• State Resolved
  • Locked Locked
  • Replies 7 replies
  • Subscribers 63 subscribers
  • Views 689 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

TPS61028 Low Efficiency

Maxim Galkin
Intellectual 460 points
Other Parts Discussed in Thread: TPS61028, TPS61200, TPS61070, TPS61220, TPS61260, TPS61221, TPS61222

Hello!

I use two TPS61028 non isolated DC-DC in portable device, which supply from two batteries (two cell 1,2V each).

One DC-DC is used for digital 3,3V supply and other DC-DC is used for analog 5V supply.

I attach both schematics.

7127.Sch_Digital_Power_Supply.pdf

7128.Sch_Analog_Power_Supply.pdf

I changed input voltage (net Vin BATP) and I measured voltages and current consumption of my portable device with two working DC-DC.

I cann't reach efficiency values from TPS61028 datasheet.

I used multimeter Agilent U1252B for current consumption measurement on net VinBATP net and Agilent DSO-X 2012A oscilloscope for average (average time is 10 ms) voltage measurement.

Multimeter can measure current in three modes: AC, DC and AC+DC. I attached picture for understanding difference between modes:

I measured current consumption in all three modes and in case without load, with resistor load 3940 Ohm on 3,3V net and with resistor load 3940 Ohm on 5V net.

I attached results of my measurements:

0602.PE_BRPD_Debug.pdf

The graphs of efficiency is below:

I have a few questions:

1. Why current consumption of TPS61028 without any load (I measured about 100000 Ohm) is more than 2*45uA = 90 uA (I used two DC-DC)?

2. What kind of current consumption value is used, when measured efficiency in TPS61028 datasheet? (DC value or AC+DC value, or AC value)

3. Why efficiency of TPS61028 is low (about 15% less)?
Input and output capacitors I placed near IC without vias and so on. I placed inductor (10uH) on bottom side of PCB and connected with IC by means of two parallel vias.

4. May I increase efficiency if I will use TPS with fixed output voltage instead adjust version?

With best regards, Maxim Galkin.

P.S. Sorry for my bad english.

  • 0
    TI__Guru**** 150950 points

    I believe these two app notes (from the product page) should address your fundamental concerns:

    http://www.ti.com/lit/an/slyt412/slyt412.pdf

    http://www.ti.com/lit/an/slva236/slva236.pdf

    But I would also note that different inductors will give different efficiencies.  

    I would also recommend using newer devices, such as TPS61200 or TPS61070.  These are smaller and offer better transient response, etc.

  • 0 in reply to Chris Glaser
    Intellectual 460 points

    Hello, Chris!
    Thanks for url's.

    I measured current and voltages same that discribed in slva236.pdf document.

    I added electrolytic (tantal, Low ESR) capacitor 100uF on input. Measured current consumption values in DC mode is not changed, but values measured in AC mode and AC DC mode were decreased in 4-5 and 2-3 times.
    Conclusions:

    1. I have not answer on question #1.

    2. I have answer on question #2: " DC value of current consumption is used in efficiency measurement.

    3. I added addition capacitor after ampermeter for AC component exclude. I have not difference (in most cases it was absolutely equal) between values, measured in DC mode and DC mode with additional capacitor. But I am understand, that right measurement with additional capacitor.

    I use inductor 744028100 (10 uH, Ir = 0,6A, Isat=0,45A, Rdc=0,67 Ohm).

    I want to try use 744028006 (6,8 uH, Ir=0.75A, Isat=0,55A, Rd=0,4 Ohm) and 744028003 (3,3 uH, Ir=1A, Isat=0,85A, Rdc=0,22 Ohm) and compare results. I will measure AC noise (peak-to-peak) in all cases henceforth.

    4. I have not answer on question #4. I saw TPS61200 and TPS61070, but they have less efficiency in ligth load (0,5....5 mA). The package of TPS61200 is pin-to-pin compatible with TPS61028 and not smaller. The package of TPS61070 (SOT-6) is equal to package of TPS61028 and TPS61200 (SON-10) about 3x3mm. But I don't know, how choose DC-DC with better transient response. Can you give me url on materials about that choose? Current consumption of my portable device can increase to 30-40mA for a while 1 msec. The DC value of current consumption of my portable device is equal to 0,5...5 mA.

    With best regards, Maxim Galkin.

  • 0 in reply to Maxim Galkin
    TI__Guru**** 150950 points

    #1 should be addressed in SLYT412 that I sent.

    #4. No, they will give the same efficiencies.

    If your load current is that low, the TPS61220 is likely a better option.  There is also the TPS61260.  You can find devices that fit your power needs with the power quick search at power.ti.com.

    Smaller means smaller solution size as well.  Newer devices usually require a smaller inductance.  All these new devices also do not require a tantalum output capacitor as the TPS6102x family does.  Generally, the transient response and other characteristics are improved with newer device topologies   In short, you want to use a new device over an old one, all things being equal.

  • 0 in reply to Chris Glaser
    Intellectual 460 points

    Hello, Chris!
    I read slyt412.pdf and understood difference between Iq and no-load current consumption. I cann't find any information about TPS61028 no-load current consumption (equivalent Fig. 4 in slyt412). I want to know about measured current consumption values (column "II, mA" for Rn 3V3 =100000 and Rn 5V = 100000 in file 0602_PE_BRPD_Debug.pdf). It's about 1mA and less.

    Why new device and old device give the same efficiencies? Graphs from they datasheet is not equal.

    I have not answer on question: "If using TPS with fixed output voltage is more efficient instead adjust version?"

    I can use new DC-DC in the next PCB revision but this is requre a lot of time. My portable device is very compact. The TPS61260 is good, but I need 3,3V and 5V output. I like TPS61221 and TPS61222.

    P.S. I don't know, how choose DC-DC with better transient response. Can you give me url on materials about that choose?

  • 0 in reply to Maxim Galkin
    Intellectual 460 points

    And more.

    How you suggest, adding a ceramic capacitor (about 22-100pF) in parallel to resistor R95 and R80 of feedback divider will change DC-DC efficiency?

  • 0 in reply to Maxim Galkin
    TI__Guru**** 150950 points

    I answered your question #4 in my last post.  I also did not say that different devices will have the same efficiencies.  I should say that the same device will have different efficiencies when operated with different external components.

    We don't have a graph for the TPS61028.  You should take this data yourself.

    I'm sorry that you have already designed your PCB with the TPS61028.  This is the wrong device for a < 50 mA Iout application.  To avoid this in the future, I would highly recommend ordering our evaluation modules (EVMs) which give you a working circuit and layout.  You can modify this with your external components of choice and see the performance changes before making your own PCB (at considerable time and money investment).

    A feedforward cap will not change the efficiency as it only affects the small signal response.

    There is no single document that compares the transient response of newer and older devices but generally newer devices have higher switching frequencies.  This allows higher loop bandwidths and smaller inductances.  Both of these improve transient response.  I would not be worried about transient response with so small an output current.

  • 0 in reply to Chris Glaser
    Intellectual 460 points

    Hello, Chris.

    "You should take this data yourself." - It's pity.

    Is it a good idea to use TPS61221 and TPS61222 in next revision of PCB?

    With best regards, Maxim Galkin.