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.

TPS65400EVM-678: Using EVM without USB Connection

Part Number: TPS65400EVM-678
Other Parts Discussed in Thread: USB2ANY, TPS65400

Hello,

I have recently purchased a TPS65400EVM board for use in powering a radar system. I would like to be able to program the board once and then have it operate on startup without connecting the board to the computer/USB2ANY plug. However, I have had trouble getting this to work - the board seems to only turn on the voltage outputs once I have connected to it via the GUI/USB. Is there any way to configure the board to operate without being programmed/connected to USB on startup?

Thanks,

Russell

  • Hi Russell

    When you first program the EVM, please click on the STORE_DEFAULT_ALL so that you can have it operate on startup without connecting the board to the computer. I cannot attach a picture here. If you cannot find STORE_DEFAULT_ALL, please let me know. Thanks.

    BR

    Ruby

  • Ruby,

    Thanks for the response. I have been using this button to save the registers, and it has not solved the problem. It does seem to save the registers into non-volatile memory, since the board retains its configuration from startup to startup; however, it does not seem that the Vout supplies are activated/enabled until I connect to the board from the PI-Commander on my computer.

    For clarity, here is what I have tried and observed:

    1. Immediately when I turn the board on by applying 12V to Vin, without being connected to the computer (note: I leave the USB2ANY plugged into the board but unplug the USB - could this be my issue? I will try this tomorrow to see), the current going to the supply from a benchtop supply I have used for testing is about 10 mA, and testing each of the voltage supply outputs reads 0 V on a multi-meter.
    2. As soon as I connect to the board from PI-Commander, the current increases to 50-70 mA (depending on how I had the board configured at the last startup), and the voltage supplies show the correct voltages according to a multimeter. The voltage supplies are not connected to any load.
    3. I can modify the configuration of the board using the GUI and it responds as expected given the commands it is issued, with the voltage outputs changing as I change the value of Vref. When I press the STORE_DEFAULT_ALL button, the board soft resets and when it restarts, it holds the same configuration. 
    4. Once the board is disconnected and turned off, when I reapply the 12V to Vin to turn on the board, it goes back to pulling only 10 mA and the voltage supplies each give 0 V when tested with a multimeter.

    I have played around with some of the configuration settings in the GUI and examined the schematic to see if there is a jumper or something that needs to be placed that I am missing, but so far have not been able to solve the problem.

    I hope this clarifies my issue. Please let me know if I can provide any more information or be more clear. Thank you again for your help!

    -Russell

  • Hi Russell

    Thanks for your clear description! When you restart the EVM after first program, please make sure that the USB2ANY is not connected to the EVM. Let me know if it still does not work.

    BR

    Ruby

  • Hi Ruby,

    I think I have discovered the issue but do not know how to fix it. Basically, each of the switch enable pins (EN1, EN2, EN3, EN4) needs to be biased high in order for the supplies to turn on. I was under the assumption that if left floating, an internal pull-up resistor kept these supplies on. However, what I have observed is that on startup, I must first tie them to GND and then let them float, and then the supply starts running. Is there something wrong that I should need to do this?

    -Russell

  • Ruby,

    Just wanted to give you an update. I think that the issue is originating with some modifications I have made to the board. We are trying to get an output voltage of 10V on each supply (which falls within the range of 0.9Vin, since Vin=12V). To do so, I replaced the onboard resistors forming the reference voltage dividers with 18k and 2.2k resistors and set Vref to 1.08 on each supply, which gives about 10V. This is when the problem starts. I have tested the supply with Vref set to 0.6 V, and while this obviously doesn't give the voltage I need, it does maintain the settings when I reboot the supply without connection to the computer. Similarly, I tested this with an unmodified supply (all the resistors were the same as shipped from the factory) and did not have problems with it either.

    It seems peculiar to me that by replacing the resistors and changing to voltage supply level to a certain point, the supply no longer operates unless I program it with the computer. Do you know why this might be happening?

    Thanks,

    Russell

  • Hi Russell

    I have no idea how to insert a picture on E2E after the upgrade.

    The EN pins can be left floating when the EVM powered on. Did you have a try when all the parameters and settings of the EVM were not changed to see whether EN pins needed to connect to GND first and then be left floating? Since in our lab, when we use the EVM, we do not have to connect EN pins to GND first and then let them floating to let the EVM work. We just let EN pins floating, the EVM performs well once it is powered on.

    There is one page which is the simplified circuit of TPS65400 in GUI. On the page, you may see the resistive voltage divider for each channel. Would you please change the resistors to keep the same as you use on board?

    After that, change the voltage reference and other settings as you want (when do all the settings, please make sure that your EVM is powered on).

    After all settings are ready, click on apply changes immediately and STORE_DEFAULT_ALL (make sure that EVM is powered on). 

    And finally, let the USB unconnected to EVM. Power on the EVM to see whether it works. 

    If it still does not work, let me know. And I will have a try in our lab to find what happens. Many thanks!

    BR

    Ruby

  • Ruby,

    The board works fine, without needing to drive the EN pins to GND (i.e. I can leave them floating), when either:

    • The Vref value is set to 0.6, even if the resistors have been changed 
    • I use any Vref value (I have tried up to 1.09), with the default resistors

    The problem only arises specifically with the modified resistors, and when Vref is sufficiently increased.

    I have tried changing the resistor values in the GUI, but as far as I know this is only a convenience tool for calculating the expected output voltage. It did not fix the problem.

    If you are able to test it and see if you can duplicate the issue, that would be most helpful. For reference:

    • I have replaced R3, R8, R13, and R18 with 18kohm resistors
    • I have replaced R5, R10, R15, and R21 with 2.2kohm resistors
    • I have set Vref of each supply to 1.08V (+-0.01V to get as close to 10V as possible)
    • My Vin is 12 V

    Let me know if you need any more information. Thank you very much for the help!

    -Russell

  • Hi Russell

    Got that. I will have a try in our lab.

    BR

    Ruby

  • Hi, Russel 

    Some comments: 

    1. Is this issue related with VIN voltage? if increase Vin to 15V, still have this issue? 

    2. Could you take a picture of the EVM setting? 

    3. You are using PI commander or Fusion to program IC?  

  • Hi Zhao,

    I think it is related to Vin. I forgot to mention in my previous post - I did try with 15V to Vin, and did not experience the issue. Unfortunately, this will not work for us since we are confined to using a 12V supply for now, and have a required 10V output.

    I am using PI Commander.

     Attached is a photo of the board. Let me know if that is not informative enough, I can take another photo if necessary.

    -Russell

  • Hi Russell

    Thanks for the information. We are trying to duplicate the issue.

    BR

    Ruby

  • Hi Russell

    I did the experiment with the EVM in our lab, but I could not duplicate the issue. Here is what I have tried:

    1. connect USB2ANY with both PC and EVM, open GUI, and power on the EVM (the load current is 0A).

    2.check each channel's initial output, and every channel works well.

    3.change R18 to 17.8kohm (I have no 18kohm in the lab), and R21 to 2.2kohm. Then Reconnect EVM with PC using USB2ANY and power up to check channel 4's output (Here reference is 0.8V). 

    4.change voltage reference for CH4 to 1.08V, and find the output changes to 10V. Then Store the registers (Here sometimes I cannot save the registers at once, so I click on the STORE_DEFAULT_ALL twice to make sure that the registers have been changed ). After that, I disconnect the EVM with the USB2ANY and repower up. The output of CH4 is still 10V as I want.

    5.change  R3, R8, and R13 to 17.8kohm resistors, R5, R10 and R15 with 2.2kohm resistors. Then Reconnect EVM with PC using USB2ANY and power up to check CH1, 2 and 3's outputs (Here reference is 0.8V).

    6.change voltage reference for CH1, 2, and 3 to 1.08V, and find the output changes to 10V. Then Store the registers. After that, disconnect the EVM with the USB2ANY and repower up. The output of CH1, 2 and 2 is still 10V as I want.

    So would you please tell me when you power up the EVM, what is the load current of each channel ? Thanks!

    BR

    Ruby