# Which low pass frequency filter to configure to measure/calculate DC-DC converter efficiency?

I am measuring the DC-DC converter active power (P in watts) efficiency of a TI (Webench) LM3150 PCB that converts 12 to 5VDC.

In the answer of question Measuring precisely the efficiency of a power converter it is suggested to use:

• an averaging function on the (multi)meter
• low pass filtering to get rid of any high frequency measurement noise

Our Zes Zimmer LMG95 precision power meter has got several low-pass filtering options available:

1. AAF: analog anti-aliasing-filter
2. AAF + 30Hz
3. AAF + 60Hz
4. AAF + 87.5Hz
5. AAF + 175Hz
6. AAF + 1.4kHz
7. AAF + 2kHz
8. AAF + 2.8kHz
9. AAF + 6kHz
10. AAF + 9.2kHz
11. AAF + 18kHz

The frequencies here are digital low-pass filters with a cut off frequency of about ... Hz.

Which low pass filter - I can choose only one - to best configure in this active power efficiency measurement setup? And why choose that one?

## 3 Answers

Input voltage is DC and this is dead easy to measure with a DC volt meter. Current into the power supply will be dc plus an amount of AC ripple. The AC part is irrelevant and any old meter reading dc amps will measure average.

The result is: -

DC input voltage multiplied by average input current = input power in watts.


It is incorrect to measure RMS current because this will give the wrong answer when the input voltage is DC.

Same with the output - if the output voltage is largely DC i.e. it has very little ripple voltage then, output power is: -

DC output voltage multiplied by average current into the load.


Applying a filter makes no sense if you are trying to measure power when one of the consituents of power (namely voltage) is constant dc.

As for using a power meter with a filter, you'll probably find that it won't make much difference which setting you use. Try it and prove it.

• The power efficiency difference with and without AAF + 30Hz low pass filter at 0.05s cycle time, also depends on how one presents its numbers. The difference is 0.2% when comparing 95.7% (no filter) with 95.9% (filtered) or 0.14% when comparing 95.73% versus 95.87%. At 0.8000A load, without filtering both the input and output wattage are lower: 4.2318W and 4.05102W (=95.73%) versus 4.24042W and 4.06529W (=95.87%). – Pro Backup Apr 21 '15 at 18:02
• I'm not sure what you are trying to point out but my answer still stands and is correct. If some form other of filtering gives the same result then that is a side issue. Please make your point clearer if you wish. – Andy aka Apr 21 '15 at 21:08

Zes Zimmer, the manufacturer of the measuring device, suggests:

### Only DC signal: AAF + 30Hz

If you have a only a DC, without any pulse signature on it, use the AAF + 30Hz low pass filter, to get the noise of channel as low as possible.

### Additional AC signal or pulse: AAF

But If you have also an additional AC signal or a pulse on top of the DC, you have to choose a higher filter like the analog anti-aliasing filter.

### No filter

My electro technical engineer tells me to disable all filtering to include each switching effect into the DC power conversion efficiency calculation.

• This isn't really an answer to your question. – Andy aka Apr 21 '15 at 21:10