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.

TPS40322: How can I replace two TPS40140 by two TPS40322 to get an accurate four channel current source?

Part Number: TPS40322
Other Parts Discussed in Thread: TPS40140, TPS40090, LM5119, LMC555, LM5143-Q1, LM5143

Hi there,

I'm using the TPS40140 in a four channel design as laser power supply. Esp for its excellent current sharing, the wide 42Vin range and the multi channel functionallity. Now its NRND and the alternative is the TPS40322.  May be bad, but also a chance for  improving ...

Special features I need (according TPS40322):

transformation to a current source: ok

four channel design:  in principle yes,   the following issues are based on the 4 channel design

frequency sync: ok

regulation/error amplifier: By manual only the EA of channel2 of one controller could be deactivated, this would lead to a system with two parallel regulated circuits. Not good for pulsed applications. Is it possible to switch off both EA of one pwm controller and to drive them by COMP from the Master channel?

output voltage (current): Other controllers could be tuned by its softstart pin (a second noninverting EA input). Here I see a "start-up test procedere" to check internal stages at first, how can I use the SS pin as output control? 

dead time control/higher output voltages: For higher output voltages I would use an external gate driver. According internal cross conduction control, what output should be used to drive the external gate driver (to prevent additional dead time)?

Thanks in advance for some help, I could and I will make some simulations and tests with an eval board, but its really time critical.


  • Hi Frank,

    TPS40322 is not stackable. That mean you can not use 2xTPS40322 for 4phase design.

    Can you let us know the Vin, Vout, Iout in your application, so we can propose other part?



  • Hi Qian,

    thanks for the quick Reply!

    The output specs are: Iout from <1A ....50A;   Vout: 3V...20V   in a first step, input voltage  24V...26V. The next step would be Vout from 3V ...65V @ Iout=1A ... 50A,  scaled by power < 2kW.

    A very important target spec is ripplecurrent << 0.1%, thats why current sharing is very important. The 40140 was excellent according ripple and very transparent described. For higher voltages I have checked the TPS40090. But its current sharing seems not as good as the 40140. I also have no Information about the functionallity and the gain of the current sense correction.

    The Problem @ higher voltages was the resistor devider to come into the CM voltage range of the CS inputs, this also attenuates the difference voltage and the over all ripple was larger than the 40140.  But in General the 40090 was not bad, I only would need some "Insider Information to optimize the performance respectively to know the technical limitations.

    To summarize it:   I need:

    - multichannel to reduce the ripple

    - current source properties, means less output capacity

    - Control Input to set the Output current, the best way by manipulating the REF of the EA

    - excellent current sharing

    If you have some ideas, you are very very welcome....


  • Hi Frank,

    Can you take a look at LM5119? Two LM5119 can work tegother for 4 phases.

    For below application, what's Vin voltage range?
    Vout from 3V ...65V @ Iout=1A ... 50A, scaled by power < 2kW.

  • Hi Qian,


    The LM5119 looks not bad.

    According external current setting:  The manual says   The LM5119 regulates the FB pin to the SS pin voltage or the internal 0.8V, whichever is lower.  This means I coud control the output via the SS Input.  Is this correct or do I oversee something?

    According four channel design:  I need two external clocks, one 90° shifted from the other, to sync the system,  correct?

    Everything else looks fine, the current sense must be tested, and 48Vout as start is ok....

    Have a great weekend



  • Hi Frank,

    I did not try using SS pin to control the output current or voltage. The typical method to adjust output is to apply a voltage to FB pin via a resistor. 

    For 4 phase, you need two external clock with 180 degree phase shift. The duty circle of both clock signal is 50%. 

    LMC555+Inverter can generate two clock signals required. 




  • Hi Qian,

    I have checked a third opportunity, the LM5143-Q1. And this part have all I need on a first view, interleaving, current sharing, high Output voltage and a way to tune the Output (via Feedback Offset or SS Manipulation). If it is stable over the wide range of vout and iout must be checked.

    To start I modified the PSpice model of the according evalboard. It works fine with a fixed output voltage of 5V, but not if I change the feedback into a resistor divider. The Problem during simulation is that LO2 keeps at ground and after SS is reaching 0.6V the Output is interrupted.  The parameters are calculated by the LM5143 Controller design tool. Could this be a problem of the model? It would be important because I develop the model circuit step by step to a current source before I start with practical work.

    Thanks in advance


  • Hi Frank

    Sorry, I'm not supporting LM5143-Q1.
    Can you submit a new post with LM5143-Q1 in title?
    Other TIer will support you.

  • Sure,

    thanks so far...