One Pin To Rule Them All

When Maxim acquired Dallas Semiconductor, they took over the popular 1-Wire product line. These are sensors that get power and bidirectional data over the same pin. However, we never liked the name 1-Wire as you really need two wires: one for the power and data and, of course, a ground wire. A new startup company, Cyclopia, has announced their latest line of truly one pin CPUs, and we’re impressed. The low power system on chip devices multiplex data, power, clock, and ground on one wire.

A company spokesperson, [Star Lipfir], noted, “Our patent-pending technology uses two well-known effects. First, a FET gate doesn’t actually draw current but works on an electric charge. Second, capacitors store charge.”

Apparently, the external system drives the CPU by sending an intermittent ascending ramp voltage over the single wire. The charge enters both a ground storage capacitor and a power capacitor. When the voltage ramp reaches a certain threshold, a FET turns off, isolating the ground storage capacitor. The power capacitor continues to charge until there is enough potential difference between the two capacitors to operate the circuitry (see chart, right). The ramp then goes to zero which allows the CPU or the external system to communicate during the time before the next ramp begins.

[Lipfir] continues, “The great thing is that the ramp sets the clock speed and with one wire, you can connect as many CPUs together as you need. This really simplifies board layout and cuts down on traditional ground plane issues.” Of course, adding too many processors to the same line will make programming more difficult.

A company insider mentioned that as a homage to the Dallas tradition, they had considered naming the devices “0-wire.” However, they decided against it as it would conflict with the company’s upcoming line of true zero wire devices, due out about this same time next year.

91 thoughts on “One Pin To Rule Them All

    1. The new HaD tradition should be un-jokifying april fools pranks by making them actually work somehow. Is there a lawyer in the house? Somebody incorporate Cyclopia and draw up the patents. We have a year to get this done.

  1. In all seriousness, this isn’t much beyond what I saw many years ago, at an air traffic control RADAR site. The RADAR system was housed in a box about the size of a 20-ft trailer, sitting maybe a quarter mile from the runway at an Air Force base. The surprising thing about it, was that the only thing that connected it with the control tower was a piece of RG-8 coax. This supplied both power and synchronizing signals from the tower, where the power supply was, and in return, RADAR video back to the tower, where the display ‘scopes were.

    1. Any sizeable radar system probably has an integrated power generator. I doubt a single RG-8 coax could supply enough kilowatts to keep a radar system running without turning into smoke.

        1. I’m sceptical about that too. I would expect that such a RADAR would be in the kW power range which would be pushing it for an RG8. The other thing is that I’m surprised there is no redundancy, if your link is cut you loose the whole system.

          1. Well, we’re solidly into April 2 now, and still I say, I’m serious. What you may not understand is how much power is necessary for pulsed RADAR. I worked on long-range systems that operated with 2-5 megawatt pulses. That’s PEAK power. Due to the very low duty cycle (you send out a pulse of a few microseconds, you wait several MILLIseconds for the returns, for duty cycles in the 10^-3 range), the average power needed is low. The 250-mile systems I worked on used average powers of less than 10 kW For a 20-mile system at an airport, used only for controlling aircraift on approach and departure, keeping in mind the square-law power requirement for transmitted power, the average power is in the TENS of watts. I’ve seen RG-8 used for hundreds of watts.

          2. Sorry Jim, didn’t mean to come off as accusing you of not knowing what you were on about, just surprised about your claims and wondering how it could be so. Thanks for the information :)

  2. Hmm, fascinating thanks for podt, a two pin CPU should be feasible with sizable ram/earom ie power and ground pins but with IR receiver & transmitter embedded in the epoxy package that come out the side(s) maybe 4 or so and capacitive touch sensor on the top of the package ie on/off/hold etc with of course a visible led or two – all of which programmable via the cygnal method (pre silabs takeover). Still have heaps of the old silabs 8051 series with the very flexible cross point matrix io selector…

    Could end up with a utility type very small CPU that just runs off the DC buss but, does something useful with multispectral IR inside products with touch sensing and led indication :-)

    1. Back in the 60’s and 70’s there used to be a company called poly packs that sold surplus parts in little poly packs. If you got one of their floor sweepings packs you always got a few 1 terminal diodes that the big companies tried to keep secret. These were known as monodes. I have seen switching monodes, small and large signal monodes, zener monodes and of of course, early versions of LEM’s, light emitting monodes. And let’s not even get into the various types of binisters….

  3. Can’t we actualy do this using high frequencies and stray capacitance for ground? But i guess that would produce lots of EMI (depending on frequency needed). So probably wouldn’t be much more usefull than RFID / NFC technology. Except for the fact that you don’t need coils. Basicaly it would be a wired RFID :-D

  4. Not as impressive as the pin less mcu from fubar elec that has a nuclear reactor onboard for power and communicates through telepathy with the other components on your pcb.

    1. I never really got Maxwell’s demon. I’s just a really, really slow heat pump. Of course you can violate the laws of thermodynamics if you just ignore the part which consumes energy by calling it a demon and handwaving how it would work. What’s the use of that?

  5. This calls for one of those LED trick hacks, where you connect just one wire from a 9 Volt battery to the LED and leave the other leg hanging in the air, and the LED still lights up.

    Trick being that you sneak an enameled magnet wire inside the sheath of a multi-strand wire and run the return current through that.

  6. I’m assuming the 0-wire technology uses an in-phase beam to transfer the data across the quantum tunnel to the unicorn horn. The power is provided by a negative phase feedback when entangled quantum bits spin in the same direction.

  7. This is a great idea, but I think the new cpu’s may be a bit I/O limited.

    These technology races just never end well. In the 60’s with the cold war, I recall the soviets trying to one up the americans by taking matell’s so called “close and play” technology and applying it to television.

    1. Considering that 25 years ago, the idea of sending 600Mbps over a telephone line with just $20 worth of hacked Homeplug adapters would be considered so ridiculously silly, the idea that someone would eventually figure out how to send a few Tbps over a telephone line shouldn’t be dismissed right away.

  8. I know its a joke, but Dallas 1-Wire indeed only needs a data and ground connection, and has an internal capacitance to remain powered while data is active. That is to say the data and power pin are the same pin. There is a power pin as well, but it is not required for operation.

    1. Little known fact: the original name of the tech was “One wire for carrying both signal and power, and the other wire for ground.” They just shortened it.

      This is why you can’t let engineers name products. It wouldn’t even fit on the box!

  9. I started reading this, and before relasiing it was actually an April Fool begun thinking “sure, they’ve got a trick to make ground and signal and power all the same wire, but what would be the use, a CPU with just one GPIO soudns a bit pointless”

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.