Hello to everybody,
First of all, I want to sincerely apologize for replying back so late. There were some issues and I was sent to the field for several weeks. I was able keep working on my design but not to write back until now. I hope not to make this long. I want to share some very insightful waveforms I was able to get with your support; I dont want to only ask questions without actually sharing the results and share something usefull here. .
I did went through the ON Semi appnote and included inrush limiting in my design (although it is maybe not necessary).
First about the design changes:
1) I changed from IRF7726PBF to the IRF9321. In my attempt to make an objective and best engineering decision, I made the table shown in attachment "Mosfet Selection Table" putting together all P-channel mosfets suggested and investigated. The winner was of course IRF9321 becuase has 2x decrease in price, 4x decrease in power dissipation and around 5x decrease in temperature rise. It would also improve input-output efficiency of the complete system.
2) The IRF7726PBF was not necessarily a bad decision in terms of performance. It does have a higher voltage drop and power dissipation compared to the IRF9321, but can handle very well the voltages and currents that I explained the first post. Obviously, in the presence of a better option (more importantly in terms of price), the change was an all win.
About the actual measurements:
1) First I wanted to identify the Turn ON regions and characteristic of the MOSFET. Something like getting the equivalent Figure 6 of the ONSemi appnote, but for my mosfet. See image "No Inrush Limiting Vds Id P Turn On Waveforms into Kindle.png". It is not as clear as in the appnote, but it is possible to see the different Vgs stages and Vds transition.
2) Inrush current limiting: The possible devices that will be connected to the outputs of my circuit are primarily cellphones (smart and not smartphones), 12V LED light bulbs and possibly 12V TVs. In figure "2.35A Load No Inrush Limiting Vds Id Pd 1.1uF Load.png" I show the mosfet's drain-source voltage in blue {Vds}, Drain current in orange {Id} and Instantenous Power in light blue {Pd} for a 2.35A resistive load in parallel with a 1.1uF capacitive load connected to the 12V output WITHOUT inrush current limiting circuit. You can see the following characteristics:
- Turn ON time aprox 1us.
- Peak current: aprox 26A.
- Peak instantaneous power: aprox 33W
- Power pulse duration: aprox 1.2us
Then I connected the resistor-capacitor network suggested in the appnote to limit inrush current and, in fact, the inrush current was significantly diminished with the expense of increased power pulse peak and duration. See image "2.35A Load Inrush Limited 330Ohm 1nF Vds Id Pd 1.1uF load.png". Specifically you can see the following characteristics:
- Turn ON time: aprox 1.5us
- Peak current: aprox 15A
- Peak instantaneous power: aprox 68W.
- Power pulse duration: aprox 1.5us
The power pulse more than doubled with the sligthly enlarged Turn On Time and reduced inrush current. In both cases the turn ON Power pulse and Vds, Id characteristics are tolerable by the IRF9321 as are under the SOA Curve. See image "IRF9321 SOA".
At this stage of design, there was still pending the actual implementation of active shorcircuit, overvoltage and ESD protection for these outputs. I investigated several methods to achieve this, looking simultaneously for "Smart Switches" as was suggested by several people. The short answer for the system I was designing is that going for the "Smart Switches" was the best option. I was completely amazed that those devices include ALL the protection characteristics I needed and even more. Now, for my application, when doing the comparison between a discrete solution and the IC, in terms of price and time of implementation, the IC solution was a clear winner. But in these sense I got to know that viable solutions for this were available for relatively low power requirements as mine. But for higher voltage and current requirements (maybe 3x-5x of mine) the IC solution might not be available or might be more expensive than a discrete solution.
That being said, these are the final devices I selected as output pass device and output protection:
12V, 5A output: TPS25910
5V, 3A output: TPS2592BA
Both devices are amazing in the rich feature set they offer and their robustness. I did not use the IRF9321 in the actual implementation, but it was a very insightful process that gave me real experience.
Beyond that, If anyone else reads this in the future for a similar application, and have different voltage/current requirements, I will recommend to look a the following:
1) Search in google for "Smart Switch" and "eFuse". Most manufacturers refer to these devices (pass transistor with a lot of protections) as "eFuses". Also can look for "Hot Swap Controllers" whose intended end application might not be only output protection, but can fill the requirements and price very well.
2) Look at the Texas Instruments range of "Hot Swap Controllers" and "eFuses" and "Load Switches". You can look for example at the TPS25910 or TPS2592 I selected, and the page will recommend other devices as well.
3) Look at the Infineon ProFET, ProFET+ or HiC ProFET families. These devices provide a lot of protection features and a wide range of RdsON devices. Some are "automotive" grade and designed for that end market, but that does not mean at all that they cannot fit other non-related applications.
http://www.infineon.com/cms/en/product/channel.html?channel=ff80808112ab681d0112ab69e2d403574) Look at the ST Family of "VI-Power" devices. At some point I tested the vnn7nv04ptr-e which in the end was not suitable, but they have several other devices. They also have some "STEF12" and "STEF5" eFUSES, intended for 12V and 5V power rails, respectively.
5) After investigating about those Smart Switches, I think it is highly recommended to analyze the use of these devices when reliable output protection is required. If you need short circuit, overcurrent, output overvoltage, thermal shutdown and/or foldback, ESD output protection, many of these devices are designed and intended for that kind of requirements.
I would like to thank everyones recommendations and support here. Also for pointing out the existence of the smart switches.
Home
Help
Search
Login
Register
