• State Resolved
  • Locked Locked
  • Replies 7 replies
  • Subscribers 65 subscribers
  • Views 1050 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

TPS62095 vs TPS54426

Bob C
Intellectual 615 points
Other Parts Discussed in Thread: TPS54426, TPS62095

I'm seeking feedback regarding the TPS62095 vs the TPS54426, both 4A buck regulators.

I have used the TPS54426 before, and I was just testing it on the lab bench again.  I get solid efficiency numbers ( high 80's ) and very low ripple on the output, which is great.  However, I'm a little concerned that when I punch in numbers for a 4A buck regulator in the TI "Power Quick Search" box, the 54426 doesn't come up.

The top choice is the 62095, and I see that it has better efficiency numbers.  Looks like the BOM cost would be close.

It also looks like the 62095 is a much newer part.  And, I see that there is a family of these DCS parts just like there was with the DCAP-2.

Is it now an obvious choice to use the 62095 (DCS) over the 54426 (DCAP-2) for a new design, or does the former have some drawbacks?  I see that it switches at about twice the frequency and accepts a lower Vin max, but that's OK.  It's input would be a 5V wall wart.

Depending on the feedback I receive here, I think the next step will be to evaluate both of these on the lab bench for data such as ripple and switching / RF noise using near field probes. 

Thanks

Bob

  • 0
    TI__Guru**** 191445 points
    I'm not sure why TPS54426 does not show up in power quick search. I usually use webench for my part searching though. TPS54426 is certainly ok for 4 A solution. There are probably 80 - 100 devices in the catalog that can take 5 V input and generate 4 A output, depending on the output voltage required and the feature set needed. I would probably start with a detailed comparison of the datasheets. That may incline you to go one way or the other before bench testing.
  • 0 in reply to JohnTucker
    Intellectual 615 points
    Thank you for the quick reply John, and I will go back to the datasheets for a more thorough read. Glad to hear the 54426 is still a solid choice.

    As I pointed out in my original post, the 62095 appears to have notably better efficiency ( looks like I can get out of the 80's and into the 90's with it). Can you tip me off on things the 55426 / DCAP-2 might do better?

    And since the 62095 switches at a notably higher frequency, I thought a lab bench comparison for EM noise might be relevant.

    The application is to bring in a 5V wall wart and regulate down to 3.3. The 3.3 would serve a dual purpose as the 3.3V rail and input to a group of buck regulators & LDOs for lower rails (e.g., 1.1.V). Since I'm using the 3.3V rail to power some ICs directly, load regulation will be important. From the two datasheets, it looks like the 55426 may have a better load transient response?

    Any further guidance will be greatly appreciated.

    Thank you again

    Bob
  • 0 in reply to Bob C
    TI__Guru**** 191445 points
    It's been a couple years since I last worked with TPS54426. Indeed a key feature of DCAP2 control is "fast transient response". The adaptive on time control allows the duty cycle to change nearly instantaneously in response to load step changes. You do have to be careful with higher duty cycle designs. 5 V to 3.3 V is getting close to the limit. I am not familiar with TPS62095, so I can't really comment on it.
  • 0 in reply to Bob C
    TI__Guru*** 149640 points
    Hi Bob,

    DCS-Control and D-CAP2 both have good load transient response and are similar. The key difference in the architecture is DCS-Control has an error amplifier to precisely regulate Vout to its level. More details are here: www.ti.com/.../slyt531.pdf You compare figure 4 in the TPS54426 D/S to figure 9 in the TPS62095 D/S for the load regulation difference. Note that the y-axis scales are very different but it seems to me that the TPS62095 has better load regulation, as it should.

    The TPS54426 is rated to 18Vin, so this gives its lower efficiency due to higher voltage FETs used.
  • 0 in reply to Chris Glaser
    Intellectual 615 points
    Thank you Chris for the information and the link to the white paper. I'm going to take a good look at it.

    When I suggested that the 54426 had a better load transient response, I was referring to the comparison of Figure 6 of the 54426 datasheet and Figure 16 of the 62095 datasheet. I think the scenario depicted in these figures could be representative of quick load changes due to memories / buses switching. Please let me know if you disagree.

    Do you think the DCAP-2 technology in the 54426 is superior in any way to the 62095? If not, I probably should grab a DCS EVM and test them both out side-by-side.

    Thanks again

    Bob
  • 0 in reply to Bob C
    TI__Guru*** 149640 points
    Yep, we have EVMs for you to test and compare.

    The test setup for those two figures is different. Specifically, the rise time for the TPS54426 is much slower than for the TPS62095.
  • 0 in reply to Chris Glaser
    Intellectual 615 points
    "Specifically, the rise time for the TPS54426 is much slower than for the TPS62095."

    Thanks for pointing that out. I missed it in my review.

    Seems like I'll want to use a DCS-based part after testing it. I'll double check the features on the parts in this family tomorrow and choose an EVM.