fluctuating output with lnk3204D
Hello All,
I'm having troubles with a design. I was reading and finally I get to design a AC/DC from 85-265 AC to 5Vdc@100mA, following the design guide (an70.pdf) I have the attached eschema, but the output is not fixed. It is depending too much on the load.
With 30mA of comsum I get 7Vdc, with 50mA=6V, 79mA=5Vdc, 85mA=4.7Vdc.
I have a variable load (is a uC and a Wifi module) and i need to fix the output voltage.
I could add a 5.1V zener but I need that the output keep over 5Vdc
Comments
I'm guessing it is your choice of diodes. It may be a good idea to make output and feedback diodes the same type, as regulation depends on them having equivalent voltage drops.
Unfortunately, that's not going to happen. Inductor current is:
i = Vt/L.
Which for me was (160V * 1us / 680uH) = 235mA. That's a lot higher than the current in feedback diode, putting it at a different operating point. This delta voltage needs to be compensated for in feedback resistor divider. Using UF4007 diodes at 100mA load I found the error to be 0.5V. Using 4.12k instead of 3.48k solved the problem.
Wait, there's more. Current in feedback diode is not constant, but a function of duty cycle, which is determined by load current. Charge to the 10uF must be replenished because of the constant 850uA load. At light loads the duty cycle of output pulses can be very low, meaning higher pulse current through the feedback diode (and greater voltage drop). This is where you are getting variation in output voltage with output load.
Best solution may be to use larger diodes?
Hi pcampos,
After thoroughly reviewing your schematic there is one important connection that was missed. Looking at CBP in the schematic, the capacitor is solely connected across the BP pin and S pin of the LNK3204. This shouldn't be the case as you can see in the reference schematic of the IC. This must be connected to the main output via a diode and current limiting resistor. In your case it must be connected to DFB via an additional current limiting resistor. If you wish to not connect this capacitor to DFB via current limiting resistor, you may also do so but make sure that the voltage acrooss the BP pin does not fall to approximately 4.5v or else your internal MOSFET will turn off causing your power supply to not behave as expected. You may check this voltage level when operating normally using 100nF. If it is falling you may want to consider 1uF but this will also affect your current limit. Otherwise just connect it to DFB via current limiting resistor. For your reference you can refer to the following designs and data sheet:
https://ac-dc.power.com/sites/default/files/PDFFiles/der507.pdf
https://ac-dc.power.com/sites/default/files/PDFFiles/rdr506.pdf
https://ac-dc.power.com/sites/default/files/product-docs/linkswitch-tn2_family_datasheet.pdf
Hope this helps.
Regards,
Leo
thanks all. Finally was an error on schematic. CFB capacitor was connected after inductor (output) instead before (S). This solved the issue.
Thanks
Hi pcampos,
Looking at your schematic everything looks fine. However critical component/s must be carefully selected especially the inductor. May we know the specs of your inductor? or particularly what type of inductor are you using? If you have the part number or BOM (bill of materilas) of your whole schematic kindly send it to us for checking and verification. But put a high priority on the inductor first. You may opt to experiment by using same inductance value but higher current specification say 1A as a starting point and compare the performance. Hope this helps.