TPS54310 Difficult Problems--Solution Sharing
[Copy link]
This post was last edited by qwqwqw2088 on 2018-9-21 15:41 TPS54310 is a TI low input voltage 3A synchronous buck converter with adjustable output voltage. It is used in many low voltage input and high current output synchronous PWM buck voltage conversion circuits. Now I summarize the various problems encountered by netizens using TPS51216 and some solutions of TI experts, and share them with the children who are learning about power supply from TI in eeworld. Welcome to share your experience in using TPS51216. 1. TPS51216 output problem Design goal: Input voltage: 5V Output voltage: 1.8V Output current: 4A The maximum circuit diagram is as follows:
Problem phenomenon: First, set S3 and S5 high, and the output of the inductor after the stage is measured to be stable at 0.736V, not reaching 1.8V. When using a multimeter to measure, it shows that it first reaches 1.8V and then immediately drops to 0.736V. Because VLDOIN is connected to 1.8V, VTT and VTTREF do not output 0.9V. VREF output voltage is 1.823V, and VREFIN voltage is also 1.823V. Then the VREFIN voltage is divided to 1.5V, and the output of the after stage is still 0.736V. 1. Is it correct to set the feedback resistor R477 to 10K and not solder R478? 2. Will connecting the MOS tube Vin to 5V affect the output? 3. Is the inductor value reasonable? I compared the reference circuit slur270.pdf page7 and did not find much difference. Please point out the possible problems in the design. Thank you TI expert reply: 1. For the value of feedback resistor and inductor, please refer to the data sheet SLUSAB9A. R478 cannot be left empty. 2. It is OK to connect the MOS tube to 5Vin. 3. It is recommended to capture the waveform of the VSW node (under no-load and heavy-load conditions respectively). 4. The internal LDO of VTT and VTTREF has no output because the voltage at your VCC1V8 is unstable, so the first thing you need to solve is the output problem of the switching power supply Another question? 1. Because the output voltage upper limit of TPS51216 is 1.8V, the reference voltage of VREF output is 1.8V, so R477 is selected as 10K, so R478 is left floating in order to make REFIN 1.8V. If R478 must be welded with a resistor, a resistor of about 1M is required to meet the requirements. 2. In order to verify whether R478 causes incorrect output, a resistor of about 50k was soldered at R478, and REFIN was set to 1.5V, but the output of the MOS tube post-stage was still 0.736V, so it is suspected that there is a problem with the control loop itself. I hope you can help review whether the selection of resistors, capacitors and MOS tubes needs to be optimized. Does the boost pin need a diode to pull it up? TI experts answered again: For the values of inductors and capacitors, refer to the formulas in the data sheet. In addition, you don't need to add a pull-up diode to the boot capacitor, as long as the voltage coming in from V5IN is a stable 5V, there is already a diode inside. For MOS selection, just ensure that its gate drive voltage is around 4V. Regarding the loop, since this is DCAP control, we only need to pay attention to the selection of the output capacitor in the loop. It is recommended to add a small capacitor of about 0.1uF at VREFIN. So it is recommended that you can capture the MOS tube driving voltage waveform, the voltage at VREF and VEFIN, and the VSW waveform. Thank you. Question: Hello, What is the selected current carrying value of the inductor? The inductor is IHLM2525CZERR47M07 (0.47UH) with a saturation current of 25A In addition, according to the steady-state requirements on page 17 of the datasheet Currently the mode pin pull-down resistor is 200k, which means the switching frequency is 400Khz According to the current design, Cout is about 330uf, and the ESR is estimated to be 8 milliohms The left side of the formula is 1000000/(2*3.14*330*8)Khz=60 The right side of the formula is 400/3Khz=133 Because other small capacitors are connected in parallel in addition to 330uF, the equivalent capacitance value will be larger and can meet the steady-state conditions. In addition, should a 1M ohm resistor be added at R478 based on the last analysis? 2. How to convert the power chip TPS54310 from 5.5V to 1.2V? The TPS54310 chip is used in the project to convert the system input voltage of about 5.5V to 1.2V to supply the core voltage of TMS320C6416 series DSP. The 5.5V is stable through monitoring, but the chip TPS54310 outputs 1.2V and occasionally drops from 1.2V to 0V, which lasts for about 20ms and then quickly rises to 1.2V, causing the DSP to fail to operate normally. In this case, I would like to ask you, from which aspects should we locate the cause of the fault? Thank you. TI expert reply: Remove R69 and C12 to see if there is still any abnormality. Also, it is best to recalculate and optimize the peripheral parameters of your comp pins. For the specific calculation method, please refer to the TPS54110 data sheet. The two chips are from the same series. I studied this document and the 54310 document itself and found that their methods of calculating parameters are different, and the parameter calculation method should also be different from the one I uploaded (I cannot calculate my parameter value from the formula). How should I calculate and optimize? If you can have such a calculation formula, but I don't have a spectrum analyzer, how can I verify the rationality of the optimization? Hello, what is the difference you are talking about? There is no parameter calculation formula in the 54310 data sheet. For verification, mainly look at your SW node waveform under various load conditions, it is best to use an analyzer 3.About the ground of TPS54310 Dear experts. I would like to ask, are all the reference grounds of TPS54310 analog grounds? ? ? There are analog grounds and power grounds. The layout and wiring can refer to the figure below Is the ground of this power chip connected to the analog ground? Because I use this chip to power the digital circuit, I am confused about this! Thank you for your reply! Answer: Yes, the analog ground and digital ground should be separated, and finally connected together at a single point In addition, please confirm again, AGND and PowerPad are connected together, and connected to the analog ground, right? Is PGND connected to the DGND digital ground? Looking forward to your reply! ! Thank you very much! TI expert reply: These three grounds are all single-point grounding. AGND and PGND are single-point grounding. DGND and PGND are also single-point grounding. TPS54310 does not have a digital ground, because this is an analog chip, there is no digital part, and there will be no digital ground. The digital ground I mean that if there is a digital part in the system you design, there must be a digital ground. The digital ground and analog ground must be grounded at a single point. For systems with relatively high power, the ground connection must also pass through a π-type filter >>More Questions and Answers
| 
提升卡
变色卡
千斤顶
京公网安备 11010802033920号
