Power measurements for 40 kHz signals

[manytoolsmike]

Hello EEV brains,

I need an instrument for accurate power measurements of High(ish) frequency signals, and for standard 230 Vac 50Hz power consumption.

I need accurate Power measurements for 18 kHz to 50 kHz AC signals. The voltage ranges from 100 to 400 Vac, the power is under 10 watts.

There are a few issues that I am having:

1. The signal is a series of 200mSec pulses with 200mSec pauses in between that repeat every 5 seconds. so I need an average of this.
2. Each pulse is a different frequency between 18 kHz to 50 kHz.
3. The "power factor" varies a lot depending on the frequency (the current can be significantly out of phase with the voltage)
4 The Voltage and Current vary wildly during the six second signal.

In the video below I am using a Signal Generator to drive the Transducer (this video is measuring the frequency response of the transducer). You can see the current and voltage traces on the screen. This is the kind of signal I want power measurements for, but the voltage and power will be much higher.

https://youtu.be/u1-cBjFJvXs


I want to measure:
1. Average power over a 30 second or more period.
2. Peak power
3. Watt hours
4. Phase angle/ Power Factor

I am having similar issues calculating the Power consumption of the switch mode power supply driving the circuit.  When I take a running average of the inline current and multiply that by the voltage I get a massive difference between the DC and AC power. I know the PSU is running inefficiency due too the extreme transient nature of the current draw, but it cant be this bad..... 16 watts on the dc side and 60 watts on the AC! I suspect the numbers are inflated by a very poor Power Factor. And so I want a Power meter that details all the values

Here is a video showing my setup and math... These videos where made for a "different" audience so I am not using technical language.

https://youtu.be/-RUdtZC3PCA

To be clear this is a separate type of measurement, here I am trying to measure the PSU efficiency and the overall system efficiency

Is there an instrument that can do all this?

Thanks for your time,
Mike


 

[mawyatt]

You might consider an old analog type multiplier for the current and voltage waveforms, then deal with the resulting multiplier output with filtering and/or AD conversion. Also you could digitize both the current and voltage waveforms with an ADC with a high oversample ratio (~10X) then do all the power computations in software.

In either case make sure your anti-aliasing filter doesn't attenuate any higher frequencies if you want to get an accurate representation of the actual waveform power.

Best,

[JohnPi]

Most digital oscilloscopes can do math on the waveform.

So, use a current probe to measure current on one channel, and measure voltage on the other.

A scope with over 1 M samples (e.g. Rigol DS1054Z can do 24 M samples) can certainly record sufficient data (24 M samples in 2 channels == 12 M samples/channel. Over 5 s, this is a 2 MHz sample rate -- plenty for a 40 kHz signal).

[David Hess]

You did not say how much you expect to pay but many modern DSOs include automatic power analysis software which does exactly what you want, and the ones that do not can still do the analysis manually.

[TurboTom]

Most digital oscilloscopes can do math on the waveform.

So, use a current probe to measure current on one channel, and measure voltage on the other.

A scope with over 1 M samples (e.g. Rigol DS1054Z can do 24 M samples) can certainly record sufficient data (24 M samples in 2 channels == 12 M samples/channel. Over 5 s, this is a 2 MHz sample rate -- plenty for a 40 kHz signal).

Be careful with this generalization: The pulsed character of the power consumption requires the scope to be capable of applying math to the complete sample memory and not just to what's on the screen. IIRC, none of the "legacy" Rigol scopes is doing this. Even the "upper" scopes of this series that may include a power analysis function probably won't perform well at this task in "on line" mode. But since the power analyzer software will work off-line with CSV files as well, in this operation mode it should be possible to solve the problem, yet it may take some time and won't be as simple as pressing a button.

[The Soulman]

If you're driving the transducers from a low impedance fixed voltage source, wouldn't it be more accurate to
make an impedance measurement of the transducer in its intended environment? (under water?)
And calculate power according to the input voltage?

Also your power supply efficiency isn't great, 50 watts in and 15 watts out? 
I don't  have a rigol but 600mA average isn't the number I get from eyeballing the trace, the 1,14A RMS figure looks more like it.
That would make it 27 Watt on the DC side.
You'd likely could get away with a much smaller power supply than you have now, that would also make it more efficient.