• State Resolved
  • Locked Locked
  • Replies 11 replies
  • Answers 1 answer
  • Subscribers 63 subscribers
  • Views 1324 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.

BQ40Z60 PACK GND usage with boost converter drive

k1mgy
Expert 1680 points
Other Parts Discussed in Thread: BQ40Z60, TPS61175

BQ40Z60

My original design calls for a +21V DC external DC supply to drive the VAC (Charger Voltage) inputs.  According to the reference design, the return terminal is connected to PACK GND rather than system GND. 

The design has changed as we are now using a TI TPS61175 Boost Converter (from +12VDC to +21VDC).  Is it still necessary to connect the boost converter ground returns to PACK GND? 


Note: I've retained a VAC snubber circuit to avoid damaging the TPS61175, as was done in the prior design.  However, I'm tempted to connect the snubber ground return to system GND.  Would rather have a current spike on that bus instead of PACK GND... but perhaps this is also not wise?

  • 0
    Expert 1680 points
    Would TI give this question their attention please?
  • 0
    TI__Mastermind 40080 points

    Hi


    It would be better to connect the PGND of TPS61175 to the PACK GND or PGND of bq40z60. The snubber should be just close to the TPS61175 SW pin.
    Could you please share your block diagram to me so that I can judge it more clearly.
    My email address is helen-chen@ti.com


    BR,Helen

  • 0 in reply to Helen chen
    Expert 1680 points

    Thank you Helen.  I have sent you schematics and a few comments in email to your address.

    For the benefit of readers, I will be happy to post the results here.  I know how valuable these public postings are when answers and ideas are hard to get from the manufacturer.

    /m

  • 0 in reply to k1mgy
    TI__Mastermind 40080 points

    Hi Mark,

     

    The ground connection is very good with TPS 61175's PGND connected to 40Z60's PGND.

     

    For the snubber part. From the 40Z60 datasheet page35, we know that the snubber is to protect the 40Z60 from damage during the adapter hot plug in.

    Now in your design, you want to protect the TPS61175. As I don't know the spike level , so please keep it. You can detect the voltage at TPS61175's VIN pin during hot-plug in. the maximum input at VIN pin is 18V.

     

    Yes, thermal may be an issue. I don’t know what is your operation frequency for TPS61175? from your schematic, it seems that you want to make it operation at 2MHz, right? Could it be lower, because the thermal is related to the operation frequency. As you already use a 10uH inductor, so the frequency can be lower. What is the maximum output power of TPS61175? I can calculate it for you.

     

    Best regards,

    Helen

     

  • 0 in reply to Helen chen
    Expert 1680 points

    Helen,

    Thanks for calling my attention to the switching frequency.  Oops!

    There's no graph in the data sheet that shows a relationship between the TPS61175 switching frequency and efficiency.

    I've found that around 600-700KHz works well in other buck/boost converters, and wonder if this might be best here, or if you can make a recommendation. 

    The requirement is to boost +12VDC (regulated, external) to +21VDC delivering around 700mA for the "adapter pack" input to the BQ40Z60 for charging.  I figure with our external +12V supply rated at 1.5A Max, it's reasonably assured we'll get 700mA out of the boost converter at +21V without exceeding the external +12V supply's 1.5A current rating.  As a rule of thumb, I am doubling the required current drive for my expected load.  However, perhaps there's a better way to determine this?

    Yes, we'll keep the snubber.  It's good insurance.

    /mark

  • 0 in reply to k1mgy
    TI__Mastermind 40080 points

    Mark,

     

    You're welcome.

    If use 10uH inductor, fs=800k is a good selection.

    This Boost converter can well output 21V/0.7A.

     

    I send a TPS61175 calculation sheet to you, hope it will be helpful.

     

    BR,Helen

  • 0 in reply to Helen chen
    Expert 1680 points
    Helen,

    Some questions:
    Boost Efficiency: Is this the ratio of current or power across the boost circuit? I need to know the load on our 1.5A external +12V supply with a 0.7A load on the boost converter +21V output.

    What is the formula for determining the switching frequency resistor? I note the table in the data sheet, but it would be helpful to do a more direct conversion.


    My latest pass at this uses a 15uH inductor (2.9A) and a 107K resistor at the FREQ pin. I've paralleled two 22uf capacitors at the output per your suggestion, but will see if we can get by with only one. I suspect the 40Z60 won't introduce transients...
  • 0 in reply to k1mgy
    TI__Mastermind 40080 points

    Mark,

    Boost Efficiency=(Vo*Io)/(Vin*Iin), in your case, the estinated efficiency is out 88-89%. So the input current is about 1.3-1.4A. So if your source have +12V/1.5A ability, then it is just at the margin.
    Sorry, there're no formula, the graph is a fitted curve by test. Since you choose a 15uH inductor, you can let it work at 750k, Rfsw=143k
    If there's no transient, then one 22uF cap is enough. But be careful about the effective capacitance under the DC bias, generaly a 50V/22uF cap under 21V DC bias is only 10uF. If transient is not a problem, you can use 2pcs 10uF/50V cap in parallel.

    BR,

    Helen

  • 0 in reply to k1mgy
    TI__Mastermind 40080 points

    For the TPS61175 layout, please refer to the TPS61175 EVM layout.

    You’d better add some vias under the TPS61175 PGND pad, make the PGND is big enough for the thermal dissipation.

    The EVM is a four-layer board. In your design, make sure, there have 2 layers left for the PGND copper plane for good thermal performance.

  • 0 in reply to Helen chen
    Expert 1680 points
    Will do on the vias. My board is also 4 layer with plenty of plane.

    Thanks for that explanation on efficiency.

    So the only piece I'm missing is the formula which selects the frequency based on resistor value. Do you have one or must I take a guess using the table?
  • 0 in reply to k1mgy
    TI__Mastermind 40080 points

    You can select one through graph. Then test the SW waveform to deternmine if  you select a right resistor. Since you choose  a 15uH indcutor, so 750k operation frequency is OK for your application.