Output voltage overshoot upon restart with TOP254 design
I built a power supply based on the attached design that I created using PI Expert Online. However, I am using Fairchild MOC206M (data sheet attached) for the feedback opto-coupler instead of the one shown in the attached PI Expert file. The supply is up and running but I am having an issue with output voltage overshoot. I added a “soft-finish” circuit that largely eliminated the overshoot at start up except for one condition. If input power is removed and then re-applied right when the 3.3V main output starts to drop, output voltage overshoots as high as 4V before settling back to 3.3V.
Worst case condition is high-line (260Vac) input and min load on the outputs. Min load is 3.3Vout at 14.1mA; 5Vout at 1mA; and 5VIout at 0.5mA.
Pictures of 1) A normal high-line min load start and 2) A re-start just as the 3.3Vout rail dips are attached. Note that the peak on normal start up is 3.37V while the peak output voltage on restart is 3.98V.
I need to identify the best approach to eliminate the re-start overshoot or reduce it to less than 3.6V.
Comments
Hello and thank you for the feedback.
1. Removing the soft-finish circuit did not resolve the overshoot issue following a line dropout. In addition, overshoot is now present during any start-up condition.
2. With the soft-finish circuit still removed, I applied 15Vdc across the C7 bias winding capacitor. The overshoot issue still persists during a line dropout condition and during any normal start-up condition.
Thank you for your assitance.
Hi,
I would like you to try couple of other things.
1. Remove resistor R8 which is placed across optocoupler LED as first step and see if this helps.
2. If removing, R8 does not help, place a 1KOhm resistror in series with C14 as a second step (R8 should still be removed in this condition as well).
3. if the above two steps does not help then, also change U2 to LTV817A and see if it helps. Last step is for debug purpose only.
Regards,
PI-NANO
Hello,
1. Max observed overshoot with R8 removed is 3.75V. It was 3.98V before. So there is some improvement.
2. Also added 1k ohm in series with C14, R8 was still removed. Did not see any additional improvement.
3. Did not have an LTV817A so I tried a Vishay IL213AT, which has a higher min CTR than the MOC206. The result was about 50mV of additional improvement. R8 still removed and 1k ohm still in series with C14.
The soft finish circuit was still in place for each of these tests.
Hi,
Good to hear that it has reduced the overshoot upto some extent.
Now I would like you to try couple of more things.
1. Replace R7 with 0 Ohms resistor and see if it improves.
2. Try reducing capacitor values of C8 and C10 to anywhere between 330uF to 1000uF and see if it gives any improvement.
3. Also increase the 1k ohm to 10k ohm in series with C14. Also try different vales for C14 from 22nf to 100nf and see if it makes any difference.
Regards,
PI-NANO
Hi,
Could you also please send me your PIExpert file as well. I would like to check with our PIExpert design team.
Regards,
PI-NANO
Hello,
1. Please remove the soft-finish circuit and verify the same test condition. Please let us know if removal of the soft-finish circuit appears to resolve the overshoot following a line dropout and restore.
2. If the overshoot problem persists, please remove the soft finish circuit (D10,C13, R6) and connect a 15V supply across bias winding capacitor C7 and repeat the experiment. If the overshoot problem appears to be cured, this suggests lower than expected bias winding voltage. For this the remedy could be a higher capacitance for C7 or one extra turn for the bias winding or both.
Please let us know what you find and we will help to resolve this issue.
Note that the removal of soft-finish circuit in points 1 and 2 above is for debug purposes only. Once we have found the cause of the overshoot issue, we will ask you to add it back.
Regards