TI-PMLK Buck Experiment Board: TPS54160 & LM3475 - Review

Table of contents

RoadTest: TI-PMLK Buck Experiment Board: TPS54160 & LM3475

Author: ipv1

Creation date:

Evaluation Type: Evaluation Boards

Did you receive all parts the manufacturer stated would be included in the package?: True

What other parts do you consider comparable to this product?:

What were the biggest problems encountered?:

Detailed Review:


The PMLK Experimenters kit is a PCB with two Buck Converters namely, the LM3475 and TPS54160 . This kit is designed for those who wish to learn about buck converters however it also assumes that you understand how buck converters work. I could start by explaining how buck converters work however considering that this is a review for the PMLK BUCKEVM, I would rather just jump right to the point. The kit is of premium design and gold contacts and high current jumpers resonate with the quality of Texas Instruments products.


My primary interest in the kit was for the simple reason that wanted a standardised hardware that I could use to explain buck conversion. This kit comes close however there are some things the designers could have done better. In this review I will go through my experience with the kit and what could have been done to make it better.

A Tale of Two Regulators

As I mentioned before, the board comes with two regulators for experimentation. The LM3475 buck regulator is the smaller cousin which operates with a hysteresis comparator. The switched MOSFET is an external one  yet the component count is pretty low. A block diagram of the application is given below.


For details on how a hysteresis buck converter works, please visit the links(TI Application Note SNVA170B) at the end of the article. The important detail here is that the switching relies on the ESR or equivalent series resistance of the output capacitor. The switching frequency depends on factors such as input voltage, output voltage, inductor value, etc which requires careful work when designing for an application. The feedforward capacitor(usually across R(f1)) is required to feed the ripples back into the current loop for the hysteretic converter to work.


This kit allows the user to add C1 via jumper 9 to vary this feedback ripple as well as experiment with the values of output capacitors using J4-6. TP6 is used to monitor the ripple feedback voltage and TP1 can be used to monitor the switching frequency. Take a look at the schematic for the kit below.




The TPS54160 is the more complicated brother amongst the two and comes with a whole lot of bells and whistles. The look at the board and the higher BOM count is apparent but the kit has a lot of features for fiddling around. Take a look at the schematic below.


The most obvious difference is the absence of an external MOSFET and is integrated between the Vin and PH pins. The switching frequency can be toggled between 250 kHz and 500kHz using J22. The experiments include observing the effect of the switching frequency, input voltage as well as the load on efficiency and output voltage ripples. Two inductors are available onboard facilitating experiments on the effects of inductor selection on output ripples. There are five experiments for this topology and sample waveforms guide you through the process of interpreting results.



What is missing?

The kit comes with a disclaimer that a user must have a fundamental understanding of switching regulators, however adding a simple explanation would have been a lot better. The accompanying manual for this part, starts with experiment 6 and feels too obtuse for basic use. True that the case studies are well drafted but felt out of place and coming from nowhere.


Another missing part is a means to sense current. A small instrumentation amplifier integrated into the kit would have been nice eliminating the need for the current probe. A smart placement of the same could have enabled the sharing of the current sensor between the two half of the board. Putting two regulators on the same board serves no purpose whatsoever.


I made a DIY load to use with this experiment which works pretty well. Adding a simple circuit like that would also add value to this board considering that this is an experimenters board and such things could reduce the wire jungle that may be a consequence of the setup. Worth Elecktroniks uses LEDs as a load and that could serve as a makeshift sink in this case. Jan Cumps used a few turns of a wire around the sense resistors which works quite well in most cases. I have a similar solution which works well in most situations and is shown in the figure below.




Where are the experiments?

I did some but not all experiments in the book. The reason was that I found myself veering off into tangents such as looking for means add other inductors to the board or finding better current sensing methods and even analysing the ripples in detail. I ended up doing 5 videos on regulators and while editing them I realised that they were more about buck techniques rather than the board which is why I am putting them into a separate writeup altogether.




This review is about the PMLK kit and here is what I think. The board needs a better manual if you are targeting people who want to get started with or want to learn more about buck converters. For the advanced user, it is just another EVM with a manual that states the obvious. I found that with a few additions, this could turn into an excellent teaching aid and I am trying my best to compile the necessary literature. The idea should be to explain the fundaments, move to ideal cases and then compare the practical deviations from expected. The TPS54160 is extremely versatile and is good for people trying to understand the difference between the different inductor types available out there.


[1] http://www.ti.com/lit/an/snva170b/snva170b.pdf

[2] https://training.ti.com/buck-regulator-architectures-hysteretic-buck-regulators

[3] http://www.ti.com/lit/ds/symlink/tps54160.pdf

[4] http://www.ti.com/lit/ds/symlink/lm3475.pdf

  • True but that is the point. They had no idea where to begin hence the necessity of some basic references and stuff.

  • Doesn't it show they started a lab experiment before knowing the subject matter?

  • I agree on the current sensor and that is why I am probing alternatives. I have an INA169 board which I soldered recently though adding it to the mix will be a challenge. I really wished they added a current sensing circuit to the EVM itself.


    For the manual, I gave the same to some students and they were baffled initially with the structure of the experiments. It took a while for the newbies to grasp what exactly they were looking for and how to go about it. Four groups of two students each reported the same result hence my inference. Once I explained buck converters to two more pairs, the feedback was a lot more positive.


    Hope to see a better load from your ongoing research.


    sir, many thanks

  • Nice road test report.



  • Inderpreet,


    Jan Cumps used a few turns of a wire around the sense resistors which works quite well in most cases. I have a similar solution which works well in most situations and is shown in the figure below.


    I can't claim it worked well. I barely got a signal out of it because I'm using air as the medium to capture the magnetic.

    Your solution with a core is much better. You 'll grab much more of the magnetic field than my clumsy coil.

    What both your and my solution fail to do is to capture the DC component.


    I don't share your verdict on the manual. When you explicitly say that it's for EE students that know the theory, omitting that same theory is ok (I think ).


    I like the load!