Part Number: PMP7282
I am working on PMP7282, 50A, Non-isolated DC-DC synchronous buck converter and we used same schematic as PMP7282.
I have gone through following issues.
1. when we are loading at 50A Input ceramic capacitor have failed and burnt; where we are using Electrolytic capacitor (4 No, 100V/100uF. i.e. 2 for master and 2 for slave) and Ceramic capacitor (10 No, 100V/4.7uF i.e. 5 for master and 5 for slave). Please suggest us the selection criteria if we are using either only ceramic or Electrolytic or both. Which capacitor type take more voltage rating.
2. All MOSFET's have shorted at very Lower input Voltage (35V) and No-load condition, please suggest us that how to adjust the dead band time for the switching,
3. which pins have to be verified the synchronization working properly or not.
4. Both Master and slave are not sharing the load equally.
5. Please suggest, PMP7282 can be used for Automotive Application or not i.e. we want use this converter for Battery charging after rectifier output.
6. what are the possible root causes for MOSFET, Capacitor and IC failures.
Thanks & Regards,
1. Ceramic capacitors usually fail due to mechanical damage. Flexing of the board can cause cracks. Parallel aluminum electrolytic capacitors are used for damping at the input, in particular when a filter inductor is used or there are long wires from the input power source. RMS currents will split between aluminum and ceramic based on the impedance at the switching frequency. This should be evaluated to ensure that the aluminum RMS current rating is not exceeded. 100V rating on either aluminum or ceramic is adequate.
2. There is no way to adjust the dead time on LM5116. The delay is set internally to around 70ns, The adaptive gate drive will time out after 150ns. See the data sheet section HO Output for a detailed description.
3. The test report has a section on Synchronization that you can compare waveforms. Please note that normal probe capacitance can load the SYNC since 27pF coupling capacitors are used. For accurate measurement and operation I used low capacitance probes. You can disable the slave and disconnect the SYNC coupling capacitor and see if the master runs properly on its own. On my prototype I had switching noise from the master inductor getting into the SYNC 2N7002 buffer FET, so added 27pF gate capacitor.
4. Check the COMP voltages to make sure they are the same. Mismatch in the 10k and 4.99k gain resistors or external reference can cause an offset in current sharing.
5. The output of the converter cannot be connected directly to a battery. A battery needs a current source for charging. PMP7282 has a constant output voltage. This would require another current sensing and control loop for the output.
6. It is likely that there is some type of system instability, either noise, input filter or control loop. If you would like further assistance, please forward your board documentation, pictures of your test setup and any waveforms or test results to me.
We are glad that we were able to resolve this issue, and will now proceed to close this thread.
If you have further questions related to this thread, you may click "Ask a related question" below. The newly created question will be automatically linked to this question.
In reply to Bob Sheehan:
Thanks for reply.
2. as we cannot provide any dead-band time externally , please suggest that How to avoid the dead short of MOSFET's.
3. please check the Load unbalancing Issue waveforms in attached file(Here, scope having some problem, ignore the ringing in inductor current).50A.xlsx
3 & 4. i will try and get back to you
5. we are connecting directly battery to Load;
Here we have setup of (AC Generator+ AC-DC + DC-DC + Load + Battery);
please suggest us any changes have to be done in Setup.
In reply to Khagesh Rao:
Hi M Khagesh,
Looking at your spreadsheet results, there is some instability causing erratic switching. It could be noise getting into the sync or current sense. It could also be input filter oscillation or control loop stability.
Since PMP7282 is stable, there might be something different with the board or setup. If you are using a heat sink on the back of the board you could get switching noise coupling into the control circuit.
The gate drive voltage looks too high at 14.5V. This is not good as the operating rating of VCC/VCCX is 15V with abs max of 16V. You will need to reduce this to a level between 10V-12V in order to have enough drive but with safety margin for transient. Connecting VCCX to VOUT is fine at 12V, but not 14.5V. The drivers are probably getting damaged due to instability causing transients and possible ringing on the gate drive.
2. What you call shoot through on the waveforms is not caused by insufficient dead time. It is retriggering due to the high side FET turning on very fast due to the abnormally high drive voltage, and coupling through into the low side gate. Reducing the drive voltage will help.
3. Once you solve the instability, the load balance issue will take care of itself.
5. A constant output voltage converter cannot be connected directly to a battery for charging. You will have no control of the charging current, so it will charge at short circuit current. The battery acts as an infinite capacitor and the control loop will go unstable.
6. Do you have a picture of the setup or system schematic/diagram?
1.i have tuned the output voltage to 12V and also i took ti suggestions to reduce instability. Now it is working fine till 40A.
2. But our final application is battery charging and also i have gone through some solutions for cc control but all are ic based only.
So, please suggest the cc control circuit of PMP7282 for battery charging.
Thanking you sir.
I do not have a simple circuit that has been tested with the LM5116 to regulate the output current. The most straightforward method is to use a high side current sense resistor and level shift amplifier to OR into the feedback at FB. You still need to go through the design analysis and prototyping to prove out a circuit. The frequency compensation may need to be adjusted for stability. An OPA191 may work with a P-CH FET.
All content and materials on this site are provided "as is". TI and its respective suppliers and providers of content make no representations about the suitability of these materials for any purpose and disclaim all warranties and conditions with regard to these materials, including but not limited to all implied warranties and conditions of merchantability, fitness for a particular purpose, title and non-infringement of any third party intellectual property right. No license, either express or implied, by estoppel or otherwise, is granted by TI. Use of the information on this site may require a license from a third party, or a license from TI.
TI is a global semiconductor design and manufacturing company. Innovate with 100,000+ analog ICs andembedded processors, along with software, tools and the industry’s largest sales/support staff.