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.

UCD7242 indicates FLT error when starting up and there's no pwm output, why?

Other Parts Discussed in Thread: UCD9222, UCD7242

Hello!

We designed a C6678 board following the standard EVM6678 board, and all schematics of UCD9222 + UCD7242 are the same. However, we have debugging this part for almost a month. Can someone help me analyze why this error exsits?

Here are some shortcuts of our waveforms and results:

1. Reports from Fusion Digital Power Designer. UCD7242 reports FLT error, and there's no pwm output.

2. When we remove the output inductor, the duty cycle of pwm rises to be approximately 98%, and the output now is 12V with heavy noise.

3. We then add the inductor and remove all output capacitors except a 330uF one, and the output now is averagely correct but still with heavy noise. I am sure that 6678 cannot work under this voltage.

  • If you are using the same design as the TMDXEVM6678L/LE Rev 3.0 EVM (same inductors and capacitors), you can try the attached configuration file for UCD9222.

    TMDXEVM6678L_UCD9222_14Mar16_ver5.xml

    Regards,

    Zhiyuan

  • Dear Zhiyuan

    After loading your configuration file, it becomes much better.

    Rail#1 works well with a little bit ripple:

    Rail#2 reports OV Fault when starting up and then gets better as Rail#1:

    Now, I am wondering if I can achieve better quality outputs through parameter tunning? If so, what are the key parameters?

  • Under Design->Linear Compensation, you can lower the gain (K) and see if there is any improvement.

    Regards,

    Zhiyuan

  • Thanks for your advice! However, there is a strange phenomenon. When I plug in the USB2GPIO cable, the PMBUS_ALERT pin is low and the output is normal. When I remove the cable, it goes high and there's no output. What could be the reason?

  • When there is fault or warning, Alert will be pulled low. You can go to Monitor page to see what caused the alert. 

    Since the rail is off when you remove the adapter, it is possible that the rail is configured to be controlled by CONTROL pin, and the CONTROL signal is from the USB-to-GPIO adapter.  You can check ON_OFF_CONFIG setting in the GUI.

  • Hello Zhiyuan!

    Sorry to trouble you again. Since last time we used your configuration file, everytime we can only manage Rail #1 under control. But Rail #2 still outputs low-quality voltage with heavy noise as shown below (we set the "OV Response" to be "Continue without Interruption"):

    We have adjusted the linear compensation Gain (K) and it got nothing better. But only onetime, the output was fine to power the DSP and we could run a "Hello world!" demo. But it never happens again even we are using the same configuration file.

    Even we removed the DSP chip from our board, every error keeps the same.

    Finally we made another test board only had the power section of UCD9222+UCD7242 on it, and the output voltage run smoothly.

    We have stuck into this problem for two months and got so confused. So PLEASE help me! 

    Regards

    Pengcheng Wang

  • I'm thinking there are additional capacitors near the DSP that are not included in the control loop design, so the loop is not stable. 

    On the test board that only has the UCD, did you install the capacitors for the DSP? If not, can you install those capacitors and see if the loop is stable?

    Thanks,

    Zhiyuan