• State Resolved
  • Locked Locked
  • Replies 7 replies
  • Answers 1 answer
  • Subscribers 62 subscribers
  • Views 373 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.

TPS7A47: Fast Charge function

John Kenny
Prodigy 40 points
Part Number: TPS7A47


I noticed in the Functional Block Diagram on the datasheet for this part a block called "Fast Charge" but no text about this function & I haven't found any in TI documentation.

I noticed in using this part that it retains it's current output right up to the set voltage which is exactly what I want in my application. Essentially, I'm using it to energise ultracapacitors & keep them at a set voltage even when 2 - 3A is being drawn from these ultracaps. So I need a fast reacting, low noise, voltage regulator to provide a very stable output voltage 

A number of questions, please:

- This behaviour is unlike most other voltage regulators which gradually drop current as they approach the output voltage setting. Is this behaviour as a result of bipolar design of the regulator?

- is there another voltage regulator with higher current output >1A which has this behaviour? 

- I'm setting the output voltage using the ANY-OUT pcb setting - does setting voltage using external resistors result in the same quick charge behaviour?

- is paralleling TPS7A4700 ICs feasible to provide higher current output?

- I also need to be able to set output voltage on-the-fly, using a potentiometer - is this feasible in a parallel configuration? 

  • 0
    TI__Mastermind 23361 points

    Hi John, 

    As you can see in the block diagram, the 100KOhms resistor and the external NR cap could create a big RC constant that will delay the startup. The fast charge block is to bypass the RC and quickly charge the node to the set voltage in order to speed up the startup. This fast charge circuit will be disabled after the device finish startup. 

    Please see my comments under your questions. 

    - This behaviour is unlike most other voltage regulators which gradually drop current as they approach the output voltage setting. Is this behaviour as a result of bipolar design of the regulator?

    It really depends on the load of the LDO, as the capacitance load will need peak current need at the beginning and drops charging current when it is being charged up. The LDO itself has a circuit to limit the maximum current it can output and once this is engaged, you would observe constant charging current if the capacitance load is huge. Most of the LDO behave this way.

    - is there another voltage regulator with higher current output >1A which has this behaviour? 

    You may consider TPS7A83/84/85 families if you don't need high voltage. 

    - I'm setting the output voltage using the ANY-OUT pcb setting - does setting voltage using external resistors result in the same quick charge behaviour?

    The ANY-OUT setting only change the resistor dividers internally and it will not change any function or behavior of the device. 

    - is paralleling TPS7A4700 ICs feasible to provide higher current output?

    If TPS7A8X family works for your application, I would suggest going with a single LDO. When paralleling multiple LDO, ballast resistors will be needed. The accuracy of the LDOs needs to be very tight, otherwise, it will result in the huge voltage drop across ballast resistors and this defeats the purpose of having two LDOs.  

    This thread will help you understand the limitation I talked about. 

    e2e.ti.com/.../832277

    - I also need to be able to set output voltage on-the-fly, using a potentiometer - is this feasible in a parallel configuration? 

    You could use a potentiometer to adjust the output voltage on the fly for most reference-voltage-based LDO. For parallel, please see answer in the previous question. 

    Regards, 
    Jason Song

  • 0 in reply to Jason Song
    Prodigy 40 points

    Thanks Jason

    I understand the fast charge now

    You are probably correct, the LDO is running at max current limit which is not great long term - I will look into that

    Can't use the TPS7A8X as I need the higher voltages than they provide

    Had a read of that document on paralleling TPS7A8X with ballast resistors - the TPS7A4700 seem to produce very well matched Vouts when using ANY-OUT.

    I presume low value ballast resistor needs to be high current?   

  • 0 in reply to John Kenny
    TI__Mastermind 23361 points

    Hi John, 

    Got you, I figured you probably picked TPS7A47 because of high voltage.

    TPS7A47 can be used for parallel applications, like you said, using ANY-OUT could generate very well matched Vouts; it can be further improved when connecting the two NR pins together. The reason I did not recommend a parallel application using TPS7A47 is that the accuracy specification of the output is 2.5% across conditions and the app notes TI currently have related to parallel application needs this accuracy numbers to calculate the Rbllast. But the value from 2.5% accuracy will result in too big a Rblast that is not practical to implement in a real application due to the high current voltage drop through the Rblast. 

    However, you can still run your own tests to figure out the output difference when connecting the two NR together. Once you have confirmed this error term, you could use this value to calculate the Rblast. This Rblast will be much smaller compared to the one calculated with the 2.5% accuracy spec. 

    Does this answer your questions? 

    Regards, 
    Jason Song

  • 0 in reply to Jason Song
    Prodigy 40 points

    Thanks Jason

    It is probably obvious but just checking - the Rblast needs to be 2Amp or greater, or I could limit the max current situation from the TPS by using a 1A Rblast?

    And final question - connecting NR (noise reduction) pins between TPS - should this have a capacitor to ground to lower noise, as per the single TPS config?

  • 0 in reply to John Kenny
    TI__Mastermind 23361 points

    Hi John, 

    Each LDO will need to have one Rblast, so when you do the calculation, the voltage drop across Rblast is up to 1A. 

    For the NR, when you connecting the two NR pins together, you still need to have the NR capacitor filter on the pins to filter noise. The purpose of connecting the two NR is to minimize the voltage difference at the NR that will help to reduce the offsets on the two outputs. 

    Regards, 
    Jason

  • 0 in reply to John Kenny
    Prodigy 40 points

    Ah, I see that Rblast is actually a PCB trace of correct length & width to give resistance & current capability needed

    So only question left is - should there be a cap to ground from NR pin - it doesn't show one in the document on paralleling? 

  • 0 in reply to John Kenny
    TI__Mastermind 23361 points

    Yes, you may use traces to achieve the needed resistance or you may also consider put real resistor. You should have a cap to ground from NR, and this is for noise filtering. 

    Regards, 
    Jason