EEVblog Electronics Community Forum

Electronics => Projects, Designs, and Technical Stuff => Topic started by: madhu.wesly01 on October 06, 2017, 08:57:43 am

Title: Impact of 10% Ripple at output supply in switching regulators.
Post by: madhu.wesly01 on October 06, 2017, 08:57:43 am

Hi,
I am working on building a power management PCB based on TPS54295 IC, for achieving the constant output loads(Dual output) of 3.5V, 2A and 3.3V, 1A. Input voltage range is 4.5 to 12V. One problem with switching regulators is - Ripple and Noise can be higher than Linear regulators. Since we need constant supply of 3.5V, it is becoming a battle-neck to achieve constant voltage, Is it?
I have implemented the power circuitry in Webench for the TI IC, Please find attachment. I got the outputs with 10% Ripple. My doubt is how will the Ripple of 10% effect the output supplies?
If it affects, how can we compensate it?

Title: Re: Impact of 10% Ripple at output supply in switching regulators.
Post by: capt bullshot on October 06, 2017, 09:25:50 am

The problem is, this report doesn't give any useful explanation of the 10% Vout ripple mentioned on the first page.
Looking at the output Vp-p graphs later on, everything appears to be OK with this design regarding the output ripple. An output voltage ripple of 10% would certainly be inacceptable for such s kind of supply.

Title: Re: Impact of 10% Ripple at output supply in switching regulators.
Post by: madhu.wesly01 on October 06, 2017, 10:19:27 am

Hello @capt bullshot,

Thanks for your response.

After I have gone through the following data-sheet, I came across few doubts, can you/(Anyone) please give your explanations in lines to the following doubts:
http://www.ti.com/lit/ds/symlink/tps54295.pdf (http://www.ti.com/lit/ds/symlink/tps54295.pdf)

Output requirements: 3.5V, 2A and 3.3V, 1A.
Input voltage around 7.5V

Ambient temp: -40 to +60 degrees

Given in data-sheet:
Junction temp: 150 max
Junction to Ambient temp: 34 degrees/watt

Doubts:
1. From above values, can you please tell me will there be any over heating issues? What is the formula for calculating Max. Dissipated Power in SMPs?

2. I want to implement the same application circuitry provided in the data-sheet, Will it work? (consider I will changing Voltage divider resistors values, and Inductors accordingly to achieve desired outputs)

3. The output ripple given in an example as 30mV, will it effect more?

4. If my load takes only 1A, then will the remaining current get converted as heat?(consider the ICs Dual output current is 2A)

Awaiting your reply.

Thanks & Regards,

Title: Re: Impact of 10% Ripple at output supply in switching regulators.
Post by: capt bullshot on October 06, 2017, 10:59:46 am

1. Can't tell this from skimming over these values. There's some efficiency / power disspipation graphs in the report, you may be able to figure out the answer from these and the datasheet values. I practice, the thermal resistance of your layout is a matter. Usually there won't be overheating issues as long as you stay within the datasheet limit of output power. Beware of optimistic datasheets, you may not be able to reach these limits in a real world circuit.
2. Usually: yes. Consider ordering a demo board from TI using this regulator, this eases your work
3. I don't understand the question. If 30mV ripple is OK depends on the circuit that you want to supply from this regulator.
4. No. The regulator will just draw less current from its input.

Title: Re: Impact of 10% Ripple at output supply in switching regulators.
Post by: madhu.wesly01 on October 06, 2017, 11:27:45 am

Thanks for your prompt response. :)

In data-sheet it is mentioned that to use bulk capacitors at input and output. Why is Bulk capacitors actually used? Is 0603 comes under bulk capacitor?

Awaiting your reply.

Thanks & regards,

Title: Re: Impact of 10% Ripple at output supply in switching regulators.
Post by: capt bullshot on October 06, 2017, 11:45:26 am

Bulk capacitor is a common term for the capacitors used at the input of a power supply circuit or at the power input of a random circuit. Often this refers to some kind of electrolytic or tantalum cap. The exact requirement on the bulk cap depends on the actual circuit, today often a ceramic cap is adequate. I you follow the datasheets specifiations for the required capacitors, you've got a good starting point. Maybe you'll need some experimentation with the real circuit to optimize it or to cover a raising issue.

One issue that might raise is the ringing of the input voltage of your regulator if you use only ceramic caps. Depending on the circumstances, the input voltage may ring excessively due to the low ESR of the cap and high Q of real or parasitic inductors. This can be cured by using an electrolytic cap in parallel to the ceramic one, or an artificially raised ESR of a ceramic cap (by putting a small resistor in series to the capacitor).

Depending on rated voltage and capacity, a 0603 may be a valid bulk capacitor. I'd expect them to be somewhat larger.

Title: Re: Impact of 10% Ripple at output supply in switching regulators.
Post by: madhu.wesly01 on October 06, 2017, 11:59:57 am

Thank you @capt bullshot.