Greetings and Help with Power Routing / Oring Board Design

[jwhitten]

Hello everybody,

This is my first post on this forum.

I'm hoping to get some assistance with designing a Power Routing Board for my robot. Some background about me. I have general electronics experience, most of it from some years ago, but I'm pretty new to working with MOSFETS and all these new fancy Power-Handling chips and such. In general, I must say that it looks like MOSFETS are awesome replacements for the BJT's that I've used in the past! ;-)

[TL;DR: My questions are on down toward the bottom of my post]

Here's my setup, I am building a 6-wheeled "rover" type platform to use as an experiments platform and one of my requirements is for a "human scale" operating span, meaning that it must have enough onboard power to approximate a human day. For my purposes that means that the CPUs and sensors must be fully operable for around 16+ hours of a day, and capable of operating in standby capacity for up to 24 hours (or longer). It must be capable of moving for at least four hours a day between charges. Longer is fine.

To this end, I have acquired two 9AH 12.8VDC LiFePo4 batteries with SMBus (I2C) communication, which I estimate gives me about 115 watt-hours each. Each battery has a built-in TI BQ3050 battery mgmt chip and provides lots of interesting operational detail about various aspects of the battery: run rate, estimated time to depletion, current usage, average current usage, voltage, etc. lots more.

The main CPU is an ODROID C1 (similar to a Raspberry PI 3 with a tad bit more horsepower), several Arduino "Nano" style MCU's to handle the various sensors, and probably an additional CPU of some sort (still to be determined) to handle one or more cameras. All-in-all, not counting the drive wheel motors, I estimate that the typical current rate will be less than 2AH. I have already empirically established that one of the batteries can keep the ODROID running at full-tilt for around 28 hours (which was around a 1AH burn rate). So far none of the sensors I'm using (infrared proximity, ultrasonic ranging, camera, etc.) are using anywhere near another AH, so I think I'm pretty well on target for my electronics power goal.

When the wheel motors start turning though, I expect they'll eat into the power budget pretty heavily. According to the specifications, the wheel motors can have a (max) stall current of around 3 Amps (times six motors), so I'm using that as my approx theoretical max battery draw parameter and selecting parts that can provide a maximum of 16 Amps (though in reality, I know that the beastie will likely be on fire if it ever gets to that point! ;-) Practically, I'm expecting the motors to draw around 2-4 Amps (all combined) as it operates, depending upon terrain and speed. And I don't really plan to have it moving for four hours continuously, I'm simply earmarking that number as my goal.

So this is what I'm thinking with respect to power routing... whenever the robot is on the workbench it will be hooked up to bench power ("Free Power", from the perspective of the robot's operating power budget) and will *always* prefer benchtop power to any other source (unless forced to switch for some test or experiment). When bench power is unavailable, it will select from one of the two batteries to provide mobile power. As the robot will be in constant communication with the batteries, it will be able to choose whichever battery is better suited for whatever task is required, and/or can switch to the other battery when/if needed, as needed. I have already worked out a simple circuit to manage the selection of batteries and chargers-- though I noticed that the power mgmt chip below seems to use a different scheme. I did not include that circuit in the attachments but I can certainly upload it if anybody's interested. Though it's simple, it's kind of neat.

My first thought was just to use some relays to handle the power routing, and some big-honkin capacitor to use for a hold-up cap. (See RobotArdi_PowerMgmtOverview attachment). Shortly thereafter, I thought maybe I would use this need to learn more about MOSFETs and perhaps make a better board with better power management qualities. So I started reading up on MOSFETs and experimenting on the breadboard to come up with a better design. I learned about building High-Side switches out of MOSFETs and designed a replacement board to use instead of the relays (See second attachment: RobotArdi_PowerRoutingBoard-2). I thought I was on the right track during an earlier incarnation of this board, but using single MOSFETs instead of back-to-back MOSFET pairs-- when I blew up my first MOSFET-- let out all the blue smoke! And that's when I began learning about about reverse power issues and MOSFETs...

So I started reading some more and discovered that there is a whole panopoly of Power Chips available nowadays for managing various power issues and so I figured that they probably had the subject covered better than I could do on my first time out-- so I decided to invest in some modern "High Side Switch Drivers" from the automotive industry. So my simplistic thought is that I could combine them into a similar arrangement as the "relays" in my original drawing (sort of a hybrid between that and the discrete MOSFET design)-- and then I started reading about specialty "Power Oring" circuits to help manage combining power sources-- which sounds like exactly the situation I'm looking for.

So that brings me to now, and to my questions...

I have selected (and purchased) the STMicroelectronics VN7003ALHTR "High Side Driver" component as the active replacement for the MOSFETs in my simple circuit. Then I started reading about the Linear Technology TLC4417 "Prioritized Power Path Controller" which is designed to assist with "Oring" various power sources together, which also sounds like it is very much in line with what I want...


Links to data sheets:

LTC4417:
http://www.linear.com/product/LTC4417

VN7003ALHTR:
http://datasheet.octopart.com/VN7003ALHTR-STMicroelectronics-datasheet-66373282.pdf


1. Can I use the LTC4417 in high power (16A) applications? Is it simply a matter of selecting beefy MOSFETs for the external devices? Or is there some other factor that I need to consider. The datasheet only ever mentions 2Amps, and that was in one of their example circuits.


2. If I use the VN7003ALHTR parts, do I need the LTC4417? Can they withstand the reverse power conditions just by tying the outputs together? (the data sheet says they can withstand up to 24 volts reverse volts). They were a little expensive and I don't want to blow them up in experimenting.

3. Does the LTC4417 simply replace the VN7003ALHTR parts-- albeit with the inclusion of other discrete high-current MOSFETs, per the example circuit?

I have looked around everywhere I can think to look and there doesn't seem to be a lot of guidance available for that specific part, or for "Oring" power sources together in general. I've read through all the discussions I could find on the subject here and elsewhere, but just haven't been able to satisfy myself that I am suitably versant on the subject to proceed yet.

I would sure appreciate it if anybody offer any information or assistance?