Computing stuff tied to the physical world

Tiny Lithium – packet test

In AVR, Hardware on May 2, 2012 at 00:01

This post continues where the Tiny Lithium discharge post left off…

To summarize: a tiny ML614 rechargable Lithium cell of a mere 3.4 mAh is used to power a JeeNode running the radioBlip sketch, which is going to send out one small packet per minute (the period is actually slightly longer).

I charged the battery overnight from a 3.0V power supply and through a 1 kΩ resistor, as described in the datasheet. As expected, the battery voltage without load is now indeed 2.97V

The test is really simple, but it’s going to take a little while: connect, see packets come in, with a counter being increased for each packet, and then just wait for the whole thing to stop sending. Here we go:

DSC 3072

The range will need to be good, since packets have to cross a reinforced concrete floor to reach my receiver.

Here’s the log of packets I got (the first packet seems to have been missed):

L 12:08:51.628 usb-A600dVPp OK 17 2 0 0 0
L 12:09:54.592 usb-A600dVPp OK 17 3 0 0 0
L 12:10:57.556 usb-A600dVPp OK 17 4 0 0 0
...
L 23:58:57.898 usb-A600dVPp OK 17 167 2 0 0
L 00:00:00.777 usb-A600dVPp OK 17 168 2 0 0
...
L 23:59:32.169 usb-A600dVPp OK 17 6 8 0 0
L 00:00:34.988 usb-A600dVPp OK 17 7 8 0 0
...
L 23:58:57.821 usb-A600dVPp OK 17 101 13 0 0
L 00:00:00.609 usb-A600dVPp OK 17 102 13 0 0
...
L 23:59:01.058 usb-A600dVPp OK 17 197 18 0 0
L 00:00:03.818 usb-A600dVPp OK 17 198 18 0 0
...
L 23:58:45.821 usb-A600dVPp OK 17 37 24 0 0
L 00:00:51.318 usb-A600dVPp OK 17 39 24 0 0
...
L 13:18:55.865 usb-A600dVPp OK 17 34 27 0 0
L 13:19:58.670 usb-A600dVPp OK 17 35 27 0 0
L 13:21:01.474 usb-A600dVPp OK 17 36 27 0 0
L 13:33:35.035 usb-A600dVPp OK 17 48 27 0 0
L 13:37:46.210 usb-A600dVPp OK 17 52 27 0 0
L 13:43:00.171 usb-A600dVPp OK 17 57 27 0 0
L 13:56:36.450 usb-A600dVPp OK 17 70 27 0 0
L 13:57:39.232 usb-A600dVPp OK 17 71 27 0 0
L 14:06:01.520 usb-A600dVPp OK 17 79 27 0 0
L 14:20:40.504 usb-A600dVPp OK 17 93 27 0 0
…silence…

As you can see, this “battery” packs enough energy to send out 27 x 256 + 93 = 7005 packets – of which 398 were not properly received (mostly in the last days). IOW, it was still sending after 5 days!

The node kept going even though several of the last packets were not received – so transmit power probably dropped off quite a bit, but the sketch was still running properly until the very end.

Note: once moved to measure the supply voltage (still about 2.5V), packets started coming in again – perhaps because of improved reception from the new location, or a slight warm-up of the battery.

This little cell just doesn’t want to give up!

  1. Nice result, do you have any idea how the radioblip sketch compares to the roomnode without a PIR?

    Wonder how it performs when charged by a solar cell ;).

  2. Following from your previous post, an ATmega328P will actually boot with 2.97V at 16MHz? Is the advantage of booting at 16MHz that the Optiboot is still the default version?

    It is also interesting to note that those batteries are rated for 3.4mAh at 15uA with a max rated current draw of 250uA. The transmitter is pulling 100x their rated current! At higher current draw (100uA) their storage is only rated at 2.5mAh. It is amazing they lasted 7005 minutes. I wonder how low the power droops during transmit. In my tests with a 100uF capacitor at 3.3V, a 3ms transmit would deplete ~0.75V from the capacitor.

    (Your link to the previous article is pointing to -lithium- and should be -lipo-)

  3. So it uses less than 1 mAh per day, which is 42 microamps (averaged). Impressive! Maybe that tiny solar cell really would work.

Comments are closed.