How to run Nano 33 BLE + NeoPixel Ring from (a single) battery?

Hi

I've just started hacking with embedded devices and love it so far. My idea is to control 4 NeoPixel Rings with the Arduino Nano 33 BLE (only one at a time) and run the whole sketch from a battery.

I've realized the NeoPixel rings require 5V to operate. Also, the NeoPixel Uber Guide suggests using level shifting to ensure stability (although not a hard requirement).

Now I am confused about the power requirements. According to some posts in this forum, the board was designed to run from coin cells like CR2020. However, they have a shallow capacity, which might be enough only without any peripheral.

The circuit must be able to run from the battery for roughly 14 hours. The whole package (meaning board + LEDs) shouldn't be larger than 7x7x7 cm.

According to the NeoPixel guide, one LED/driver consumes roughly 20 mA (not at full brightness). So for a 16 LED ring, the following calculation should be correct. 16 x 20 mA / 1'000 = 0.32 A.

Further, the Nano itself requires roughly 1A max (Nina B306 = 15mA + LSM9DS1 = 2mA + User application = 950mA = 967mA)

Which leads to a required capacity of at least 18.018 Ah (14h x (0.32 A + 0.976 A)), correct? :-/

Assuming the math is correct, the battery would have to be quite a beast. Is this setup powerable by a battery in a practical manner? I mean, according to the size and runtime requirements mentioned above?

How would I design the (battery-driven) power supply to feed the Arduino and one NeoPixel ring at a time? Should I go for a high-capacity LiPo and step up the voltage with a booster?

Thanks in advance

You have a lot of misinformation here, many things you think you know are wrong.

Utter rubbish, please point to these posts and I will shoot them down.

Again rubbish, it requires no such current. 1A is a very large amount and you will fry your board if you try and sourc or sink that much current from it.

I read it again it says no such thing.
Each LED consumes 1mA when it is off, and up to 60mA wheN. It is on full brightness. You can set a brightness limit in software if you want.

You can chain all the rings off one pin and have the software control what parts of each ring gets turned on at any one time.

Avoid voltage boosters here performance is not a good as EBay figures would have you believe. use a buck regulator ( reduces the voltage ) from a reputable supplier.

I took this figures out of the datasheet. Aren't these the max current values that can be used to size the power requirements?

The NeoPixel guide says that 20mA corresponds to 1/3 of the max current. From the guide:

So this is also not correct?

In the meantime, I came across the idea and use 4x 18650 (3Ah) in series, which would result in 12Ah @14.8V. These batteries are not too big. According to Nano's datasheet, the MPM3610 DC-DC regulates input voltage from up to 21V.

So If I get that right, I should be able to connect the batteries to VIN (PIN15) and GND (PIN19) and the Nano works, right?

But then, how can I take 5V from the battery to feed the NeoPixel ring? There is VUSB (PIN12), which can be connected to VUSB pin of the USB connector by shorting a jumper underneath the board. However, this won't help, as there won't be any USB power supply plugged in. So I guess this is where the buck converter fits in.

Maybe it's better to stick with 3x AA batteries supplying 4.5V and connect both the NeoPixel ring and the Nano to it? I have read somewhere that "extra amps are good, extra volts are bad"

Sorry for the noob question and thanks for any help that points me in the right direction!

BLE was designed to allow the development of wireless devices powered by coin cells like heart rate monitors and similar sensors. That would usually require you to have a very minimalist PCB and a software that sleeps most of the time. You would need to disconnect the power circuit of the Arduino and other components on the board to have a chance to get ultra low power and you would have to analyze and modify the software of the Arduino too. There are some posts in the forum where users tried to get as low as possible. Not sure what the lowest power consumption was. But then you would likely not be able to drive Neopixels with some changing patterns.

I think you will need to do some experiments, because all types of batteries and power circuits have there advantages and disadvantages. The voltage drops over time and you will need to see at what levels your pixels and the Arduino still work. So even if the batteries has x mAh you may only get half of it while the voltage is above the required level.

With multiple LEDs you have options to optimize your power in case you cannot get a large enough battery inside the space you have. Of course this depends on your application. Think of what smartphones and laptops do .e.g. low power mode when the battery drops below x % e.g. dimming brightness or switch some LEDs off ...

1 Like

Thanks for replying. Does it make any difference in terms of power dissipation (heat), if I attach a 14.8V battery source (4x 3.7 18650) versus e.g. 4.5V battery source (3x 1.5V AA)?

I was looking for a step-down converter and came across this cheap MP2307-based 3A buck converter module. However, according to this article its efficiency is very bad, meaning it consumes up to 60mA by just sitting idle. What would be a reasonable alternative?

So assuming I have a proper step-down converter that provides 5V output voltage, the following diagram should work, right? The Nano's internal DC-DC converter should further step-down the voltage to 3.3V, where as the NeoPixel consumes the 5V...

For all the datasheets I've read, the 60mA per unit sounds right, x16 is less than 1A.

The on-board power supply might be able to handle that (it being the MP2322 (datasheet Nano BLE) or MP3610 (mentioned above)).

The Nano BLE doesn’t have 5V on board (even worse: the USB-part of the NINA needs 5V from the other side).

I doubt the Ring will work properly on 3.3V, but it's worth a try (the 2322 and 3610 are both thermally protected). If this fails, you'll need a step-down converter (I found DFR0831 at dfrobot.com, you might look at that) (you might as well try the 'crappy' one you already have, might be good enough).

As for the data connection: data sheets often suggest a resistor between your output and the input of the LED-chip (470R).

Anyway, if the Ring is supplied with 3.3V, I expect the signal to work fine.

If the Ring has 5V, such is not guaranteed, but: you might just try it, it can't do any harm.

If it doesn't work, a voltage adapter takes only a PNP-transistor and 2 resistors (one resistor from output to base of the transistor (emitter to ground of course) one resistor as pull-up from collector to 5V, collector to input of the Ring (I’d suggest both 1k but it's not critical)).

Most optimistic scenario: Battery to Vin, 3.3V out to V+ of the Ring, output to input Ring (mind the series-resistor).

Connecting the USB shouldn't affect this set-up, nice for testing.

Most evolved scenario: Battery to Vin (Nano board using it's own converter) and to step-down converter powering the V+ of the Ring, with signal adapter.

This set-up even allows you to use all 4 Rings simultaneously (at the cost of battery life).

I will calculate with the here above mentioned 4x 18650 Li-ion batteries (3Ah), giving 3Ah @14.8V (not 12 Ah!!)

The Nano board is supposed to consume 15mA @ 3.3V, so less than 4mA from the battery.

One Ring at the above mentioned average of 20mA x16 = 320mA @ 5V.

Let's assume an efficiency of 75%, @14.8V it comes to 145mA from the battery.

Adding up to 150mA, it gives you 20 hours of battery life.

Using connectivity of the Nina will consume more current, but it's negligible to the Ring's power need.

Addition: I'm assuming Sj4 (marked 3.3V) is connected.

I recommend you look into series and parallel battery configurations. This combines cells in series to increase the voltage to an optimal level and at the same time parallel to increase the ampere-hours. You can find lots of information about this using Google. And likely you can find a pack that has the configuration and size you need.

Primary batteries e.g. most 1.5V are most likely not a good choice. Their voltage drops over time much more than rechargeable which tend to have less drop over time with a sharper drop at the end when the battery is nearly empty.

There are thousands of power conversion circuit designs. Many engineers spend their life developing them. Each has a working point where the loss is minimal. You will need to decide what requirements are important to you. e.g. loss in standby mode is only important if the circuit needs to be powered while it is not fully active. If you disconnect the system anyways you do not need to worry about that.

1 Like

Why do you want AA-cells? Didn't you need 5V? (Not to mention the dropping of the voltage during use.)
And: with luck they are 3000mA too, 3 pieces make 3 Ah @ 4.5V = 13.5Wh. Compare this to those 4 18650 cells, 3Ah @ 14.8V = 44.4Wh. Even if you could use all the power, you are down to 6 hours.

Is the following example correct?

  • 2x 3.7V 3000mAh in series = 7.4V 3000mAh = 22.2Wh
  • 2x of these packs in parallel and I'll get 7.4V 6000mAh = 44.4Wh

Yes

What?

If "these packs" mean the same thing in both examples, then you need 4 packs to achieve this second combination, as two 3.7 V batteries in parallel will give you 3.7V at 6000mAh which is still 22.2 Wh

I’m assuming you are putting time and effort in selecting converter boards, series parallel connections of batteries and such. This is highly interesting stuff, I’ll agree to that. But if the bottom line is the performance of your project, I’ll give you some thoughts:

If you raise the efficiency of the conversion from my (pessimistic) estimation of 75% to say... 85% you’ll gain (roughly) 2 hours on the 20 hours I calculated. And you needed only 14 hours.

In the converter the difference between in and out (i.c. 25% or 15%) is turned into heat. If the manufacturer specifies a certain max output current he must have calculated the worst case heat production and designed the unit to ‘take the heat’.

In general it is a bad idea to connect batteries in parallel. It is possible, but you have to make sure they stay connected all times to guarantee they have the same voltage, the same level of charge.

The brightness of the LEDs has more impact on battery life than anything else.

1 Like

What makes you think that?
These things a put together with a marketing overview. It is certainly not the case with data sheets of FETs. They boast maximum currents that can never be achieved in practice.
Sure there are test conditions when these claims can be justified for example in very short pulses, like 1uS every second.

So you have to look very carefully about maximum claims in these situations. I would suggest that the input and output voltages have been optimised to make the best of these figures.

Another common practice is to assume an infinite heat sink. This is one that does not rise in temperature no matter what heat is pumped out of the device. An infinite heat sink can be approximated by forced cooling of a heat sink to maintain its ambient temperature.

Another common technique is assuming that the maximum junction temperature of the device is the one it is running at. A typical maximum junction temperature is 130˚C In practice you would not run a junction temperature that hot.

For every 10˚C cooler you can run the junction temperature the reliability ( time it takes half the devices to fail ) approximately doubles.

If I buy a module (be it a converter board or a USB charger) the manufacturer has printed on it something like: 5V 0.5A. If the module can’t live up to it’s moniker, the manufacturer is a fraud and by law liable to prosecution. (And some manufacturers make it a close call.)

You are right about reliability. It is wise not to use a module or a component up to it’s absolute maximum rating, safety margin is good.

The tests on individual components you refer to usually also show longer periods up to DC. These are done on infinite heatsinks. They give you proper information if you know how to read them.

This topic was automatically closed 180 days after the last reply. New replies are no longer allowed.