TPS62130A: Light load efficiency 1.5uH, 2.2uH

Hello Hirokazu,

Here the value that will actively change your efficiency is the DCR of the inductor you are using. 

L=1.5uH has a better efficiency than L=2.2uH, because it has 10.7mΩ DCR and L=2.2uH has a 19mΩ DCR.

The less DCR you have for an inductor the better your efficiency will be. It will imply less loss on your power line.

I also recommend you to have a look at the newer, smaller and newer TPS62147/8 which offer a bit lower Iq, some extra features like output discharge , and tighter output accuracy.

Thank you very much!

Best Regards,
Dorian

Hello Dorian-san

Thanks for your reply soon.
I would like to know the reason why the efficiency at 1.5uH is better than one at 2.2uH in case of PFM mode at light load.
I guess that the efficiency at 2.2uH may be better since fsw at 2.2uH will be lower than one at 1.5uH on PFM mode in light load.
Could you please explain the reason by using internal phenomenon at PFM mode?

Best Regards,
Hirokazu Takahashi

Hello Hirokazu-san,

If inductor is 2.2uH, indeed the fsw will be lower than when L=1.5uH because inductor will store energy more easily.

What you can try to see that DCR (losses) impacts efficiency is to use the 74438357015 (Wurth Elektronik) 1.5uH/20mOhm in your system and you will see that efficiency at 0.001A is slightly lower than when you use the previous inductor you mentioned with L=2.2uH/19mΩ DCR still at 0.001A.

Thank you!

Best Regards,

Dorian

Hello Dorian-san

Thanks for your support.
I tried simulation of Webench with 1.5uH/19mΩ, 15uH/10.7mΩ and 2.2uH/19mΩ.
The result at light load 0.001A is below.
(1) 1.5uH/19mΩ : 84.3%
(2) 15uH/10.7mΩ : 84.2%
(3) 2.2uH/19mΩ : 81.2%

The detail is the attached file.

TPS62130 effeciency(2019 06 25).pdf

(1) and (2) are same value DCR. But 1.5uH case is better efficiency than 2.2uH one. Do you have any reason except DCR?

Best Regards,
Hirokazu Takahashi

Hello Hirokazu-san,

At light load, as you mentioned because of a lower frequency, the inductor value is the main constitutive element of switching losses in the system. So a higher inductance value is good for light loads.

Also in PFM mode, WEBENCH model accuracy can be slightly worse. That's what you are seeing here.

You can see this by using for a fourth example (4) 2.2uH/35mΩ : 81.13%

Thank you!

Best Regards,

Dorian

Hello Dorian-san

Thanks for your reply.
We set below L and did the both of simulation and actual measurement.
(1) 2.2uH/82mΩ
(2) 1.5uH/58mΩ
(3) 1.0uH/40mΩ

TPS62130 effeciency 1.0-2.2uH(2019 06 25).pdf
As the result, 1.0uH case has the best efficiency.
Of cause, 10uH has the lowest DCR.
Do you agree that our result is correct?
Could you please check each inductance value are proper for TPS62130A?

TPS62130A has DCS-control, is it effect to improve the light load efficiency at lower inductance?

Best Regards,
Hirokazu Takahashi

Hello Hirokazu-san,

There is one other topic you might have missed. In topic 9.2.2.2 (External Component Selection), we recommend to set FSW to low when a 1uH inductor is used in the system. When you want to use a 2.2uH, we then recommend to set Fsw to High.

With this change you will see sensible efficiency values for both case.

If you want to know more regarding the light-load efficiency, you may have a look at the Power Save Mode topic of the following App Note:

http://www.ti.com/lit/an/slyt531/slyt531.pdf

Thank you!

Best Regards,

Dorian

Hello Dorian-san

Thanks for your information regarding External Component Slection.
How about 1.5uH? Which do you recommend 1.25MHz or 2.5MHz as Fsw?

Best Regards,
Hirokazu Takahashi

Hello Hirokazu-san,

I would recommend you tie FSW to a High logic state if you want to use a 1.5uH inductor.

Thank you!

Best Regards,

Dorian