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.

PMP9053: UCC28910 cannot start up

Part Number: PMP9053
Other Parts Discussed in Thread: PMP9608, , PMP, UCC28910

Hi all, 

I tried to follow the power supply reference designs, PMP 9053 and PMP 9608. These two designs are very similar. PMP 9053 outputs 3.3V while PMP 9608 outputs +-3.3V. However, I found the diodes used for PIN 1 and 2 of transformers are different in these two reference design (D4 D6 in PMP 9053, D2, D3 in PMP 9608). I am using the diodes in PMP9608 design. Their reverse voltages are lower than them in PMP9053.

Right now, UCC28910D cannot start up in my circuit (it might be broken, too) because the frequency of DRAIN (PIN8 of UCC28910) is only 15Hz. I double checked the connections and orientation of the diodes. Is it possible that the problem might be caused by the diodes I selected. Does anyone help me understand the functions of these two diodes and how to select the diodes for this circuits? 

Meanwhile, I am not sure what the frequency of the voltage from DRAIN (PIN8) should be under the working condition and how this frequency is controlled?

Thanks.  

PMP 9053: 

PMP 9608: 

  • Yin Li,

    Thank you for your interest in TI. Can you post your schematic and excel calculator filled in? This will help us understand your design details.

    Frequency is chosen by selection of magnetizing inductance of your transformer and peak current setting. Please share your transformer details as well like magnetizing inductance and turns ratios.

    You can also use this resource to debug while we review:  

    Regards

    John

  • Hi John,

    Thanks for replying. The schematic I am using is from TI reference design on Smart Plug. Here is the website of this reference: 

    I am using its power supply part which converter 120VAC to +-3.3VDC but it doesn't work right now. All of components I am using are the same as them in the reference design including transformer. I attached the schematic of the power supply part here.  .  The transformer is customer designed by TI and I purchased from Wurth Electronics. I also attached the datasheet of the transformer here. XFRM_750314752.pdf

    Thanks. 

  • Hello Yin Li,

    If your power supply board is connected as the schematic diagram in your previous post, then I think the start-up problem may lie in the GND connection of the UCC28910 (U8). Your schematic shows U8 pins 1,2,3, shorted to pin 4 and to 2 resistors R36 and R37, but NOT connected to the common GND return for the primary-side circuit. This makes a high-impedance connection from C14 and for the return path of the Drain current, preventing normal start-up. The result is a repetitive attempt to start every 67 ms, which appears to be frequency of about 15Hz. This is not the designed switching frequency of the controller, but the frequency of the re-start attempts.

    I think U8-1,2,3 should connect to the network from the right side of L7 over to T1-pin5. I also think that you probably do not want to connect U8-4 (IPK) to U8-3 (GND), but that U8-4 should connect only to R37 (to match the PMP circuit).

    Please check you board connections for proper primary-side grounding.

    Regards,
    Ulrich
  • Hi Ulrich, 

    Thanks for your replying. I am sorry for the previous schematic. That schematic is not the lasted one we are using. We already correct that wrong connection. The following is the latest version we are using.  . The results I provided in the previous post were based on this schematic. In another word, after I corrected the wrong connection, UCC28910DR still cannot start up. Is it possible caused by D7 and D8 I select? 

    In addition, if I want to test the chip individually, how should I input the voltages to different PINs of the chip? 

    Thanks. 

  • Hello Yin,

    Sorry for my delayed reply. I'm glad that you found the wrong connection. It is not clear from your post, but do I understand it correctly that you powered the circuit up with the wrong connection, got no start up, found the wrong connection and fixed it, repowered the circuit and still get no start up? If so, I think the controller was probably overstressed when powered with the wrong connection and is now damaged.

    I suggest that you install a new controller at U8 and follow the start-up troubleshooting guide that John Stevens recommended above.
    I also recommend changing D8 from RS1G-13-F to US1J-13-F. This changes a 400-V, 150-ns recovery diode to a 600-V, 75-ns diode. I think the original design for this power supply was for 120Vac only and a 400-V diode is okay for that. But for an input voltage up to 265Vac (even up to 300Vac) a higher reverse rating is necessary, and the faster recovery time is also helpful.
    As long as you are debugging at low voltage, you can keep the RS1G in place until the higher-rated parts come in.

    Also, I recommend to change D7 from SMAJ100A to SMAJ130A or maybe up to SMAJ150A. This will reduce the leakage inductance energy loss and improve efficiency. It is not necessary to change this for debug, just a recommendation for long-term improvement. But check the peak voltage at the DRAIN under all operating conditions.

    It is not easy to test this kind of controller (PSR) because you must provide the same waveform at VS as shown in Figure 22 (page 18 of the datasheet). The IC relies on detecting this waveshape, not only at the sense-point "knee" but also at the zero-crossings. This is not easy to recreate in an open-loop test circuit, so the actual power supply is the best test circuit to debug in. Keep the input voltage low and have no load or very light load on the outputs. Make sure your ac or dc source is isolated from the power mains.

    I hope that beginning with a corrected connection and a new controller will lead to proper start-up.
    Regards,
    Ulrich
  • Hi Ulrich, 

    Thanks you. My problem has been solved. The chip, UCC28910, has been broken due to the wrong connection. After I changed a new one, it works fine. 

    Thanks. 

    Yin