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.

TPS61032: TPS61032 PCB Layout

Part Number: TPS61032
Other Parts Discussed in Thread: TPS61030EVM-029, TPS61023, TPS61030

Hi,
With reference to the Case CS0634034 and CS0558572 here is my updated design. I request you to have a look at the layout. I tried to match it with TPS61030EVM-029 board and place the components closer. I chose the same recommended inductors/capacitors and placed them near to the IC. Also, the IC tracks have a bit more width (normal are of 10 while tracks around this IC has 15). I have attached the overall design circuit and PCB. While I have also attached only the booster IC TPS61032 section's layout and circuit.
Please have a look!

Components: Tantalum Capacitors - Solid SMD 100uF 16volts 20% C case Conformal www.mouser.de/.../594D107X0016C2T Multilayer Ceramic Capacitors MLCC - SMD/SMT 1206 10uF 6.3volts X5R 20% www.mouser.de/.../C3216X5R0J106M Multilayer Ceramic Capacitors MLCC - SMD/SMT 0805 2.2uF 10volts X5R 20% www.mouser.de/.../C2012X5R1A225M Fixed Inductors 7uH 4.5A www.mouser.de/.../CDRH104RNP-7R0NC

PCB LayoutPart of the PCB using TPS61032TPS61032 Layout Part OnlySchematic_escooter_sajjad - rev5_2021-08-12.pdf
Sorry for repeated cases.

Thanks in anticipations.

Regards,
Sajjad Hussain, TECO, KIT Germany

  • Hi Sajjad,

    Thanks for contact. Next time, please share the schematic and PCB layout with me privately, through E2E chat. 

    For the case number, I am not able to find it. Would you mind sharing the link here directly, if the background is important?

    I assume you want me to review the schematic and PCB layout for TPS61032 on your board, right?

    Would you mind letting me know below information? Thank you very much!

    • Input voltage (VBUS2)
    • Maximum output current 

    -Wenhao

  • Hi, Thanks for the response.

    Sorry the previous case link are:

    https://ticsc.service-now.com/csm?id=csm_ticket&table=sn_customerservice_case&sys_id=6060d9f81bb47410ad92ba63164bcb6a&view=csp
    https://ticsc.service-now.com/csm?id=csm_ticket&table=sn_customerservice_case&sys_id=f01551731bb5b014ad92ba63164bcb93&view=csp

    Yes, I want to review the PCB layout basically. As I had some problems (in previous cases).

    Input voltage (VBUS2)

    VUSB2 is equal to VUSB is 5V from USB.

    Maximum output current 

    The maximum output current would be 1.A.

    1. The PCB and schematics until now working on, are as below.

    Schematic: Schematic_escooter_sajjad - rev4_2021-08-15.pdf

    PCB: PCB_PCB_PNR_R12_2021-08-15.pdf

    Power Block Diagram:

    2. The schematic and PCB intended to be manufactured. Here I basically revised the TPS61032 layout.

    Schematic_escooter_sajjad - rev5 copy_2021-08-15.pdf

    PCB_PCB_PNR_R13 copy_2021-08-15.pdf

    Basically, I have a MCP73871 charging management IC which takes input from USB or from a charger. It is used to charge the 3.7 LiPo battery and then drives the TPS61032 booster IC to produce a 5V for remaining circuit. The input to booster IC is either (3.7 to 4.2V)  when MCP is drived from the battery or 5V when MCP is derived from USB/Charger.

    What will happen to TPS61032 when battery is overdischarged and it produce voltage less than 3V? I don't know, whether I need a battery protection IC to disconnect the battery in case of overcharging or overdischarging and MCP does not produce any output and therefore TPS is not derived anymore and remains safe. I also think that, now my TPS61032 is being damaged, may be, due to the overdischarging of the battery and battery has very low voltage and booster IC tries to expect more current from the battery to output 5V and in this process it is damaged.

    Sorry for the unconvenience. This is my first experience with the charging circuitry.

    Thanks in anticipation and waiting for the response.

    Thanks and Regards

    Sajjad Hussain

  • Hi Sajjad,

    May I firstly check with your working spec. The booster IC's input is either (3.7-4.2V) or 5V and output is 5V, so we need pass through function (5V to 5V). But TPS6103x doesn't own this function so cannot generate 5V when input is 5V.

    Because I think this design has already layout several versions, would you prefer to change another IC or let the system doesn't work at 5V to 5V? If another chip is preferred, TPS61023 with pass through is recommended.

    BTW the previous case link don't exist, would you mind re-sharing the content in this threat directly, or your latest reply has included all important contents.

    Please kindly let me know your updated decision and we can move on reviewing layout or re-designing based on other chips.

    Best Regards,

    Eric Yue

  • Hi Eric,

    thanks for the response. 

    The booster IC's input is either (3.7-4.2V) or 5V and output is 5V, so we need pass through function (5V to 5V).

    The datasheet of TPS6103x says that: Input Voltage Range: 1.8-V to 5.5-V

    and also I am getting 5V when there is 5V input to the board. I have problems operating in TPS61030 discussed in first case and then switched to TPS61032 and has been working properly. but during thorough testing of the board i found that after 2-3 days the TPS61032 is being damaged. I operated it from 5V charger for few ours while battery is attached (charging mood). Then I unplug the charger and let the board operate from the battery. But it fails after some time. When it fails, It was operated from a 3.7V during these tests. I keep monitoring the battery voltage randomly, but missed to check the battery voltage when TPS61032 is damaged. My doubt is that may be battery voltage was too low and battery can't provide sufficient current for TPS61032 and it is being damaged (this is my guess).

    That is why I am considering to have a battery protection IC to disconnect the battery for the MCP73871 IC when battery voltage is overdischarged.

    Do you think that this may also be an issue?

    Considering the above, still do you think that we should replace TPS61032 with TPS61023. In my case it hase been working for 4 days when directly operated from 5V charging input.

    Previous case:

    https://ticsc.service-now.com/csm?id=csm_ticket&table=sn_customerservice_case&sys_id=6060d9f81bb47410ad92ba63164bcb6a&view=csp

    https://ticsc.service-now.com/csm?id=ticket&table=sn_customerservice_case&sys_id=5be6650f1b3d70d0ad92ba63164bcb6d&view=sp

    Thanks and waiting for the response.

    BR, Sajjad

  • A similar to my case, the following post also shows that it can work with +5V input supply.

    https://electronics.stackexchange.com/questions/94518/powering-5v-device-with-usb-and-or-battery

    Regards

  • Hi Sajjad,

    Figure 7 in datasheet doesn't show Vout. TPS6103x is a boost chip and typically Vout should be higher than Vin. It may output 5V when Vin=5V, but the ripple can be large. Because I don't have EVM currently, could you further check Vout ripple, inductor current ripple and SW pin waveforms? Large ripple may cause damage to the system.

    TPS6103x has UVLO voltage is 1.5V and 4A peak current limit. Through your description, will input voltage lower than 1.5V? I think it's better to test Vin voltage and inductor current waveform to confirm.

    Best Regards,

    Eric Yue 

  • HI

    i will check the voltage ripples using microscope. but how to check inductor current ripple?

    I have checked the IC using Adafruit Evaluation board and it has been working since 5days. i.e, it gives 5v when Vin=5v.

    Regards,

    Sajjad

  • Hi,

    I am summarizing again.

    I am using MCP73871 charging IC to charge 3.7 Li-Ion battery. This IC can take either input from charger or from battery and generate the output accordingly. TPS61032 is used to boost the output from MCP73871 to 5v to drive the remaining system. The system block diagram is shown below:

    • MCP73871 gives 5V when only charger/USB is connected MCP73871 gives
    • 5V when both charger/USB and battery is connected MCP73871 gives
    • battery voltage when only battery is connected

    My questions are:

    1. do we still need a battery protection IC between battery and MCP73871 to prevent overdischarge?
    2. If we need, can anybody suggest any?
    3. what if there is no battery prtotection and battery is around <1.8v and charger is not connected, can TPS61032 still provide 2A or so?
    4. I replace the booster IC to TPS61023. still I need any battery protection ic to make sure that TPS61023 has sufficient input and battery is not drained too low.

    Thanks and Sorry for repititions.

    Regards

  • FYI:

    osciloscope shows the following waveform:

    1. 5V input when not connected to the board

    2. 5V input when connected to the board

    3. 5V output of the TPS6103x evaluation board

    4. accross the SW input

    .

    With Kind Regards,

    Sajjad

  • Hi Sajjad,

    1. I think we need to find out what causes TPS61032 damaged. Is the purpose you mentioned we need the battery protection IC to protect TPS61032? Typically, the chip will not damage if Vin is very low. I am thinking large ripple may cause this problem. 

    2. If we decide to add battery protection IC, I can add corresponding colleague to help.

    3. If Vin decrease to trigger UVLO (1.5V), the chip will shutdown and the switch is turned off. Between 1.5V-1.8V, the chip attempts to operate, but the specifications are not ensured.

    4. If the purpose is to protect TPS61023, I think we don't need it as well. do you have any other concern to use battery protection? e.g. protect the battery.

  • Hi,

    1. Yes, I was only concerned why TPS61032 is being damaged. I was thinking to add protection IC to protect it. But I think we don't need it for TPS61032 as it will be shut down when voltage is low.

    2. May be, it is useful for the battary itself.

    3. I tested it, it will be shutdown when vin is less than 1.7V

    4. I was only guessing to protect the TPS61032, but if the reason of damage is ripple then ofcourse we don't need it.

    So, how to decrease the ripple? or should I replace the TPS61032 to TPS61023? But TPS61023 has less output current I need atleast 1.5A output current, do we have any other with pass-through feature?

    Thanks and Regards

  • Hi Sajjad,

    You can float one pad of the inductor and weld a wire, then use current meter through the wire to test inductor current.

    1. What is the decreasing peak in figure 2 of input when connected to the board? Is this Vin or SW? It is strange to have this peak at Vin.

    2. Through output ripple, the ripple is about 100mV-200mV, typically ripple will not be such large in PWM. I think the chip is working in PSM mode, inductor current can further confirm it.

    In PSM mode, the ripple is large and the loop doesn't work. We have not used this chip in this spec and may cause damaged.

  • OK, I will measure the current for inductor today.

    1. This is the waveform of the VIN (5V) and the input to the evaulation board.

    Another important point: The TPS61032 is not being damaged on the evaluation board (tested for 5days(, but it is being damaged on my board (after 2-3 days). therefore I conculded that may be I don't have a very good layout on my PCB and also I don't have an inductor of very good specification. I used 2.5A specs inductor while on the EVaulation board it is of 4.5A.
    That is why I am concerned about the layout/components also. May if we have a good layout/components, the TPS61032 life is increased. I don't know! 

    We have to finalize the reason.

    Thanks and Regards

  • Hi Sajjad,

    Please test the inductor current to further confirm it. I think TPS61032 cannot generate 5V to 5V normally. 

    Because I think this design has already have several version tested boards, can you change the IC yet? If can, I can recommend change other IC for safety.

  • Hi,

    Sorry I can't test the inductor current until now, as I could not manage to have a current probe for the oscilloscope. I will get back to you as soon as I get it.

    But I used different input voltage i.e., 3V to 5V and see the output voltage on oscilloscope which is 5V but the ripple varies from 167mv to 210mv.

    Going from TPS61032 to TPS61023 

    I have done revision for the design and layout of TPS61023 for 5V output:

    1. Can you please check The schematic:

    2. Can this ciruit with TPS61023 will provide 1.5A (maximum) but 1A in normal conditions? 

    3. Can you please check, The layout for that.

    4. Please remember that the components are as it were used in the TPS61023EVM-052 (https://www.mouser.de/ProductDetail/Texas-Instruments/TPS61023EVM-052?qs=BJlw7L4Cy78hIEUz%252Bz9fXw%3D%3D)

    5. The complete schematic is:

    Schematic_escooter_sajjad - rev6_2021-08-19.pdf

    6. The complete PCB layout is: 

    PCB_PCB_PNR_R14_2021-08-19.pdf

    Thanks and Best Regards,

    Sajjad

  • Hi Sajjad,

    When input is 3-5V, output ripple should not be so large, you can directly measure C16 ripple through minimum gnd ring to test.

    I think you have two choices now.

    1. Keep TPS6103x in use. From 5V to 5V,  TPS6103x will have large ripple and typically we will not use it when Vin ≥ Vout, but it can generate 5V and will not damage easily because of large ripple I think. The current limit will protect the chip. You can maintain TPS6103x if not convenient to change new chip, but be aware of large ripple. I think layout can be optimized to try to avoid the damage. Below is my suggestion.

    The schematic is okay.

    For layout, ground route line of Cin should be wider and place polygon and add more vias in this area is recommended.

    The ground pins of TPS6103x is recommended to place polygon and add enough vias too.

    Cout loop is very important for boost, so minimum loop area from vout-Cout-gnd. place polygon (wide route) and add more vias in this gnd area. Make sure GND area on second layer is entire.

     

    The signal loop (FB loop) is recommended to separate from power loop. The R30 GND pad can directly connected (polygoned) to chip's GND pin (not PGND).

     

    I cannot see second layer, but I think it is separated to 3 parts: Vin, Vout, GND. Please make sure GND part is complete and entire especially Vout loop and under the chip and SW.

    2. Change to TPS61023. The current limit is 3.7A so it can output 1.5A above Vin=3V. The schematic is okay as well, R26 can be omitted if no need to test loop stability (bode plot). 

    Cout is recommended to be placed similar to EVM and Layout Example from datasheet. Loop area of Vout-Cout-GND should be minimum. 

    ground route line of Cin should be wider, place polygon and add more vias in this area is recommended as well.

    GND area (polygon) is also important.

    Best Regards,

    Eric Yue

  • Hi,

    Thanks for the detailed answers.

    I have only two layers. The top is routing layer, the bottom is the ground layer only.

    1. TPS61032 Layout Revised

    2. TPS61023 Layout Revised

    Is that improved?

    Best Regards,

    Sajjad

  • Hi Sajjad,

    1. TPS61032 Layout Revised

    I think for most pcb manufacturing process, vias cannot be placed on pads, it is recommended to place these vias close to these pads.

           

    Through second layer is all GND, but it is separated by different parts of other signals wires. For example, Vout-Cout-gnd loop is important, But GND of C17 and C18 will loop a large circle return to PGND of the IC, because GND in second layer is cut off by several wires. The critical GND loop should keep entire.

    C16 is the nearest Cout, it recommended place polygon (wide route) and add more vias in this gnd pad.

    R30 GND pad's via can be omitted.

    2. TPS61023 Layout Revised

    Cout and SW is recommended to place like below which is layout in datasheet and EVM. GND of Cout placed on first layer to minimum this loop

    The problem in TPS61032 is also important. For example entire GND on second layer.

    Best Regards,

    Eric Yue

  • Hi,

    Thanks alot dear!

    1. TPS61032 Layout Revised

    I removed the vias from IC pads.

    I removed the wires from 2nd layer under this sections so that ground plane seems as a whole. Place more vias.

    Polygon placed around C16.

    Via of R30 is removed.

    2. TPS61023 Layout Revised

    I also changed the placement of c16 and c25 as you suggested. Also, cleared the bottom ground layer from other nets under the area of consideration.

    Hope this is better now?

    With Kind Regards,

    Sajjad

  • Hi Eric,

    If possible, can you please recommend a battery protection ic for my design for Li-Po 3.7V battery.

    Regards,

    Sajjad

  • Hi Sajjad,

    The vias from ALL components' pads should be changed, not only the IC.

    Sorry, I am not familiar with battery protection IC, you can open a new threat if needed and corresponding colleagues can help you.

    Best Regards,

    Eric Yue

  • Hi Eric,

    Thanks for the response. 

    There is no problem in manufacturing the PCB with vias at the pads. We also had the vias on the pads of some components in our last manufactured PCB.

    If this not affects the performance of the IC, I will keep it as it, it is crucial for this IC's I will change it?

    Thanks and Regards,

    Sajjad

  • Hi, Here i have removed the vias from the pads of the components.

    1. TPS61032 Layout Revised

    2. TPS61023 Layout Revised

    Thanks and Regards,

    Sajjad

  • Hi Sajjad,

    I am not quite sure whether it will affect because we typically place vias close to pads before. I think maybe both is okay.

    Best Regards,

    Eric Yue