• State Resolved
  • Locked Locked
  • Replies 6 replies
  • Answers 2 answers
  • Subscribers 63 subscribers
  • Views 517 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.

Original question:

TPS22810: Using a mechanical slide switch to control the Enable Line

TPS2121: Can this Power MUX or another in the TI portfolio meet my use case?

Charles McClelland
Prodigy 250 points
Part Number: TPS2121

I am looking for a device that can do the following:

Two inputs - single output

Input A - Lower voltage - a LiPO battery can be from 2.7-4.2V - Primary source

Input B - Higher voltage - DC in from 5-9V - Secondary source

If Input A is present - MUX selects this input 

In Input A is absent - MUX selects secondary source

Limitations / conditions:

Input B, the MUX will enforce a 5.5V overvoltage limit.

No reverse current allowed

This is a low current application - ideally would have the ability to set a current limit or at least short protection.

Does something exist that can do this?

Thanks,

Chip

  • 0
    TI__Mastermind 32650 points

    Hey Chip,

    I do not think one device in the PowerMUX portfolio can meet all these requirements but TPS2121 can meet most of them.The lowest current limit that can be set is about 1.5A for this device. The minimum input voltage is 2.8V. Over voltage lock out can be user configured to 5.5V. It will block reverse current and priority can be set for input selection.

    Unfortunately, your secondary source input voltage and the need of an OV at 5.5V limits the use of TPS211xA devices.

    Let me know if you have further questions.

  • 0 in reply to Shreyas Dmello
    Prodigy 250 points

    Shreyas, 

    You have been quite helpful.  I am very happy you mentioned the TI power Mux offerings to me as I can see many applications in the IOT space.  That said, there is always the first implementation, and I would greatly appreciate it if you could validate my approach.  I always want to reduce risk before getting boards made and if you could check my work here, it would increase my confidence.

    I am using the TPS2121 and will be referring to this data sheet: https://www.ti.com/lit/ds/symlink/tps2121.pdf

    My use case again:

    - Primary supply - LiPO battery (3.7V to 4.2V) 

    - Secondary Supply - A 5V DC @1A   - OR - it could be a solar panel but in this case there will always be a LiPO battery

    - If there is a LiPO battery, I want to use that one until it gets to 3.7V

    - on the Secondary supply, I want to have a 5.5V over voltage limit.

    So, here is my approach.  I could not use the most straightforward solution which is the Manual Switchover as there are some system states where I would not have the 3.3V to connect to CP2 as pictured in Figure 20.  So, I selected the  Automatic Switchover with Priority (XCOMP) circuit , The schematic of which is Figure 28 on page 25 of the data sheet.

    In my example, I made the following changes:

    - I connected the LiPO to IN1 and the DC input to IN2

    - I don't need over voltage protection on IN1 so, I connected OV1 to ground

    - I want to limit the voltage on IN2 to 5.5V so, I put a voltage divider of IN2 -> 20.47k ->  OV2 -> 5k -> Ground

    - If the LiPO goes below 3.6V, I want to switch so, I put a voltage divider of IN1 -> 12k -> PR1 -> 5k -> Ground

    - If the DC-IN goes below 4.5V, then we are done so, IN2 -> 16.2k -> PR2 -> 5k -> Ground

    So, My hope is that when a LiPO battery is attached and it is over 3.6V, the Mux will connect IN1 to OUT.  If the battery falls below that voltage or is not connected, OUT will be connected to IN2 as long as it is between 4.5 and 5.5V .

    My circuit is below. 

    Am I on the right track here?  Any advice is appreciated.  

    Thank you,

    Chip

  • 0 in reply to Charles McClelland
    TI__Mastermind 32650 points

    Hi Chip,

    I took a look at your schematic and here are my observations:

    1. You OV protection is set to around 5.38V nominally. Is this okay? Do note that there is some variation on the V_ref.

    2. Your PR1 voltage is set to 1.08V when IN1 is 3.7V. I would suggest to set PR1=1.10V when IN1 is 3.7V. 

    3. In XCOMP mode, CP2 is compared with PR1. As of now this means that IN1 will switch to IN2 at IN1=4V. This is because the voltage on CP2 is 1.17V when IN2=5V. If you want to use IN1 to a lower voltage, this divider needs to be changed.

    4. Your SS cap is quite large but do note that in XCOMP mode there is no soft start when switching between two valid inputs. This could result in inrush currents and thus to maintain the input rail, I would add input capacitance on both rails.

    5. ST is a useful open drain status pin and for further troubleshooting. I would recommend using the ST pin and not grounding it.

    6. The MUX will not turn off any input as long as it is valid. When both PR1 and CP2 fall below V_ref, there will be no prioritizing inputs. The device will pass through the highest voltage now as it works in VCOMP mode. The input is passed through until it either rises over OV or falls under UV. 

    7. Because of the inrush current during switchover the device could bounce between the inputs as the sudden loading droops the input rail. As such, I think adding hysteresis to the switchover will be useful. You can do this with the help of the ST pin. Here is a TI Design that provides the details: http://www.ti.com/lit/ug/tidue50/tidue50.pdf

    Thank you for choosing TI Power Switching in your designs. 

  • 0 in reply to Shreyas Dmello
    Prodigy 250 points

    Shreyas, 

    Thank you again for your help and advice.  I have looked at your comments and have a few follow ups.

    My values for the resistors are slightly different than yours.  I used this formula to calculate the Top Resistors and kept the bottom Resistor at 5k ohms and Vref at 1.06V.  Is this correct? 

    Rtop = (Vmax / Vref) * Rbottom - Rbottom 

    1) The Real Time Clock I am using can operate up to 5.5V (and Vmax is 6.5V) so, I will use a 20.9k ohm resistor

    2) On the PR1, by my calculations setting Vref to 1.10 at 3.6V would require a resistor value of 11.4k.  

    3) You are correct.  If I change the CP2 resistor value to 20.9k, then IN2 would need to be 5.5V to reach Vref of 1.06 which is the over voltage value.

    4) This circuit is only powering a Real Time Clock and a few passive components.  Estimated current draw is only 70uA.  Do I really need more caps?  

    5) OK, I can connect the Status pin to an input on my microcontroller.  I assume I can use the GPIO pin's internal pull-up right?

    6) If neither IN1 nor IN2 is below Vref, they circuit will be non-functional.  I am OK with this.

    7) OK, I see that adding a 1M ohm resistor between ST and PR1 could reduce hysteresis.  But, does this preclude me from using the ST pin for troubleshooting as you suggested in point #5 above?

    Here is my revised circuit.  Please take a look - I think we are close.

  • 0 in reply to Charles McClelland
    TI__Mastermind 32650 points

    Hi Chip,

    From your second post I thought your load was in the A range or 100s of mA range not uA! Lets take a step back here! 

    70uA is quite small so I think hysteresis may not be necessary. I also don't think extra caps may be necessary. On the other hand, 70uA is a very low load current. TPS2121 can leak more current than your load requires! Are you certain you want to use TPS2121 for this application?  TPS2121 was built for much higher power applications! I'm glad we caught this now!

    I think it would benefit you to take a look at TI's Signal Switch portfolio over the Power Switch portfolio: http://www.ti.com/switches-multiplexers/analog/overview.html I cannot help much with these devices but feel free to post an E2E question about those devices and the responsible engineer should follow up with you.

  • 0 in reply to Shreyas Dmello
    Prodigy 250 points

    Shreyas, 

    Good catch, in that post I was referring to the supply side not the required current on OUT.  I will look at the switch line but you have opened my eyes to a new tool that I will certainly find applications for going forward - this learning will not go to waste.   Thank you for all your help.  

    Chip