Official Raspberry Pi Page
https://www.raspberrypi.org/products/pi-zero/
Core Differences To The Raspberry Pi 2
Broadcom BCM2835 (as in Raspberry Pi 1, but @ 1GHz) single core processor instead of newer quad core BCM2836 (RPi 2).
512MB instead of 1GB RAM.
Processor
Broadcom BCM2835. This contains an ARM1176JZFS (ARM11 using an ARMv6-architecture core) with floating point, running at 1GHz, and a Videocore 4 GPU.
Memory
512MB LPDDR2 SDRAM
Connections
USB On-The-Go port
Mini HDMI
40pin GPIO header with identical pinout to Model A+/B+/2B
Unpopulated composite video header
CSI camera connector (newer version from May 2016)
LED
A single green LED. Normally on, turns off to indicate disk activity.
Power
The Zero uses similar power as the A+
Our tests using fresh installed Raspbian OS
HDMI + USB hub connected (no USB peripheral load)
Max current during boot up to command prompt (GUI not loaded): 197mA
Idle at command prompt (GUI not loaded): 95mA
Nothing connected other than power input:
Max current during boot up to command prompt (GUI not loaded): 185mA
Idle at command prompt (GUI not loaded): 82mA
Nothing connected other than power input, HDMI port and LED disabled (see here)
Max current during boot up to command prompt (GUI not loaded): 185mA
Idle at command prompt (GUI not loaded): 63mA
Powering with less than 5V
The RPi zero runs at 3.3V, there is an on board voltage regulator which creates this voltage from the 5V power input. Whilst the Raspberry Pi is generally specified as operating from 4.75V to 5.25V (the USB acceptable voltage range), the processor runs at 3.3V. No schematic is available at the current time for the RPi zero, but as it is simply just the processor and SD card, which both run at 3.3V, there should be no reason not to be able to power it with a voltage below 4.75V. It should be fine to power it with 3.3V directly via the IO connector 3V3 pin (we haven't tried, but no reason not to), or use its voltage regulator to create the 3.3V with an input voltage below 4.75V. The running current will be higher than at 5V as its voltage regulator is switch mode, but the voltage regulator should still do its job OK as long as there is a bit of voltage headroom to its Vout (3.3V).
Powering with 3.5V, HDMI disabled, USB hub connected
Powers up fine. 3.3V rail looks fine on an osciliscope.
RJ45 USB ethernet adaptor in the USB hub (powered by the USB hub not the RPi) works fine, SSH connection fine.
Max power consumption during boot: 268mA (this was just by watching PSU display readings by eye, most of the time was < 180mA but peeked at times – using an inline max/min multi-meter resistance caused too much voltage drop)
Power consumption idle: 98mA
Power consumption after sudo shutdown completes: 22mA
Powering with 3.4V, HDMI disabled, USB hub connected
Powers up fine. 3.3V rail looks fine on an osciliscope.
RJ45 USB ethernet adaptor in the USB hub (powered by the USB hub not the RPi) works fine, SSH connection fine.
Power consumption idle: 100mA
Power consumption after sudo shutdown completes: 22mA
Powering with 3.3V, HDMI disabled, USB hub connected
Powers up fine. 3.3V rail has a bit of wobble, min 3.16V and typ 3.29V on an osciliscope.
RJ45 USB ethernet adaptor in the USB hub (powered by the USB hub not the RPi) works fine, SSH connection fine.
Power consumption idle: 101mA
Power consumption after sudo shutdown completes: 22mA
3 years ago
The opposite question: I am interested in knowing how high you can go with an RP0’s input voltage – I want to power it using 4 x 1.2V NiMH batteries, nominally providing 4.8V but when they are fully charged these batteries have a higher voltage which (x 4) could go beyond the 5.25V spec – since the RP0 has a regulator would it cope with e.g. a 5.5V input voltage? Might it cause problems with the Wifi adaptor?
3 years ago
I just tested with 3.3v and there is a problem on PiCamera. The moment pi camera running for a while even voltage is stable at 3.3v it will hang. Unless you are running the PiZero without the PiCamera and just plain PiZero then should be working fine.
4 years ago
Just throwing my two cents in here, I was very skeptical of this at first. I thought there was no way the processor could run at 3.3V with the drop-out of the regulator. I almost agreed with Anonymous’ comment on here that the USB hub must be powering the Pi through backfeed since that’s a common issue with cheap hubs, but then I realized that would make no sense because the 3.3V supply and 5V supply would be shorted together and at the very least that would cause incorrect current readings but everything seemed normal here.
So, I tried it on my Pi Zero W. Had to disable HDMI and WiFi can’t be used below around 4.5V so I disabled it as well. However, directly connecting to the GPIO headers with an FTDI board, I was able to power the Pi off of voltages even lower than 3.3V, albeit unstably. I was powering it through the regulator still as well, not bypassing it like another commentor mentioned. The drop-out is just surprisingly low.
I was able to get down to 3.0V on the 5V line (Running `top` so I could see the Pi hadn’t died and was still updating) before it cut out entirely. It’s very unstable at this voltage though. For example, while I could navigate directories at 3.1V, starting something more intensive like the interactive Python 3 shell would crash it. I wasn’t able to do anything to cause the Pi to crash at 3.3V, but that still doesn’t mean it’s entirely stable at that voltage either. I feel like this is more of a technical novelty than anything, as it wouldn’t be wise to connect the Pi directly to a Li-Ion cell, at the very least because of the inability to use the full capacity of the battery.
Also, one more thing to note, some lower quality SD cards may have more issues than the ARM CPU at lower voltages as well that may be hard to trace. I noticed read errors before the CPU cut out at some voltages so that may be the first thing to fail.
4 years ago
My guess is the powered USB hub is providing the 5v for the RPi to operate correctly, since the power rail is the same for the power port and the usb port.
4 years ago
Pi Zero uses a dual regulator that has 3v3 and 1v8 outputs. The processor uses both, so I don’t see how the processor could operate if you bypass the voltage regulator by feeding 3v3 directly as described here.