BBC micro:bit battery power

Hi Stephen,

It's a JST 'PH' connector. I wouldn't try anything higher than 3V, since that is what's marked on the board. Anything lower could be tested, to see if it works. 2.4V might work. LiPo cells are nominally 3.7V but actually have a wide range between fully charged (around 4.2V) down to fully discharged (<2.9V). Many ICs are not designed to operate with this range, so I think it would be very risky to try this on the micro:bit. If you really wanted to push it, perhaps 3.3V might be a limit but I'd strongly suggest sticking to 3V since this is what is marked.

I've seen no spec, so this is just a best guess.

So, I'd rule out LiPo, and just try 2xAA/AAA primary cells, or 2xAA/AAA NiMH for example.

shabaz thanks for your response. I suppose I can just cut the lead off the battery box that comes in the micro:bit kit. (to attach to my own power supply) I think what is marked on the board and stated in the documentation is just classic British 'dumbed down' info. I know the micro:bit is aimed at 11 year old school children. But even an 11 year old can understand the concept of a range of values. And even if some can't, just print 3V with prominence and put a note in the teacher's notes or something. After all, renewable energy (re-chargable batteries) is supposed to be a thing in schools.

I found this http://infocenter.nordicsemi.com/pdf/nRF51822_PB_v2.5.pdf which is Nordic Semiconductor's Product Brief doc on the nRF51822 SoC which states 'Wide supply voltage range (1.8 V to 3.6 V)' for the SoC itself.

The whole reason I questioned this was that I am unaware of any 3V 'chips'. Usual voltages are 5V, 3.3V, and 1.8V (also 2.5V but not so common?). But then again, maybe there is/are some other component(s) that are more voltage sensitive?

JST is a company name. If it is a PH type, here are the bits you'd need:

2-pin shell:

http://uk.farnell.com/jst-japan-solderless-terminals/phr-2/housing-2way-2mm/dp/3616186

crimp pins:

http://uk.farnell.com/jst-japan-solderless-terminals/bph-002t-p0-5s/contact-loose-piece-0-05-0-22mm/dp/3617210?MER=en-mer-0713-pd-r1-acce

In theory you need the proper tool for the crimps, but if you're not manufacturing it's quite easy to crimp them with pliars.

For prototypes I like to solder them, as that's usually more reliable, but that's just a comment on my crimping skills really. (To solder: strip the wire end; twist; tin; cut to a couple of millimetres; lay the end where it would normally crimp - the inner of the two crimps; solder; then crimp the wire support - the outer crimp. It helps with inserting them in the housing if you bend the inner crimp tabs so that they're upright rather than slightly splayed before you solder. The reason to tin first is that you then pre-shrink the insulation, otherwise it will shrink back and the cable support won't properly clamp the insulation. You need to watch when you solder that you don't apply so much solder that it flows down to the contact area.)

If you knew all that, just ignore me. It might help others.

It seems strange to me that the micro:bit documentation is so poor; they obviously see the children as becoming coders rather than engineers! I asked about the suggestion here (in the music box project) to apply an 8ohm speaker to an IO pin (a nasty load for a processor output to drive directly) and was just ignored. [I presume the reason that the poster was so casual about it is that the board probably has series resistors to improve the ESD protection on the IO pins and to prevent damage when they get shorted, but there's nothing on the micro:bit site that I can see to tell you if they are there and what the value is, so how would you sensibly design anything to connect to it?]

The Nordic part isn't the only device on the board. You'd need to check all of them. There's also the changeover from USB power to battery to consider - the USB will be regulated down from +5V to give the +3V [if it is +3V]. The designer might have assumed that the battery voltage would never be more than the 3V you'd get from a pair of fresh alkaline cells. Who knows in the absence of documentation?

They have not released schematic for the bbc micro:bit? That's pretty pathetic.

Detailed specs will help, but I suspect they're concentrating on their core use-cases currently. I'm sure more information will come in time for exactly these types of reasons, i.e. the need to know limits for connecting batteries.

The 3V on the edge connector is actually closer to 3.3V, and from looking at the board there are a couple of diodes, so (speculating) there is probably a 3.3V regulator off the USB input, and this and the battery connector are combined using the two diodes. So, the battery input can possibly be acceptable up to 3.3V, which still prevents LiPo cells without additional circuitry. So for this reason I'd still recommend only dual 1.5V primary cells (e.g. AA or AAA sized) are used, or 2x AA or AAA NiMH rechargeable cells.

If an external regulator is planned to be connected to the battery connector (say to allow the use of LiPo) then I'd pick a 3V one, but for the reasons above I wouldn't be surprised if anything up to around 3.3V is ok.

According to this article, the voltage was brought all the way up to 4v.

micro:bit Power Consumption | The REUK Blog

Not sure how much strain (if any) this would have placed on the components.

Interesting info.

As you say, that will likely stress the parts if the circuit is not designed for this.

Just checking, looks like the NXP chip on-board is rated up to 3.6V for correct behavioiur, but behaviour is undefined but won't cause damage up to 3.8V. Beyond that damage may or may not be caused for that chip, in the short term or over a long term.

In practical terms, I would just use a 3V or 3.3V regulator. Since it is designed for connecting to a 2xAA battery pack, 3V is a known good voltage to select for the regulator if a higher voltage source needs to be used.

A LiPo cell directly connected is not advisable given what information we have currently, since it has such a wide voltage range, up to 4.2V.