• State Resolved
  • Locked Locked
  • Replies 3 replies
  • Subscribers 63 subscribers
  • Views 1482 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.

BQ25306: External Power Path Alternate Options

Uzair Mohammad
Prodigy 140 points
Part Number: BQ25306
Other Parts Discussed in Thread: TPS62136, BQ24610

Hello everyone!

I've got a three part question regarding Power Paths using the BQ25306. First, a little bit about my application:

  • Uses a Lithium Ion 2S4P battery: 8.4V, 8000mAH
  • The system has 2 loads: 
    • VSYSTEM is a set of 5V and 3.3V regulators that power a microcontroller. This requires uninterrupted power, with no voltage drops that will result in microcontroller resets. Consumes a peak of 100mA in total
    • VLOAD goes to a set of TPS62136 Buck Converters. Each one powers a DC heating system that consumes an initial peak of 3A. There are two such TPS62136's, totaling 6A peak drain from the battery/power source.
  • Goal is to charge the Lithium Ion Battery at 3A, while also being able to support 6A VLOAD.
  • When the charger is plugged and unplugged from the device, a short interruption in VLOAD is acceptable, but VSYSTEM needs to remain on at all times. 

Onward to my questions. I see several methods to incorporate a Power Path, both in the spec sheet for BQ25306 and in TI's application notes: https://www.ti.com/lit/an/slua376/slua376.pdf?ts=1616431553750

First, I'll reference figure 10-8 in the BQ Spec Sheet, showing a power path application using this particular IC. I see in another TI E2E thread that the Power Path current is limited to Iinpdm, which is 3.7A. Therefore, figure 10-8 is not capable of delivering 6A to my primary load while the battery is also being charged, is that correct?

I have one possible modification of this schematic- disconnect PMID, and connect Vbus directly to VSYS. Schematic here:

https://imgur.com/a/HmEzXuJ

By bypassing Q1 and the entire BQ IC, doesn't this give me direct non-limited current to my primary load? In my application I mentioned that my microcontroller needs uninterrupted voltage, which is why it connects directly to the battery, but the rest of my system load can tolerate a quick voltage drop. Is there something that I'm missing here? I see that when a charger is plugged in, Vbus is high (9V, for example). It charges the battery through the SW pin, while mosfet Q4 is opened and the battery is not connected to the main load any longer- but the charger is. The reverse blocking diode in Q4 prevents any reverse conduction, it seems. 

The only thing I do not like about this setup is that, under regular conditions, ALL of my heavy load current is passing through Q4 from the battery. Even with a low RdsOn MOSFET, I lose some performance because of the voltage drop across the MOSFET, especially at high currents. 

I then considered doing this same power path setup with a SPDT relay. Schematic here:

https://imgur.com/a/UAwn7AA

In this case, the resting position of the relay connects VBAT to VLOAD, giving me a direct connection with zero voltage drop. Upon plugging in a charger, Vcharger triggers the coil of the relay, breaking the connection with the battery. There is a slight delay inherent to the relay, in which no current is conducted as it switches from one contact to the other. Then, it eventually connects Vcharger directly to VLOAD. Because the relay physically cannot connect both Vbat and Vcharger, there is no risk of improper conduction. Again, my load can tolerate a slight period of zero voltage (unless the TPS62136 cannot handle VIN dropping out and coming back like that), so the fact that the relay is not as fast as a MOSFET is tolerable. 

Physical space is not a concern for my application, and I can get the necessary relay at an acceptable cost. Am I missing something here? Because my load can tolerate a voltage drop it is somewhat unique compared to most applications, so I'm hoping that I can use this method, or the method I previously described. Do note that VSYSTEM is powered by the battery directly and therefore never experiences a voltage drop. Is this kind of "relay hot swapping" going to cause any kind of improper voltage or current spikes that the BQ cannot deal with? Would you recommend I place an inductor after the relay or MOSFET to help smooth any current changes? 

Thank you so much, I recognize that this is a mutli-faceted question and I deeply appreciate the help. 

  • 0
    TI__Mastermind 24075 points

    Hi,

       Your 6A load is quite high for the IC to dissipate amongst the internal FETs and is more of a limitation of the package constraints itself. I don't believe you will find an integrated charger to be able to handle this powerpath, so your idea of using external FET to connect adapter to SYS is a good idea. With these power requirements I would recommend you the BQ24610 device, as it is a charge controller and has the features you need integrated. Have you considered this device? By integrating the drivers, and not the FETs this charger IC can handle higher power like that required for your application and also accomplishes direct power path without external components, which is essentially what you are trying to do with BQ25306.

  • 0
    Prodigy 140 points

    Hi Kedar,

    Going to the BQ24610 is a more expensive IC, and our application is particularly cost sensitive. Also, the BQ24610 requires more external components (some of which are related to the Power Path, but some appear related to the function of the battery charging). If I am able to implement the BQ25306 with one of the alternate designs, it has some benefits for my application.

    Are you able to identify any electrical issues associated with the options I suggested? What about the relay, does "hot-swapping" the load pose any problems downstream for the TPS62136? 

  • +1 in reply to Uzair Mohammad
    TI__Mastermind 24075 points

    Hi Uzair,

       Yes, your relay combination should work. It is the same principle of having the relay/FET disconnecting battery to SYS when adapter present, and connecting battery to SYS when adapter is present. This should be a fine workaround and also reduce your voltage drop form BAT to SYS. The issue with relay is the delay between connecting. With a FET you can slow down the turn on to reduce transient, but this would be harder with your relay.

    The main issue people run into is that this powerpath has a voltage swing between the battery and the adapter. As you have downstream buck this is fine. For a 5V rail off of the buck converter, you should consider the corner case of no adapter, and discharged battery below 5V. If this is a possibility, then battery could not provide uniterrupted 5V rail.