spikes.

How about using hall effect based current sensor such as Allegro ACS758LCB-050B-PFF-T which has 100 μΩ current conductor? What if I use this sensor in series with inductor and also between MOSFET source terminal and ground?

 

[KlausST]

Hi,

Before you asked about current measurement .. now you talk about performance.

Why now "performance" is important? And in which way?

If you see a problem in the SOURCE path, why not use the DRAIN path?

We don not know what current you expect, thus we don´t know if 100uOhms is a good or not.
(When a 100uOhms hall effect is OK, why not use a 100uOhm resistor?)

We don´t know the waveform and we don´t know the timing, thus we cant say if Allegro part is good or not.
We dont know the series stray inductance of the Allegro part.
We don´t know what accuracy you expect, regarding current level, current waveform and current timing.


Klaus

 

[frankrose]

Also hard to imagine a voltage source with 100uOhm output resistance for me, I mean why not good enough 100mOhm in series with the V1 or V2 source then? More details would be really better.

 

[D.A.(Tony)Stewart]

Generally current sensors designed around 50 to 100mV drop at max current to limit power loss under 1/4W unless this is extremely low or high power.

Impedance control, BW and choice of high side, low side or different, is done with many IC's, floating? grid tied? earth grounded? rise time?

The best design has specs!

Consider ground side resistive 50mV sense into 50 Ohm connectors for coax into 50 Ohm termination in DSO for optimal BW. with dual channel for single and A-B measurements and eliminate probe error.

 

[std_match]

Normal current probes are clumsy. If you don't need exact readout of the DC current level, you can use a Rogowski Coil Current Probe. They are available in types that can measure one leg of a power transistor.
The DC reference level can often be calculated in some way, and is trivial if the inductor current goes to zero in the switch cycle.

 

[Kiriakos-GR]

Lets start over from scratch, what Oscilloscope model do you have available there ?
Toy scopes does not support current probes.

On the other hand, while I own excellent instrumentation regarding oscilloscopes and active probes, measurement accuracy comes only with specific class of bench-top DMM, these having by design lowest burden voltage at 400mA range and at 10A range.
The very few best in quality active probes, they can not measure lower than 200mA, this is why I ordered one DMM4050 (8846A).

There is also a specific probe made for mosfet with a tiny loop which it can pass thru mosfet leads, but I do not remember who makes it.
Mini rogowski coil.

 

[emaq]

If the conclusion boils down to use series resistance and measure voltage across it, the resistor itself dissipate lot of power... for example, for 10 A current and 0.02 ohm resistor, the power will be 2 Watts! And the peak voltage measurement would be 0.2 V... which perhaps would be dominated by noise.

 

[Kiriakos-GR]

You are unable to use any resistor because (Bench-Top) oscilloscope this is NOT isolated from earth ground.
If you need help ? then be honest and inform us of what measuring equipment you have there.

 

[KlausST]

Hi,

Are you aware that you never mentioned the current range?

It's on you to give useful and complete information.

You can't expect good assitance as long as we have to guess the requirements.

Peak current range, expected resolution or SNR, frequency range, voltage drop or power dissipation, and so on...


Klaus

 

[emaq]

I have Tektronix TDS 1002 oscilloscope for measurements.

Peak current range = 10 A
Expected resolution = 5 A/division
Frequency range = 100 kHz
Voltage drop < 1 V (of course, more voltage drop will create voltage imbalance and current spikes, see the schematic)
Power dissipation < 0.5 W

I restate the problem...
how can I display and measure waveform of a current through an inductor and current through MOSFET switch using a digital storage oscilloscope?

 

[BradtheRad]

IPeak current range = 10 A

This is in a range that creates several milliVolts across several inches of copper wire. I used this method to build my homebrew Amp-hour meter. I seem to recall I used 12 gauge. I looked at a table of wire gauges to find out resistance. I cut about 12 inches. Its ohm value was so low that it shouldn't make a difference in the overall circuit.

I connected DC power supplies to send a range of current through the wire, from zero to several Amps. I set my digital meter to the 10A range, and watched to compare its readings with voltage readings on another meter. I adjusted the setup to see what length of wire gave me the most usable readings. 8 or 9 inches of wire was the right series resistance which I used to measure current.

- - - Updated - - -

So if you connect your scope leads to each end of a thick wire, and insert the wire inline with the inductor, your scope should be able to display the Ampere waveform.

 

[KlausST]

Hi,

Peak current range = 10 A
Expected resolution = 5 A/division
Frequency range = 100 kHz
Voltage drop < 1 V (of course, more voltage drop will create voltage imbalance and current spikes, see the schematic)
Power dissipation < 0.5 W

I assume these are the scope's specifications.
But more interesting is what's the application's specifications and what requirement you have on the measurement results.

Already mentioned solutions are:
* resistor as I to V ... then as voltage measurement to the scope, or
* using a dedicated current probe

But you still ask:

I restate the problem...
how can I display and measure waveform of a current through an inductor and current through MOSFET switch using a digital storage oscilloscope?

What other solution or information do you expect?

With 10A peak and 1V max voltage drop ... you could use a 0.1Ohms resistor.
Power dissipation depends on RMS current ... which we do not know.
What's wrong with this solution?
What's wrong with a current probe?

Besides this you may use:
* a current transformer
* a current transducer (like LEM ones, with the ability to measure DC values)
* Hall sensor

Just do an internet search about "current measurement methods"....to find a lot of ideas, pros and cons...

Klaus

 

[Kiriakos-GR]

I have Tektronix TDS 1002

I restate the problem...
Frequency range = 100 kHz
how can I display and measure waveform of a current through an inductor and current through MOSFET switch using a digital storage oscilloscope?

Tektronix has for you a specialized probe this worth over 1200$
You may go around this problem by finding a used probe and this is the end of the road.
Feel free exploring by your self Tektronix product catalog, have a look at entire current probe sector, and then select your new probe.

 

[D.A.(Tony)Stewart]

Your rise time may be >10MHz if BW=0.35/Trise

For current sensing use a low side current shunt SMD R that drops 50mV max at max current. or less than the Ron, then amplify with DSO x1 coax or amplifier. For > 10MHz it is best to terminate in 50 Ohms at DSO for 1:1 coax.

Use calibrated 10:1probe with spring probes (no clip and gnd lead) for improved response near +20MHz due to probe ground SRF from long ground leads.

For high side current sensing use two 10:1 probes calibrated for a flat line on noise in A-B mode. again using coil spring for tip and ring on probes. ( for <=5mm gnd lead)

For clean ripple voltage use AC coupled 50 ohm terminated signals.

The key to DIff Mode sensing on DSO ( A+B invert) is short gnd leads and calibrated probes both on a large test 5V signal to get <5mV error or 60dB CMRR if possible.

 

[Akanimo]

Hi,

How I´d do it: Resistor
I can´t see why you can´t use it. Tell us more about it.
Klaus

Klaus had given you this very simple and low-cost solution in the first response to your question but you rejected it outrightly. Why?

I think it's simple to go with the low-ohmic value resistor in series with the inductor. Measure the voltage across it. If necessary, amplify the voltage with an Opamp. If necessay, filter out noise. Apply maths function and you're done. What's wrong with this solution? Performance deterioration? How much of it? Resistor - Current spike? How?

Please try out this cost-saving solution before you reject it.

 

[emaq]

Klaus had given you this very simple and low-cost solution in the first response to your question but you rejected it outrightly. Why?

I think it's simple to go with the low-ohmic value resistor in series with the inductor. Measure the voltage across it. If necessary, amplify the voltage with an Opamp. If necessay, filter out noise. Apply maths function and you're done. What's wrong with this solution? Performance deterioration? How much of it? Resistor - Current spike? How?

Please try out this cost-saving solution before you reject it.

Ok... I will give it a try... the prototype is ready with 5, 0.1Ω parallel resistors (0.1Ω/5), each of 5W.
I think 10A inductor current can be measured in this way (though DCR of inductor increases losses) but not sure about MOSFET current... although I also have (0.1Ω/5) between source terminal and ground to measure MOSFET current.

- - - Updated - - -

Your rise time may be >10MHz if BW=0.35/Trise

For current sensing use a low side current shunt SMD R that drops 50mV max at max current. or less than the Ron, then amplify with DSO x1 coax or amplifier. For > 10MHz it is best to terminate in 50 Ohms at DSO for 1:1 coax.

Use calibrated 10:1probe with spring probes (no clip and gnd lead) for improved response near +20MHz due to probe ground SRF from long ground leads.

For high side current sensing use two 10:1 probes calibrated for a flat line on noise in A-B mode. again using coil spring for tip and ring on probes. ( for <=5mm gnd lead)

For clean ripple voltage use AC coupled 50 ohm terminated signals.

The key to DIff Mode sensing on DSO ( A+B invert) is short gnd leads and calibrated probes both on a large test 5V signal to get <5mV error or 60dB CMRR if possible.

I couldn't understand what you mean... please explain in simple English.

What do you mean by low side current shunt SMD R? If that is a resistance between source terminal of MOSFET and ground, then I am already using a 20mΩ resistance to measure current (by measuring voltage) which is less than Ron = 40mΩ. What do you mean by "to terminate in 50 Ohms at DSO for 1:1 coax"?