• State
  • Locked Locked
  • Replies 4 replies
  • Subscribers 63 subscribers
  • Views 1500 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.

LMR62421 dual output rails

c_ 64
Intellectual 630 points
Other Parts Discussed in Thread: LMR62421, LM3224, LM2733, LM2731, LM27313

Hi,

I'm considering using the LMR62421 DC/DC converter in a design, requiring +/- 10V dual output rails. I was wondering if one can create the -10V rail in the exact same manner as is shown for the application on page 13 for the spec sheet for LM3224. Or are there any possible issues I have overlooked?  I'm using Webench to create the design for a single output voltage of +10V. Are there any known issues concerning using Webench? Or can I follow it "blindly" so to speak, given the correct input and output information?

Thanks!

  • 0
    TI__Expert 4365 points

    Hello,

    Yes it is easily possible to create both + and - 10V simultaneously with a negative charge pump attached to the switch node of a conventional boost. There are a number of pin compatible devices from the former National (now called Silicon Valley Analog division). The device LMR62421 is a new release that will work fine assuming you are starting from an input voltage less than 5.5V, which is the devices maximum Vin. There are other devices such as LM2731 LM27313 LM2733 which have a maximum Vin range from 2.7V to 14V. These other devices have lower switch current ratings and will not give as high a maximum +/- 10V output current as the LMR62421 device. The are certainly some other TI brand parts as well.

    The first thing to recognize is that only one output will be regulated, The +10V output in your case. The -10V output will be low in magnitude because it undergoes two diode drops in the inversion cucle and the boost side only suffers from a single diode drop. So the negative voltage will be low by a diode drop and have poorer load regulation because of the different forward drop across the two diodes in the negative path versus the single diode on the positive path. Balance (aka tracking) between the + and - supply could be improved by placing two diodes in series on the positive path at the expense of efficiency.

    The second thing to recognize is that there should be a small resistor (1 or 2 ohms ) in series with the second inverting diode and the negative output capacitor. This helps prevent the recharge of the inverting pump capacitor from tripping the switch current limit as the next switching cycle starts.

    Attached is a proposed schematic:

    0825.Schematic Prints.PDF

    Alan Martin

     

     

  • 0 in reply to Alan Martin
    Intellectual 630 points

    Hi,

    Thank you for your feedback and proposed design. I'm afraid I haven't had time to look at it before now. I would be interested in doing some Spice simulations of the circuit (the circuit it will fit on is very space limited, so I probably have to change some of the components that you suggested). Do you know if there are any Spice models of this component? And does it make a difference if I were to simulate it using PSpice or LTSpice?

    Thanks!

  • 0 in reply to c_ 64
    TI__Expert 4365 points

    Hello;

    I'm inquiring locally about the availability of models for the suggested devices.

    What are the specifications for your input power source; specifically input voltage min and max?  How much load current is being drawn from each 10V output rail?

    We might be able to simulate here and send some results.

    Alan Martin

    TI/SVA

  • 0 in reply to Alan Martin
    Intellectual 630 points

    Hi, and thank you for your repsonse and assistance.

    The input power source will be a single cell Lithium battery, with input voltage ranging from 2.7V to 4V (depending on the battery used). The output current draw is based on the other components' spec sheets, and amounts to 50 mA for the +10V rail, and 30 mA for the -10V rail. Due to tough space constraints, I won't have room for some of the specific components that was suggested. However, using the Webench tool (unfortunately, I could only do it for the +10V rail, but set the output current to be 100 mA), I was thinking about using as diodes the PMEG6010CEH,15 from NXP (specs: VFatlo=0.57V, Io=1A, VRRM=60V), and as inductor the SDR0302-120ML from Bourns (specs: L=12uH, DCR=0.28 Ohm, IDC=0.75A).

    Thanks!