Very Low Efficiency of LNK305
ello,
I've currently designed a power supply using LNK305, which the output is 5V 150mA.
I've followed all details found in AN37, except that, despite all the components characteristics recommended are followed, the components part numbers itself weren't followed because of cost.
The power supply worked fine, but the problem is its efficiency which is very low, different than datasheet describes.
I already tried increasing input capacitor values, what didn't change anything in efficiency and even tried to replace input diodes for a diode bridge, which helped a little in efficiency but not sufficient.
Just for the record, the power supply efficiency after change input diodes for diode bridge the results was a little better: 18% for light load (17mA) and 40% for a load of 100mA.
The design of the power supply was checked in the PIXI Designer, but I didn't find any mistake.
I've attached the .pixls file and the schematic.
Please, could someone help me with these issue?
Kind regards,
Marcelo Peres.
Comments
Hi,
Sorry for the delayed reply.
Can you try and replace R3 with a fuse and see if the efficiency improves.
Identify the components where you are getting most of the losses and any of the devices has more losses than expected.
Can you send me your efficiency data?
Regards,
PI-NANO
Hi,
In fact, I've already changed a bit the circuit. The R3 was burning up very often, even when I use RF1 as a fusible resistor of 8,2Ohms as recommended. I belive that such a low resistence of RF1 doesn't prevent R3 to suffer the high current pulse when the input capacitors are discharged in the sinusoid peak.
So, I've replaced the RF1 for a common through hole resistor, at least of 100 ohms, 1/4W and the burning R3 problem now seem to be absolutely solved.
Below is the efficiency data I recorded, by measuring input and output current and voltage.
| Input Resistor RF1 | Input Current for 80mW load [mA] | Input Power [mW] | Efficiency % |
| 100 Ohms | 4,6 | 0,5842 | 13,69 |
| 500 Ohms | 4 | 0,508 | 15,75 |
| 1K Ohm | 3,7 | 0,4699 | 17,02 |
| Input Resistor RF1 | Input Current for 470mW load [mA] | Input Power [mW] | Efficiency % |
| 100 Ohms | 12,5 | 1,5875 | 29,61 |
| 500 Ohms | 10,6 | 1,3462 | 34,91 |
| 1K Ohm | 10 | 1,27 | 37,01 |
I couldn't identify any component that are getting more loss than expected. Any suggestion about that?
I hope to hear from you soon! Regards!
Hi,
Where is RF1 in your schematic? I could not find the reference designator.
Can you send me your drain current and drain voltage waveforms at point of measurement.
Can you describe the measurement procedure and what are the mesusring instruments used for output voltage, output current, input power etc.
Can you do the measurements with DC input voltage and use multimeters for measuring output voltage, output current, input voltage and input current.
Is your output voltage is stable?
Can you send me the board layout as well?
Regards,
PI-NANO
Hi,
The RF1 is in the same place as recommended by PI application note. Attached is the schematic.
The procedure to do the measuring was the following:
- The input current was measured using an TRUE RMS multimeter (AC mode), between the Line and the FASE point;
- The input voltage was measured using the multimeter between Line and Neutral (AC mode);
- The output current was measured using the multimeter (DC mode), between the R4 and P1, by taking out FB1 from circuit.
- The output voltage was measured using the multimeter (DC mode), between P1 and GND.
- The input and output power was calculated using P = VI.
The output voltage is stable.
I would like to send you the board layout, but I can't send it in this forum. Could you provide any non-disclosure way to do it (email, skype, or anything else)?
| Attachment | Size |
|---|---|
| mb_schem2.PNG (49.8 KB) | 49.8 KB |
Hi,
You are measuring Apparent Power = RMS Voltage X RMS Current,
However to calculate the efficiency you have to measure the Active input power = RMS Voltage X RMS Current X Power Factor.
You can use power measurement meters to measure the active power and then recalculate the efficiency.
Regards,
PI-NANO
Dear all,
It's been almost 2 weeks and I still have no answer from any PI engineer. Is it always such a poor support like that?
I've seen PI engineers answering some questions from others, but there are some questions like this one that are completely ignored by them.
Could some one be kind and help me with the efficiency on LNK305?
Thank you all very much.