Tiny-4 design, efficiency calculations

Posted by: micra on 07/13/2014

Dear colleagues,

I've designed and tested 5.0 V 2A AC-DC converted for universal AC supply (85-260 VAC) based on TNY288DG chip. Everything seems to working per reference book, however the problem comes up when measure efficiency of the converter.

I measure input rms current using Yokogawa 50 MHz current probe connected to oscilloscope (Tek or Le-Croy), or wideband Fluke multimeter. Any method gives me a very similar figures. If load current is set to 1 A and rms input AC voltage is set to 115V then I get ~110mA rms input current, thus efficiency is 5*1/0.11*115=39.5%. If I replace AC input voltage with 162.6V DC which is equivalent to 115VAC and applied directly to AC input I get input current of 42 mADC thus producing efficiency of 5*1/162.6*0.042=73.2% - an expected figure. The are no overheating components on the PCB during AC operation, input capacitance is 30 uF. Where do I make an error?

Thanks.

Comments

Hi,

If I understand well, you get a pretty normal efficiency data when you supply the power supply with DC voltage, with DC measurment. The data you got with a ac input does not look correctly, the reason is the AC input current measurement error. As the input current is a current is just a short pulse current with a lot of harmonic because of the diode bridge, you have to set up the scope correctly to do the rms current measurement. Could you show me the screen shot of your scope and how you did you ac rms current measurement?

In our lab, we are using the power meter (for example, Yokogawa WT210) to measure the input power.

Hi Terry,

Thanks for your response. Yes, you understood my problem. Please, find attached files as requested.

Thanks,

Michael.

Hi,

Thanks fo the photos. Now I see why you got a wrong efficiency data. As you know the input current and input voltage has power factor less than 1, so if you simply multiply the rms current and voltage, you will get the apprant power instead of the real power. That is why you got very lower efficiency data.

PI actually has a very good video to teach out end user to do the correct efficiency measurement by using the multimeters. You can refer to the PI university in the following link:

http://www.powerint.com/en/pi-university/courses/techniques-measuring-efficiency

Thanks a lot, Terry, it's all clear now after viewing video.

Regards,

Michael.