• State Resolved
  • Locked Locked
  • Replies 3 replies
  • Subscribers 62 subscribers
  • Views 91 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.

Original question:

TPS40170: Paralleling two TPS40170

TPS40170: Issues with Parallel or Multiphase Operation. Purpose of Sync Pin.

Nicholas Rodriguez
Prodigy 10 points
Part Number: TPS40170
Other Parts Discussed in Thread: LM5145, LM5148, LM25148, LM5143, LM25143

My question is similar to those found in the following other threads:

  1. https://e2e.ti.com/support/power-management-group/power-management/f/power-management-forum/971312/tps40170-paralleling-two-tps40170?tisearch=e2e-sitesearch&keymatch=TPS40170#
  2. https://e2e.ti.com/support/power-management-group/power-management/f/power-management-forum/920981/tps40170-tps40170-can-supports-multi-phase-function?tisearch=e2e-sitesearch&keymatch=TPS40170#

Question(s):

Is it possible to connect two TPS40170's in parallel to provide power supply redundancy? If not, what is the intended use case for the "Sync" pin?

Details:

In our prototype design, we have two TPS40170 converters created. They are independent circuits except for the input voltage and output connection.

  • They have independent feedback/compensation networks.
  • They have independent switches.
  • All other input pins other than Vin and the output from the inductor are independent.

When we run this combination only one of the converter's seems to be supplying the load (the master) and there is a terrible imbalance. Additionally, efficiency is reduced from what was achieved when using a single TPS40170. These results are consistent in simulation and in-lab testing.

With this information in mind, can anyone address the two questions I asked above?

  • +1
    TI__Guru 53312 points

    Hi Nicolas,

    For voltage-mode PWM controllers, the LM5145 is preferred to the TPS40170, as it provides better performance and is more cost effective.

    However, current-mode control is much easier to implement in a 2-phase design, as it provides cycle-by-cycle current sharing. Take a look at the LM5148 (80V max input) or LM25148 (42V max input) current-mode controllers. The SYNC functionality for this controller family is purposely designed to accommodate 2-phase operation, whereas the SYNC feature of the TPS40170 and LM5145 voltage-mode controllers is just to operate 180° out-of-phase interleaving for input ripple current cancellation and beat frequency reduction.

    Regards,

    Tim

  • 0 in reply to Timothy Hegarty
    Prodigy 10 points

    Thanks Tim,

    That helped a lot. I think I still have a lot to learn about Voltage Mode Control vs Current Mode Control but I was thinking that connecting the two TPS40170 outputs together must be messing with the Voltage Mode Feedback. I will look into the LM5148 and try to understand why CMC allows multiphase where VMC doesn't.

    Just to clarify, I think you're saying the SYNC pin on the TPS40170 wasn't really intended to allow multi-phase operation and connecting two outputs isn't really supported. Instead, it is there simply to reduce the input current ripple by offsetting timings while driving two different outputs. Is that right?

    Thanks,

    Nicholas

  • +1 in reply to Nicholas Rodriguez
    TI__Guru 53312 points

    Yes, that's correct. Voltage-mode control is not really suitable for 2-phase operation, as the inductor current is not directly controlled. This is unlike current-mode control where the COMP voltage (voltage-loop error amplifier output) sets the current command for the inductor current. So by connecting the two COMP pins, each inductor has the same current both steady state and during a transient, i.e. natural current sharing.

    And yes, your description of the main function of SYNC to phase align the two power stages by 180° is correct. This is mainly to improve the input-side behavior in terms of EMI and noise.

    Take a look also at the LM5143 and LM25143 controller family, which are configurable for dual outputs or one 2/3/4-phase interleaved single output.

    Regards,

    Tim