Powering Devices within the Body



  • I think I brought it up. I guess I should stop doing that before someone loses a limb.

    Well, if induction charging is the way to go, how can we improve on it? Currently my main issue with it is size. We should definitely try to source or make smaller ones.

    I've been told that the coil can be used as an antenna for communications, would this be possible?
  • rdbrdb
    edited August 2012
    Induction charging doesn't need to be large.  The battery is probably the largest thing, but the circuitry can be relatively small, and perhaps the coil could just be coiled around the whole board.

    It's just that we first need to develop a solution, and then we can look into making it smaller.  I can help where necessary by designing and manufacturing smaller boards.

    You could modulate the field to transmit some kind of signal to the device, and decode the modulation at the device.  On the device end, you can use load modulation, which can be detected by the receiver module.
  • rdbrdb
    edited August 2012
    I just stumbled upon some transponder interface chips from Atmel, like the U3280M (now obsoleted, unfortunately), which can be interfaced to a microcontroller, and the ATA5570, which can't but has a sensor pin for hooking up a temperature probe or so.
    Both of these chips have built-in circuitry for contactless power and load modulation for communicating back (it uses the same coil for both energy transfer and bidirectional data transfer), and the U3280M switches automatically from battery power to the coil when appropriate, so a charging circuit could probably be built around that.  It allows pulling up to 15 mA from the pin, which could be used for charging the battery.
    There are more advanced chips out there like the ATA5790, which have a full-blown AES-128 cryptography engine, which sounds fantastic but I haven't been able to find a supplier carrying that chip.  Maybe I'll be able to get some samples for it.

    Point is, it's probably easiest for us to use an integrated solution such as the U3280M that handles this stuff for us.  I just haven't been able to find an adequate replacement yet.

    I found this resource, describing a design for an implant built around the U3280M chip and an ATTiny13 to monitor a mouse's vital signs.  Pretty cool stuff.
  • edited August 2012
    Turns out what I was remembering was the original ITAP paper...

    Direct link to an image of the implant (showing the structure) here.

    But yes, I agree with @ThomasEgi: subdermal things are going to give us enough trouble without borrowing future pain quite yet. Let's learn how to make implants, before we try to figure out how to make implants with transdermal components - insert tired saying about walking before one runs here.
  • edited August 2012
    that U3280M chip really is a gem for what we plan to do. a chip like that significantly reduces the ammount of parts we'd had to place on a pcb and thus bringing down the size. a shame it is obsolete.

    that crypto enabled version might not be all bad either if communication is required. in that case it might be possible to switch between the microcontroller and a dedicated charging circuit with some analog-circuitry-magic. more effort and parts but still an option.

    texas instruments seems to have some frontends too, http://www.ti.com/product/tms37157 bidirectional, battery charging included ( simple battery-levels can even be reported to the microcontroller, so you could warn the user to recharge soon). best of all, it can be bought in single-units starting at aroun 3 bucks each.

    adding to the list of awesome parts: bq25504 which can harvest power from extreme low DC voltages such as thermal generators or solar panels. talking bout solar panels, sanyo produces special solar cells for indoor use. they come in useful size and might be good enough to provide useful ammounts of energy when walking around outside with the solar cell below the skin surface. listed on digikey under the part nr AM-1417CA. just in case someone wants to test the output of it with a layer of pig-skin on top under different lighting conditions.
  • rdbrdb
    edited August 2012
    Awesome find!  I'll pick up some, next time I order from a major distributor, so that we can play around with it.

    It's a shame that it only provides a 2 mA max charge current, which is significantly less than the other one, but it should still be quite useful.  The built-in battery management circuits are a great plus.
  • kinetic generators from watches were exactly what i had in mind when i asked about what sort of amounts of power you were looking for. it seems like it should be enough for something like a watch implant that only lights up LEDs when you tell it to.

    i would think solar power would be difficult, if you got a tan then your implant would recieve less power, but if you kept it covered so that it didnt tan then it wouldnt get any power. do you know if enough light penetrates the skin to power a solar cell in the first place?
  • also, i had trouble finding more information about this, but it seems like a modified version of this could be used to power small implants: http://www.gizmag.com/go/7584/
    it is part of a project run by a university as the article says, so there are probably others that may be more suitable, but its small and is designed to power things that are inaccessable and even things inside the body. if it provides enough power it could be promising.

    also it might be useful to read this incase it sparks any good ideas:
  • @Squeegy that vibration thing is more useful for the use in car's or so, where you actually have vibrations. about the kinetic generator for watches … a regular LED consumes bout 50000 times more energy than a wristwatch. with the same energy a watch needs over 24 hours, you can power an LED for 1 or 2 seconds.

    @rdb: yeah 2mA ain't much. but if the actual implant consumes less than 100μA on average, you can still charge a week worth of energy in just 8 hours of sleep.
  • yeah, i know you probably wouldnt have the right vibrations in your body for it, but i would think with that principle in mind you could make something that would work? or maybe for people who spend a lot of time driving, if your flesh doesnt stop the vibrations too much. it was just suggesting an idea that i wasnt sure of, my background is all chemistry, biology and a bit of physics, i know very little about electronics and i didnt find anything specific on how the vibration thing worked, but i thought it might be of interest.
    as for one similar to a watch, i read more about it after i posted that and youre right, it doesnt seem very feasible.
  • I don't know if one of you already read it, but today I saw this little thing:
    Their "power cube" is only 0.8mm in length and apparently able to generate 50mW (given the right inductive field of course)
    Whilst being out of reach for now, this makes me hope for nice solutions for the power (at least loading) problems.
  • @Ben this is pretty much inductive charging, with all parameters optimized for the human body. it's good to know that the body conducts pretty well at 1.7Ghz. That might open up quite some bandwidth for communication in future. we can pretty much have that today, just with less efficiency, the current circuit which is up for testing should be around 6 to 10mW, and would have a larger receiver coil, with less distance between sender and receiver. But that should be plenty for this early stage of development.
  • I know, that it is inductive charging, but what I gathered the frequency was a new thing in the way that before they thought that only lower frequencies would work within our bodies.
    Could we build this today? It seemed pretty small and we do have some trouble with smd size, right?
  • UHF stuff is a bit tricky to tune correctly without the right lab equipment. smd itself is not that much of a problem. what i am more worried about is the 1.7Ghz band.. on first glimpse i couldn't find any good information about the use of radio bands. i mean you don't want to experiment with that in case it is used for airplane-navigation or so. at least not outside a shielded lab.

    i'd keep an eye on this, but go with 125khz for now. it might not be as efficient, but it's still good enough for what we need. and we can get all the parts off the shelf.
  • As for the through-skin connectors, that is one of biggest pains in medicine. In general we always try to keep anything going through skin for as short time as possible, because even with specialized materials, implanting in sterile conditions, proper medical care of the skin around the implanted tube (most often) etc. etc. they get infected way too often.
    Anything that doesn't have to have connection through skin - doesn't ;P

    Probably until the glucose cells get really good, the inductive charging is the only way to go really... And/or minimizing power requirements of the implant, like it's done with pacemakers... Though modern ones aren't that simple - they also record a lot of heart activity according to preset settings, which can later be read by using a computer with a special inductive probe connecting with inductive antenna in the pacemaker ;P

    I was thinking lately if it would be possible to use Peltier device, also known as Peltier cooler or heater... It's an electronic component/device that when one connects it to electricity does cool on one side and heat on other... But it can be used in 'reverse' mode - when heated from one side and/or cooled from the other side it produces electricity on its leads...
    Has anyone thought of that? The temperature between our bodies inside and outside is usually quite big, so if one would implant it just under skin it possibly could work on this heat difference...

    I'm just not sure how big surface it'd need to have and if its work in room temperature vs inside arm temperature would be enough to provide enough power to make it useful for powering any electronic devices...
  • we had peltiers in our discussions. but there are too many uncertain variables to figure out. especially the temperature difference could be simply too small to make it useful.

    so for now. we have inductive power transmittion (which also allows us to communicate with the implant at the same time so, that's cool anyay). and maybe some tiny flexible solar panels.

    but atm. the only production-ready thing is inductive charging.
  • @TI If you used an ice cube against the skin directly above the implant you could exaggerate the effect. Not sure if it would be effective for charging a battery though.
  • putting ice cubes on your skin, thus inflicting lots of discomfort while hardly getting any energy output... still not an option.
  • http://www.kjmagnetics.com/blog.asp?p=shake-flashlight Haven't seen any mention of these, it says power output for the one they made is about a third of a watt. No idea what the energy requirements for most implants are, but in the Southpaw discussion you guys were talking pretty small amounts. 

    If one of them won't generate enough, would there be an advantage to using multiple?
  • rdbrdb
    edited September 2012
    The magnet they use is really heavy-duty, especially compared to the tiny weak ones that are used for implanting.  They also have the magnet sliding inside the coil.  I really doubt you'll get anything out of this approach, though it's easy enough to try out and get the actual numbers.

    For now, the approach that has the highest probability of working is having a second coil in an external charger to generate the magnetic field.  This approach allows for bidirectional communication over the same line with the charger.
  • well, this thread is old, but it's the most recent of the 'power' threads, so I'll post this here.

    Some time ago I looked for thermoelectric generators small enough to be implanted just under the skin so as to have a large enough temperature difference to power implants. I couldn't find anything small enough, but it looks like people have been working on this.

    and here is the research company making them
  • peltier elements were already in discussion. miniaturisation of them is not the problem. the point is, they require a temperature difference. the links you supplied are all refering to clothing. given cold weather, there is a pretty good temp difference between the body (bit below 37 on the surface) and the outside of your clothing. maybe between 0 to 20°C. the temperature gradient under your skin is a lot smaller. and we have no solid numbers indicating that noticeable ammounts of energy could be harvested there using thermoelectric effects.
  • Also, the problem with peltier junctions in skin is that they generate power not by there simply being a thermal difference, but by the movement of that heat energy from one "side" to the other. That means it will actively be leeching heat out of your body and your blood vessels will react to that, constricting to minimize heat loss in that area. Unless you are working your body hard enough to be producing excess heat, your power output will be minimized by the body's natural defensive reaction... and if you're working that hard, you might as well just covert some of that mechanical energy.
  • I see, that is unfortunate, I saw this article 

    it says it can only generate a few milliwatts, but that it works with as small a difference as 1 degrees C.

    so, even if it wouldn't work implanted, could it worn on the skin above the device, and transmit the power? or maybe transmitting power even over such a small distance would require more than this generates? but if you had a large enough area could it work?
  • A transdermal may have to many draw backs, but perhaps we should consider something akin to a portacath. Perhaps the entire implant could be subdermal, but charged via passing something through the skin. Another similar is dialysis. People often have dialysis 3 or 4 times a week, which entails cleaning the arm and then accessing a large prepared vessel. As long as one takes care to clean the site well, infection isn't to much of an issue.


    The simplest manifestation of this would be inserting two conductive "needles" into appropriate regions to recharge a battery. I'm not sure how this could be bioproofed and I wouldn't want to have to be too accurate. I'll have to think about how such a port could work.

    A little further along... this type of design would be great for a glucose fuel cell. One could refill it with a syringe rather than trying to have it function off of blood glucose... and the fuel would be entirely non-toxic.

  • "The simplest manifestation of this would be inserting two conductive "needles" into appropriate regions to recharge a battery. I'm not sure how this could be bioproofed and I wouldn't want to have to be too accurate. I'll have to think about how such a port could work."

    This is one of the(I think) better ideas I've come across here.  Assuming your body has no problems with gold(or silver?) the rest could be bioproofed and inserting two needles every once in a while(whenever you'd like to use the device) would should not be an issue if proper sanitation is assured on every refill.Well its at least something till we get a good inductive charging setup.

  • Well, I think it has potential, but overall I with ThomasEgi that a system rechargeable through the skin is the most viable option at this time. Inductive charging is probably a touch larger, valid. The problem with a "through the skin" method is... the body is conductive. The signal would pass through the tissue from one electrode to the other unless you had some kind of shielded port.... a shielded port means a reservoir for infection or some kind of mechanical structure that can fail. With proper design... something I haven't thought of or seen yet... or even better, using a glucose cell (Not yet available, thus not viable), this may be awesome. At the moment though... I'm not sure this supersedes inductive charging.
  • hey everyone,

    i have an idea for a possible solution to this problem. i've been talking to people (EE types) about creating this with me since I don't have time (or some of the skills necessary) to do it on my own, and no time to learn either... but i want to create an implantable power module that incorporates the latest high power density Li based cells, 13.56MHz power induction coils, power control circuitry, and battery status LED indicator... all on a flexible polymer substrate which can be charged by an HF charge pad, but also kept "topped up" by running an app on NFC enabled smartphones that will increase the duty cycle of NFC interrogations in order to move more power. 

    there is still a lot of work to do, but I believe it's possible to create such a power module. i'm looking for volunteers with electrical engineering expertise, time, and perseverance. my intent is to sell these modules on www.dangerousthings.com and split profits with collaborators.
  • rdbrdb
    edited February 2013
    Thomas and I have been looking into some chips by Atmel and TI that are designed for this purpose, and have charging capabilities built-in.  The one we've currently sampled (though have yet to play with) is the TMS37157, which can be hooked up to a coil, a microcontroller and a battery and handles charging inside the chip.  It does bidirectional data transfer over the same coil that is used for power transfer.  It has a charge current of up to 2 mA, which should be more than sufficient for most purposes assuming a low-power microcontroller running at a low frequency.

    We've been meaning to turn this into a generic platform to use as a base for developing implants, though neither of us has much time to devote to the project at the moment.

    There are more of these chips out there, and they are very appealing since they are very integrated and don't require many external components.

    Keep in mind that LEDs require rather high amounts of current so if you turn one on continuously, the battery will be drained very quickly.
  • I figured the LED would be a "death knell" type of signal... the final moments of battery life calling out from beneath the skin :) I'd imagine whatever circuitry you connected the power module to would have it's own battery monitoring and signalling, so it's probably moot... i just thought it would be cool to notice your forearm flashing and then you put your phone on it :)
Sign In or Register to comment.