Linkswitch-PH efficiency question

Kris -

Could you please attach a schematic for the design you're working on (a screenshot/PDF is fine). Also, what is your test/measurement setup? Are you using a power meter for these measurements, a DMM, a scope, etc?

-The Traveler

Hi,

Thanks for looking into this. However, I am more worried about what PIs design SW says about this. I attached the saved file in the previous post. The numbers quoted are frem the design SW, *not* from measurement.

To answer your question: Irms and voltage were measured using a Fluke DMM (True RMS measurement).

Kris

Kris -

I think the problem you're experiencing comes down to your method of measurement. For this kind of situation, you'll need a power meter such as the Yokogawa WT210's that we use in our lab.

Also, the 0.13A value that you're using appears to be the primary winding RMS current value taken from the PIXLS spreadsheet...this is not necessarily going to be the value of the PSU input current.

It is possible (but not recommended) to make this measurement using an oscilloscope with a current and voltage probe. However, it's a bit complicated and you have to take the time delay of your current probe into account.

-The Traveler

Hi,

Thanks for the pointer on the fact this is the primary winding RMS current. That did solve part of the puzzle. However, I'm still stuck with an efficiency of about 68% where it should be about 85%.

The method I'm using is the one in the PIuniversity for those poor folks without a regular power meter... Using a true RMS DMM. I am, however, wondering if this method holds for the Linkswitch-PH device as it is a single stage PFC topology... Perhaps even a true RMS DMM doesn't like 66kHz on/off?

Schematic in attachment. Please note the CM choke, fuse and the input cap of the filter are on another page.

In general, DMM's have a fairly low cut off frequency. Even if your DMM can give an accurate RMS input current at the frequencies we're dealing with, it's not going to relay anything about the phase relationship with the input voltage. As you know, this is critical for calculating the total input power.

Without an actual power meter, you're going to be stuck with using an oscilloscope to calculate your input power. You'll need a current probe, x100 voltage probe and an oscilloscope capable of performing a math waveform that is the product of the current and voltage waveforms. You'll then need to make measurements over several complete line cycles to come up with an average input power.

Depending on how much ripple current there is in your output, you can use a DMM to make the output power calculation. Measure Vdc, Vrms, Adc and Arms at the output of the power supply. Ptot = Pac + Pdc = Vdc*Adc + Vrms*Arms.

If your measurements with the DMM are correct and your system efficiency really is 68%, your design is dissipating somewhere around 4.5W based on a 9.5W output. This wasted power should be easy to find as the guilty components should be getting pretty hot.

In regards to the power meter, this is one of those situations where having the right tool for the job is necessary. It will save you time, frustration and money (in the long run). Having the right tools for the job is critical in any field you work in.

-The Traveler