5
\$\begingroup\$

I'm carrying out a research project this year at uni, which will include creating small electronic sensory devices for children (under 7) with autism. Obviously, safety has to be taken very seriously.

Only relatively small amounts of current will be used - probably powering simple sensors on an Arduino board - but seeing as my knowledge of all things electronic is pretty limited, I thought I'd try and get some suggestions first.

From the research I've done, I found PoE at littlebird: http://littlebirdelectronics.com/collections/freetronics/products/4-channel-power-over-ethernet-midspan-injector - which looks like a pretty good solution, seeing as I want the ability to control the device remotely, over a local network. A benefit also being that I can keep the number of wires poking out of the device to a minimum and also not have a direct lead going to a wall socket.

So far, I'm really only at the stage of planning the 'core' of these devices, which would be an Arduino board with limited sensors attached, possibly using the internal parts from a Wii remote. Because I don't yet have confirmed participants, I don't know exactly what sensory needs I'll be working with (each autistic child will have specific sensory issues) and therefore exactly how much power I will need for the devices. I will aim to keep it at a minimum though, simply for safety reasons.

\$\endgroup\$
5
  • 1
    \$\begingroup\$ Reading your question it looks like you have an issue, maybe "what is the easiest way to power this device where children cannot reach it?" and now have done some research. After research you found something that might work. Your question is geared to your solution, if you were instead interested in some original question asking it directly often give very very good answer, sometimes informing you of something you would need to be in the field to easily learn about. \$\endgroup\$
    – Kortuk
    Commented Mar 20, 2011 at 9:42
  • \$\begingroup\$ You're absolutely right - it was very much a leading question. From the research I've done PoE looks to be a good solution, but because of my limited knowledge, I'm definitely open to any suggestions. \$\endgroup\$ Commented Mar 20, 2011 at 11:29
  • 1
    \$\begingroup\$ @kodampixel, if you edit your question to include what you need, all of your specs and let us decide we can help, but we really need all of your requirements. It may take a bit of time to type it up, but your question was well written and I am sure you will like the result of telling us what you need and experts weighing in. \$\endgroup\$
    – Kortuk
    Commented Mar 20, 2011 at 11:52
  • \$\begingroup\$ Thanks for the suggestion, Kortuk. I've updated the post, which hopefully makes things a bit clearer now. \$\endgroup\$ Commented Mar 20, 2011 at 20:46
  • \$\begingroup\$ it is what they don't pay me for : ) \$\endgroup\$
    – Kortuk
    Commented Mar 21, 2011 at 2:46

3 Answers 3

3
\$\begingroup\$

The Power over Ethernet (PoE) is likely most suitable if the devices are network-aware, otherwise one Ethernet cable is pretty much equivalent to a (low current) power cord or a USB cable.

For any devices that to be "worn" or "attached" by test subjects, I would strongly consider an un-tethered design, powered using rechargeable batteries. For small signal (i.e. sensors, data logging, no motors, very simple LED lighting) this option should be easily economic and practical for up to 24 hour periods or longer. NiMH would be my first choice due to low cost and wide availability, with Lithium as a second. Just ensure you use an appropriate charger for the battery type, and things should just work.

Note that most rechargeable batteries such as 'AA' size cells, only provide 1.2 rather than 1.5 volts, so 4 of them is not sufficient for 5V needed to stably power an Arduino, while non-rechargeable (i.e. disposable) cells such as alkaline or zinc-carbon would.

With electricity you need to be aware of both voltage and current. The Arduino itself takes an input range of 7-12 DC, up to 500mA (I don't have a reference on average / typically current, but would guess around 100-125mA or less), from an external DC power source. The USB port can also be used to draw 5 volts, up to 500mA from a powered USB hub or powered USB port.

Using a low-cost low current (e.g. 250-500 mA output) AC-DC power adapter (wall wart) would be a default method if there is non-trivial power consumption, or needs to continuously operate for a long period of time. A modern switch mode power supply (SMPS) based unit can be had for a modest cost, and is light weight, being able to dispense with the need for a large power transformer encased in it. Combined with the power limiting capabilities already included in the Aruino (UNO) from the resettable fuse (PTC, I believe) used for USB power source, and/or the linear voltage regulator used to regulate power from a AC-DC power adapter which also includes over-voltage, current limiting, and short circuit protection unless you need addition power requirements (e.g. motors, high power LEDs) you can use the protection built into the Arduino as sufficient for electrical shock/burn protection.

\$\endgroup\$
2
  • \$\begingroup\$ Thanks, that was a really helpful response - loads of new ideas for me to look at. Fortunately, I don't need to contend with attaching anything physically to participants in the study, so I think I will try PoE to start with (as you say, it also ticks the box of being network aware). I hadn't thought of the Arduino being capable of acting as shock protection for low currents though - that will also being really helpful. \$\endgroup\$ Commented Mar 22, 2011 at 7:37
  • \$\begingroup\$ To oversimplify a bit, when working with low voltage and low current electronics, the biggest remaining risk is shorting (short circuiting) the power supply. So current limiting (such as fuses & resettable fuses, circuit breakers and over-current protection devices/ICs) deals with the majority of the issues. The second is power supply polarity protection which can be handled with a simple diode (or P-channel MOSFET with its gate tied to ground) to ensure that the positive voltage input is correctly connected. \$\endgroup\$
    – mctylr
    Commented Mar 23, 2011 at 21:34
5
\$\begingroup\$

A PoE switch is normally used to power IP phones over an Ethernet cable without having to have an additional power adaptor. It sounds like you're doing a something similar with your devices.

The midspan injector will certainly be a cheap way to add power to an unpowered Ethernet port, with the added advantage of being able to control the supply voltage. You can get PoE switches with everything built in, so I also recommend looking at these as it might be cleaner and easier to use one, but it might cost a bit more.

You could also make your own version of the midspan injector for even less. You'll need some RJ45 sockets and a power supply, and just simply connect the data pins straight through, and the supply to the output sockets to the power pins.

Here is a table of the pinout for RJ45 ethernet with and without power:

PINOUT

One thing you need to be careful of is that 'standard' PoE is 48v, which will drop to as much as 32v depending on the length of your ethernet cable. Make sure that when you split out the supply from the ethernet on the Ardunio end you have something to reduce the voltage to 9-12v. If you're using an injector then you can control this and supply a lower voltage, but if you use a standard PoE switch you need to be aware of this.

What type of sensory devices are you going to be connecting to your PoE ports?

\$\endgroup\$
1
  • \$\begingroup\$ Thanks for the speedy reply! I'm planning on using very simple and limited sensory inputs: perhaps a pressure sensor, or accelerometer. These would possibly trigger something like a small motor, or LEDs. So battery power could be an alternative option, but I would still like to be able to communicate data with the local network, meaning cabling or a transmitter of some description. \$\endgroup\$ Commented Mar 20, 2011 at 11:27
3
\$\begingroup\$

You probably have a human research review board to pass you decision thru (at least in the us ). Batteries are often the simplest safe power supply. You still need to isolate from any other electrical source, probably including your network. Check with the review criteria.

\$\endgroup\$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.