Solution Finder Get Tech Support

Load regulation problem with TOP261EN

Posted by: Shikhar Tuli on

I have used the TOP261EN to design a 48V 3.5A supply using PI Expert tool. Schematics attached. The supply shows correct voltage at no load. Even at slight loads, the output voltage drops substantially. At 1A load, the output voltage is just around 40V. Is there any way I can correct the fault?

Thanks and regards
Shikhar

Files

Attachment Size
TOPSwitch-HX_PIDesign4.pdf 2.24 MB

Comments

Submitted by PI-SevenOfNine on 07/07/2018

Hi Shikar,

Thank you for using PI product on your design. In order to help you we also need the transformer specification. Kindly provide us the information. Thanks.

Best Regards,
PI-SevenofNine

Submitted by Shikhar Tuli on 07/07/2018

The transformer specifications are in the document previously attached. I have attached them here as well. The transformer core is ETD39/20/13. I have also attached the test report of the transformer manufactured. The nominal primary inductance is coming out to be nearly the same as that recommended, but the leakage inductance is slightly higher - 3.9uH compared to the recommended value of 2.53uH.

Attachment Size
Transformer test report 80.7 KB
PI Transformer recommendation 138.45 KB
Submitted by PI-SevenOfNine on 07/12/2018

Hi Shikhar,

Thank you for the information. It seems that your design is correct, can you double check your feedback network? Is it getting the correct biases? Also, can you disable the current limit first by shorting X pin to S pin. Anyway, you can also try the attached design by changing the Kp to 0.9 instead of 0.94. Thanks.

Best Regards,
PI-Sevenof Nine

Submitted by Shikhar Tuli on 07/16/2018

What biases should I check in the feedback network? I had checked some biases in the feedback network earlier (test report attached). After disabling R7 current limit resistor, slight improvement is observed in load regulation (in the attached report). I also tried to reduce Kp to 0.9. According to PI-Expert, for this, R11 (as in schematic previously shared) reduces from 1210 ohm to 1180 ohm. Not much change was observed in load regulation with this approach. (No load voltage was increased to around 58V, using a potentiometer at TL431 input).

Sincere regards
Shikhar

Attachment Size
TOPSwitch_supply_tests.pdf 66.43 KB
Submitted by PI-SevenOfNine on 07/19/2018

Hi Shikar,

In column 1 of your table, is it the output load? If that's the case, it seems that the regulation is dropping once you load it. Is V(B) the voltage across C8? Can you monitor the drain current? Cut the trace or lift D pin then insert a loop of wire to measure the current. Please check the frequency and duty cycle. Please Zoom In up to 2 switching cycles. Thank you.

Best Regards,
PI-SevenofNine

Submitted by Shikhar Tuli on 07/21/2018

Hi PI professional,
Column 1 represents the output resistive load in ohms. The V(B) is indeed the voltage at C8. We didn't have access to current measurement probe at the moment. I have attached drain voltage output at the oscilloscope. The ON-time is almost constant and the frequency is changing substantially. PF and THD are also coming out to be very poor - PF nearly around 0.5 and THD around 100-150%.

I will send current measurements as soon as possible.

Regards
Shikhar

Attachment Size
No_Load.jpg 4.42 MB
0.5A_Load.jpg 3.74 MB
1.3A_Load.jpg 4.46 MB
Submitted by PI-SevenOfNine on 07/23/2018

Hi Shikhar,

Can you measure the switching frequency? Or at least give us the time/division setting of the oscilloscope. This is a flyback design and don't have a power factor corrector so we expect the PF to be lower than 0.7. Thank you.

Best Regards,
PI-SevenofNine

Submitted by Shikhar Tuli on 07/28/2018

Hi PI professional,
The time/division is given in the bottom of the screen in every image. The frequencies for different loading conditions are as follows:
No load - 30 kHz
0.5a load - 21.7 kHz
1.3a load - 47.6 kHz
According to the datasheet, the device is working in variable frequency mode, which is because current into the control pin is greater than I(C01) and less than I(C02). Can there be any issue in the bias winding or the optocoupler? Is it because of the light loading condition? Should I try with a zener feedback?

Sincere thanks and warm regards
Shikhar

Submitted by PI-SevenOfNine on 08/02/2018

Hi Shikar,

Thank you for the information. Yes, there must be something wrong with your feedback network, can you double check the CTR of your optocoupler. Also please check the bias voltage whenever you load the output. Thanks.

Best Regards,
Ian Barrameda

Submitted by Shikhar Tuli on 08/04/2018

Hi Ian,
The CTR value of the optocoupler. is 160%, as recommended in the PI-Expert tool. The bias voltage measurements were previously shared (post #5), and attached here as well.

Regards
Shikhar

Attachment Size
TOPSwitch_supply_tests.pdf 66.43 KB
Submitted by PI-SevenOfNine on 08/10/2018

Hi Shikar,

I will give you the contact of our sales offices and you may call and ask our FAEs near your area to help in your concern. Below is the link:

https://ac-dc.power.com/sales/

Thanks.

Best Regards,
PI-SevenofNine