> Right now the battery-free phone needs a custom base station to transmit and receive calls. But the team says there's no reason why the technology couldn't be integrated into standard mobile network infrastructure.
This could only be called a cell phone, if the cell phone network changes to accommodate this radio. It's not a cell phone, because it doesn't work with cellular phone towers.
I agree calling it a cell phone sets the wrong expectations, but it's still neat and I think a good application could be an elder care pendant (help, fallen, can't get up, yadda yadda) where close promity and custom base station aren't an issue, but never needing to worry about the status of battery would be a big plus.
Those applications are called medical alert systems, and there are various kinds. Some of the systems contain GPS, while other work with a radio base station to inform if the person has fallen and can't get up.
Modern Cellphones and not compatible with all cellphone networks, making your argument moot. Anyway, Cellphone just means it can handle hand offs between base stations without dropping the call, which this could do.
That's silly. Modern cellphones don't need to be compatible with all cellphone networks to be cellphones any more than a shoe needs to fit on every foot in America to be a shoe.
> Anyway, Cellphone just means it can handle hand offs between base stations without dropping the call, which this could do.
No. It literally cannot do that, hence my comment.
It's a cellphone prototype. You don't need to do 100% of the functionality to demonstrate functionality.
But you may have missed this part "or ambient light collected by a solar cell the size of a grain of rice". So, yes they built a cellphone, even if they can't gain power from arbitrary sources.
It has none of the qualities of a cellphone other than being a radio, so it is not a cellphone prototype. It is a passively-powered radio prototype. They may later build a cellphone prototype, but this isn't such an animal.
> But you may have missed this part "or ambient light collected by a solar cell the size of a grain of rice". So, yes they built a cellphone, even if they can't gain power from arbitrary sources.
A cellphone is not "a radio that cannot connect to a phone network and that uses a solar cell". That would be a solar-power radio.
Look I like low power devices. Anyway, the base station does not say interpret voice data from a radio and then decode that to make a call. You have an actual touch pad to use to dial so it handles the dialing part of making skype calls. But yea it's not in any way running skype, but that's > pure radio functionality thus what I said is literally true.
There are a ton of things this device does that are cheesy, because they are trying to push the limits on low power. But they could have jacked up the received power by 30x and fooled most people reading, but they actually did create something novel by pushing almost all functionality to the base station.
It the use case is IMO not general calls it's more like a limited call 911 when battery is dead mode that could still be useful.
Debatable. I'd say it's a cellphone for which no infrastructure currently exists, which admittedly doesn't make it a very useful cellphone unless and until those changes occur. But a car is still a car even if there's no gas stations.
I'm not even sure if multiple of these devices could be used simultaneously like traditional cellphones, which change frequencies/channels as they move between cells and to avoid collisions with other mobiles. Judging by their description, it works on a fixed frequency and the "phone" itself has basically no active circuitry, being nothing more than an analog audio-to-RF modulator.
So this uses vibrations caused by our voice to change the effectiveness of the antenna to absorb power from the base station right? Base station then interprets the varying current draw as amplitude modulated audio and Bob's your uncle? If I'm correct (no sure that I am) wouldn't this require being pretty close to the base station?
Hehe that would work, and is in store. I had something a little more subtle in mind. I remember Seiko selling "Kinetic" watches with internal counter weights. Maybe it's possible to scale up a similar mechanism in a phone case (granted you accept the doubled width) and enjoy the free energy. Of course this came to mind when running yesterday.
Youtube has some videos about a thing called ampy, a tiny case you attach on you. I'm not sure it made it to market though.
Even if it was transmitting instead of using a side-channel like current draw... With only 3.5 microwatts, the phone would still have to be very close to the base station. As a comparison, "normal" cellphones transmit something around 100 to 1000 milliwatts, so at least 10^5 more powerful.
EDIT: This Wired article has a lot better information:
"... Some key components of Talla's phone are housed remotely to save power. A nearby basestation has circuitry for converting and connecting to the digital cellular network, currently via Skype. The prototype basestation uses an unlicensed frequency, limited to low-power transmissions. Because the phone relies on those signals for its energy harvesting, it has a range of just 15 meters from the basestation."
And it looks like they use backscatter, which would be "transmitting" signal and not using a side-channel.
When tech writers write about things they don't understand, they tend to omit all the caveats, like "cripplingly short range" and "probably illegal to operate as a consumer device due to RFI".
Probably legal due to the frequencies involved, but also useless in anything but a completely radio-quiet environment (the ISM bands are specifically allocated to common interference sources --- and the only reason other wireless devices like WiFi and BlueTooth can survive there are due to processing techniques that can transmit/receive below the noise floor, like spread spectrum.)
Even if it operates in the unlicensed freq. bands, ie ISM bands, there are still limitations to how much you may interfere with others, expressed in eg duty cycle requirements (ie not sending too long at a time) and energy levels (ie not sending too strong). That's what the FCC/ETSI/CE etc etc regulations do.
Wait. Depending on how loud this thing can go, it could be 100% effectively pivoted as "world's first battery-free wireless headphones". Start a company selling that and use it to fund R&D to remove the PTT switch from the communicator design etc.
Is this design analog or digital? If it's analog, all those issues are replaced by static noise, which could very probably be largely mitigated by amplification power and possibly audio processing on the transmitting side.
I'd like to try to make a mobile device like this that can run with ~100mA input power from a solar cell. I thought about using the Particle Photon & an E-ink display, but my knowledge of hardware isn't the greatest. Any EE out there think it's possible?
It would be a bit of a learning curve, but certainly doable.
Especially if you find a schematics that you want to assemble and try to follow the instructions and assembly. It can be like a giant puzzle with too many rules. But doable. If not on the first try, then on the subsequent ones :)
How large of a device do you have in mind? 100 mA is more than enough power for the Photon module, but would require a relatively large solar cell -- probably at least 3 to 4 inches square under ideal conditions.
That's roughly the size I was thinking. I was thinking of adding a battery too, less as a storage bank and more as a way of papering over the ideal conditions issue.
Have those graphene supercapacitors become commercially available? If so, you may be able to do without a battery at all. Well, depending on use, available charge, and efficiency.
That's mostly radio power, right? To compare to the system in the article, we don't need to have our signals reach normal towers. We only need to reach a femtocell, and those can be as low-power as 10mW. For example a quick search found the ADF4602, a 3G transceiver that runs fine at 50mA. And that's transmit/receive power; a good 3G+ receiver can go idle between packets.
Such a solar panel is quite large/powerful and you can drive quite a lot on that, depending on how and where you deploy it. One of the most important factors on how to make it live longer is to limit the steady-state static draw (ie ensure you go to sleep modes properly etc). Plug, but _just now_ our related post got to the first page: https://news.ycombinator.com/item?id=15354741
The steady-state static draw is the current consumption that is always there even if nothing is going on. Everything else that happens is added on top of it, so to speak. When the microcontroller is at sleep it will consume a certain amount to current to eg keep registers and RAM intact (otherwise it would become corrupt).
It is also due to leakage currents from capacitors, and any leakage that might go out through GPIOs. Sensors also have a static draw when not in use. Batteries self-discharge. And so on.
For example, a cc1310 system on chip has a 0.8uA current consumption when in deep sleep, but the real static draw is more than that. Periodically, the chip will run a little charge pump to energize RAM so that it doesn't evaporate (metaphor). There might also be a timer active, and that consumes 100uA. And so on.
When the device wakes up to run at 24 MHz, it uses about 1.9mA, plus any peripherals (SPI 13+93uA, timer 113uA), plus any sensors or other hw peripherals. This might be very very short amount of time, like 10 ms every two hours-kind of short.
So it turns out, that for a device that wakes very rarely, the static draw can be a significantly decrease lifetime. It's small, but it is _always_ there. It is hardware-dependent and is the "black art" within embedded development. It's very easy to reach arduino-levels, pretty easy to start using low-power modes, time/money-consuming and requires expertise to squeeze the last years/months out of the battery.
Oh man thank you very much I' just starting to get into hardware/embedded systems coming from web and this is so cool.
I don't know if this is uncool to ask I'm also late a few days to respond.
With regard to connection to the web generally (to my knowledge) you need some way to communicate say send a POST request to some API and I'm not sure if this means that you need an OS which when I think of an OS you need to keep it on or even with sleep... I don't know. I don't know if you can design hardware/firmware meant to interface with the web, still be programmable and simple/light on resource without a full blown OS like raspberry pi.
Yeah, connecting to the web can be stressful to a small device, but it's very feasible. A basic OS that handles timers, a communication stack, a way to abstract tasks and communicate between them, can be quite small. Contiki OS for small devices can run on tiny devices - the amount of RAM and flash and CPU speed isn't much a limiting factor until you start with communication. Then you want RAM to hold queues of packets, neighbor tables, etc. You want flash to hold the code necessary to use the protocols you need. We support a range of devices at Thingsquare, eg the cc1310 with 128kB flash and for that we have space for a full-fledged IPv6 stack and all that's necessary for low-power wireless (ie specific radio duty cycling protocol, routing protocol, etc), our own libraries and abstractions to make it simpler for customers to use the system, with room left for customer application and drivers.
Contiki OS is event-based, ie when it is not actively doing something, it simply waits (for something to happen on a physical pin, or radio event, or timer expiry) and can go to sleep. That way, and with lots of hard work to make everything efficient on every level, you save power. Raspberry pi and similar ones makes it simpler (lots of developers know the stuff, linux, etc) but has drawbacks (power, complex system, lack of low-level control and kernel timing requirements might not suffice, cost, IP, in some cases security, licensing, etc).
Thanks a lot for the info/resources a lot of this not familiar to me at this time. The queues if packets/neighboring tables,etc...
Yeah it's much easier from my lerspective/experience tojust throw on Python/PHP on there and do stuff granted that's more "webserver" I didn't mention databases but you could just write to filesystem (txt) and my intent was transmission.
Apple is building this so they don’t have to give you a battery at all.
I’m always confused by people complaining about iPhone battery life. My wife’s iPhone 6 lasts days, it’s called a battery pack. And she gained this incredible battery life without forcing me to carry a heavier bulkier iPhone. That’s what Apple understands, making their devices heavier and thicker to please a minority of users while inconveniencing the majority makes no sense when battery packs exist.
Related, this spying microphone bug first used by the Soviets in the 50s that has no battery or electricity wire and can keep listening "forever", which was discovered by accident in a "gift" eagle sculpture given to the US ambassador, after years of staying in that embassador's office, listening https://en.m.wikipedia.org/wiki/The_Thing_(listening_device
This has been posted here earlier and HN couldn't get over how it's not perfect, just like Atom. HN isn't the audience for "big-picture" technologies that will get more efficient over the years and have broad applications eventually.
The low power usage of the device is impressive (I think), but the battery free part not so much. It seems as though this is just the modern day version of the crystal radio sets many of us built (or tried to) when we were kids.
My dream phone would have a small e-ink display that can maybe read wikipedia or text based websites, simple physical buttons, and a 5000 mAh battery, maybe even some low discharge chemistry instead of lion, shooting for a charge maybe once a month.
Naturally GPS would be right out...
Pretty sure the potential market size is in the single digits, but I'd be happy.
> The biggest step was eliminating the analog-to-digital converter that turns your voice into data. Instead, they took advantage of tiny vibrations that happen in a microphone when a person is talking into it, or a speaker when someone is listening. An antenna converts that motion into radio signals in a way that uses almost no power.
Does it mean the voice is transmitted in plaintext using analog modulation of (reflected) carrier wave? Privacy concerns aside, how can it handle calls from multiple stations over same frequency range?
Voice is always transmitted in a "plain text" is it radio or phone except if you use some special device (it's never easy to get one) and protocol (good luck to access it for cheap).
It runs on 3.5uW, and the homemade cellphone I built last year with off-the-shelf components drew 2A when syncing with GSM antennas.
Some people don't realize it, but GSM is very expensive power-wise, especially on boot when it has to find antennas. Having a device that can be solar powered or even with ambient radios is extremely useful!
Right now the battery-free phone needs a custom base station
to transmit and receive calls. But the team says there's no reason
why the technology couldn't be integrated into standard mobile network infrastructure
Well, there's the trick. But an awesome proof of concept nonetheless
I'm a bit puzzled by the statement “But the team says there's no reason why the technology couldn't be integrated into standard mobile network infrastructure”
There's obviously many reasons. Transmission range, security, privacy, reliability, ...
Is this sort of passive device practical for applications such as wireless mouse and keyboard? It would seem like a great selling point (similar to Logitech's attempts to make solar powered peripherals)
81 comments
[ 5.1 ms ] story [ 137 ms ] threadhttps://news.ycombinator.com/item?id=14659236
https://news.ycombinator.com/item?id=14716986
> Right now the battery-free phone needs a custom base station to transmit and receive calls. But the team says there's no reason why the technology couldn't be integrated into standard mobile network infrastructure.
This could only be called a cell phone, if the cell phone network changes to accommodate this radio. It's not a cell phone, because it doesn't work with cellular phone towers.
Those applications are called medical alert systems, and there are various kinds. Some of the systems contain GPS, while other work with a radio base station to inform if the person has fallen and can't get up.
For example, http://www.best10medicalalertsystems.com/
> Anyway, Cellphone just means it can handle hand offs between base stations without dropping the call, which this could do.
No. It literally cannot do that, hence my comment.
Of course it could be changed into a cellphone!
The device on this page is billed as a cellphone, but it is not a cellular device. It is not even a phone.
Of course it could be made into a cellphone, just as my shoe could be made a cellphone. In the interim, neither my shoe nor this radio is a cellphone.
But you may have missed this part "or ambient light collected by a solar cell the size of a grain of rice". So, yes they built a cellphone, even if they can't gain power from arbitrary sources.
> But you may have missed this part "or ambient light collected by a solar cell the size of a grain of rice". So, yes they built a cellphone, even if they can't gain power from arbitrary sources.
A cellphone is not "a radio that cannot connect to a phone network and that uses a solar cell". That would be a solar-power radio.
But, this also handles dialing to make Skype calls so it's not just a radio.
Wrong. This communicates with a base station that made skype calls. This is a radio.
You sound as if you have a vested interest in continuing to mislabel this device.
There are a ton of things this device does that are cheesy, because they are trying to push the limits on low power. But they could have jacked up the received power by 30x and fooled most people reading, but they actually did create something novel by pushing almost all functionality to the base station.
It the use case is IMO not general calls it's more like a limited call 911 when battery is dead mode that could still be useful.
A cellphone that doesn't work with cellular network cells and isn't a phone really shouldn't be called a cellphone. It's a passively-powered radio.
I'm not even sure if multiple of these devices could be used simultaneously like traditional cellphones, which change frequencies/channels as they move between cells and to avoid collisions with other mobiles. Judging by their description, it works on a fixed frequency and the "phone" itself has basically no active circuitry, being nothing more than an analog audio-to-RF modulator.
3.99 with free shipping.
Youtube has some videos about a thing called ampy, a tiny case you attach on you. I'm not sure it made it to market though.
EDIT: This Wired article has a lot better information:
https://www.wired.com/story/this-cell-phone-can-make-calls-e...
"... Some key components of Talla's phone are housed remotely to save power. A nearby basestation has circuitry for converting and connecting to the digital cellular network, currently via Skype. The prototype basestation uses an unlicensed frequency, limited to low-power transmissions. Because the phone relies on those signals for its energy harvesting, it has a range of just 15 meters from the basestation."
And it looks like they use backscatter, which would be "transmitting" signal and not using a side-channel.
Do you know any off the self components which can fully operate on 100mA?
Don't you need to keep draining power produced by solar panels? What about the heat problem and decreasing efficiency of solar panels?
It is also due to leakage currents from capacitors, and any leakage that might go out through GPIOs. Sensors also have a static draw when not in use. Batteries self-discharge. And so on.
For example, a cc1310 system on chip has a 0.8uA current consumption when in deep sleep, but the real static draw is more than that. Periodically, the chip will run a little charge pump to energize RAM so that it doesn't evaporate (metaphor). There might also be a timer active, and that consumes 100uA. And so on.
When the device wakes up to run at 24 MHz, it uses about 1.9mA, plus any peripherals (SPI 13+93uA, timer 113uA), plus any sensors or other hw peripherals. This might be very very short amount of time, like 10 ms every two hours-kind of short.
So it turns out, that for a device that wakes very rarely, the static draw can be a significantly decrease lifetime. It's small, but it is _always_ there. It is hardware-dependent and is the "black art" within embedded development. It's very easy to reach arduino-levels, pretty easy to start using low-power modes, time/money-consuming and requires expertise to squeeze the last years/months out of the battery.
I don't know if this is uncool to ask I'm also late a few days to respond.
With regard to connection to the web generally (to my knowledge) you need some way to communicate say send a POST request to some API and I'm not sure if this means that you need an OS which when I think of an OS you need to keep it on or even with sleep... I don't know. I don't know if you can design hardware/firmware meant to interface with the web, still be programmable and simple/light on resource without a full blown OS like raspberry pi.
Anyway thanks much to learn.
Yeah, connecting to the web can be stressful to a small device, but it's very feasible. A basic OS that handles timers, a communication stack, a way to abstract tasks and communicate between them, can be quite small. Contiki OS for small devices can run on tiny devices - the amount of RAM and flash and CPU speed isn't much a limiting factor until you start with communication. Then you want RAM to hold queues of packets, neighbor tables, etc. You want flash to hold the code necessary to use the protocols you need. We support a range of devices at Thingsquare, eg the cc1310 with 128kB flash and for that we have space for a full-fledged IPv6 stack and all that's necessary for low-power wireless (ie specific radio duty cycling protocol, routing protocol, etc), our own libraries and abstractions to make it simpler for customers to use the system, with room left for customer application and drivers.
Contiki OS is event-based, ie when it is not actively doing something, it simply waits (for something to happen on a physical pin, or radio event, or timer expiry) and can go to sleep. That way, and with lots of hard work to make everything efficient on every level, you save power. Raspberry pi and similar ones makes it simpler (lots of developers know the stuff, linux, etc) but has drawbacks (power, complex system, lack of low-level control and kernel timing requirements might not suffice, cost, IP, in some cases security, licensing, etc).
Have a look at these: http://www.thingsquare.com/blog/articles/sensortag-power/ and http://www.thingsquare.com/blog/articles/ti-cc1350-power-con...
Yeah it's much easier from my lerspective/experience tojust throw on Python/PHP on there and do stuff granted that's more "webserver" I didn't mention databases but you could just write to filesystem (txt) and my intent was transmission.
Contiki OS sounds cool
Anyway I'll refer back to this post, thanks.
I’m always confused by people complaining about iPhone battery life. My wife’s iPhone 6 lasts days, it’s called a battery pack. And she gained this incredible battery life without forcing me to carry a heavier bulkier iPhone. That’s what Apple understands, making their devices heavier and thicker to please a minority of users while inconveniencing the majority makes no sense when battery packs exist.
Which makes this all the more impressive!
The Thing actually needed power to operate, btw. The power was just beamed in externally.
I could use this in my spy operations though. Not that Im a spy or anything.
Definitely, not a spy.
Naturally GPS would be right out...
Pretty sure the potential market size is in the single digits, but I'd be happy.
Does it mean the voice is transmitted in plaintext using analog modulation of (reflected) carrier wave? Privacy concerns aside, how can it handle calls from multiple stations over same frequency range?
It runs on 3.5uW, and the homemade cellphone I built last year with off-the-shelf components drew 2A when syncing with GSM antennas.
Some people don't realize it, but GSM is very expensive power-wise, especially on boot when it has to find antennas. Having a device that can be solar powered or even with ambient radios is extremely useful!
Well, there's the trick. But an awesome proof of concept nonethelessThere's obviously many reasons. Transmission range, security, privacy, reliability, ...
Can you burst with capacitors or no? (Not a hardware person)
I'd love a device that ONLY does sms or text based comms (fb, whatsapp etc..) but only the text. and nothing else.
that would sip battery.