OT. Dirt Powered Fuel Cell

[Dean]

[T]

Something does not add up. I think we've been punked.

[rbowman]

I think it's possible. There are billions of bacteria chomping away

The technology has been around,

For sensors some of the microcontrollers have very low power requirements. Typically the expensive part in terms of power is communications but there are low power devices available for that too.

you're talking microamps on stadby and it wouldn't be transmitting continuously, only when there was data of interest to upload to a node with a better power supply.

You're not going to power your Tesla but that's not the point.

[T]

Sort of like the science experiment where you poke probes into a potato and an onion.

[Snag]

No , more like a sensor telling your irrigation system that sector

4A3 has low soil moisture and needs water . Only it's this device sending this info to a nearby computer (possibly thru an intermediate processor/amplifier) that controls the irrigation system .

[Dean]

But the unit wouldn't need to transmit continuously. All it has to do would be let the farmer know it's dry where the unit is. Once or twice a day would be often enough. Text message maybe? Farmers can use their cell phones to start pivots if the wells are on commercial power. I don't think I"d want the unit to start irrigating without me making the decision. What if rain is predicted? A 1300 foot long center pivot in my area of south central Nebraska typically makes a full circle in 22 hours or so running full speed. It would apply maybe 0.20" of water. Water use charts are readily available. People know how much water corn uses at various stages of growth.

[Dean]

I posted the same link at sci.electronics.design. The guys there don't seem to share your concerns. They got into a discussion of infinity.

[Mark Lloyd]

[snip]

I had a lemon-powered radio in the seventies.

[rbowman]

The Nordic Semiconductor chip I referenced sends a 32 byte string using the enhanced ShockBurst protocol. There are 125 channels so one master can listen for 125 slaves. You can do anything you want with the 32 bytes.

Where this gets to be interesting is with machine learning. I'm only throwing out some possible sensors, soil temperature, moisture, pH, air relative humidity and temperature, solar intensity, and so forth. These are factors a farmer might take into account when deciding whether to irrigate or not. A low cost and low power microprocessor can take the factors into account and send a status to the master rather than a stream of data to be processed upstream.

Chatbots are the headline AI applications but in the long run I see IoT as the area where the real impact will occur.

The first paragraph points out that 'internet' is misleading. Connecting to the internet is relatively expensive in terms of power requirements. Low power local networks require much less power which is extremely important when you're talking about remote monitoring that has to be powered by batteries or some other source like the fuel cell. You already see little solar panels in odd places but they have drawbacks.

[Bob F]

I had a lemon car in the seventies.

[Clare Snyder]

I can see this technology working for milli-watt sensor applications using micreo-watt bluetooth for communications. The power output will be significantly lower than a #10 hearing aid battery - which is

100mah - but will last an incredibly long time (hearing aid batteries do not - even with no load after activation they go deadin a matter of weeks,at best). Most hearing aids using a #10 battery will be drawing in the neighbourhood of 2.5ma - making them good for about 40 hours of use - or roughly 3 days at 13 hours per day. My hearing aids use #312 batteries, which are rated at 125mah? and fresh batteries last me roughly 5 days 16 hours per day - so draw about 1.5 or 1.6 ma at 1.4 volts - so roughly 2.2 mw. I would expect these "fuel cells" to be good for MAYBE 1/4 of that - being optimistic. - like .5mw (.0005 watt or .00035 amp) - and I mean OPTIMISTIC - quite possibly more in the Micro-watt/micre-amp range - like 1/1000 of the milliwatt/milliamps I'm talking - - - As noted by previous poster this draw would not necessarilly be constant - and the fuel cell could be charging some sort of supercapacitor that would enable pulse /burst transmission of data at significantly higher power outputs - loke running milliwatt transmitters on microwatt fuel cells.

[micky]

This is a great idea.

I'm working on a way to power electic appliances using ear wax.

[Xeno]

They were common in the 70s.

[Jim Joyce]

I saw a recipe that included navel lint and lemon juice, with liquid soap used as a binding agent.

[micky]

Damn. How am I supposed to compete with the really smart people.

Maybe I can find some way to use Guatemalan fruits or vegetables. A mixture of bananas, papaya, and coffee might make electricity directly.