6.1.11

new year's resolutions

i. never to read threads about myself on other websites
ii. to talk more with my family
iii. to try and set up some sort of Southpaw development space rather than keeping it here, since a blog isn't ideal anymore for the volume of people coming in

any ideas on the latter are welcome.

L

70 comments:

Unqualified to speak said...

Time to once again pester you about email. Should we just set up some sort of incredibly-long-piece-of-string based communications system? (",)

Sebastian said...

Why refrain from reading articles about yourself. Apart from some comments the articles themselves are generally positive.

Anonymous said...

you gotta update your profile... you aren't faceless anymore. btw: nice talk you gave, one of the more interesting ones

-anonymous coward

Anonymous said...

i totally agree with sebastian, if people say bad stuff thats actually a good sign. like you get 300.000 hits on google if you lookup 'dell sucks' but still my school has like 50 dell computer that are running very nicely

Ian said...

L-spoon has been discussing the idea of a development space (not just the southpaw, but other enhancements as well). if you wish to, you can comment on his blog post on the subject.

~Ian

Ellume said...

Well I posted a couple time on the previous blog entry, but apparently neither of my posts actually showed up. Third time posting a similar message...

The ideas of this stuff have starting spreading all around the world. There is some hype right now, with Lepht largely at the center. The hype will be over in a couple weeks. In the end more people will know about, be interested in, and actively pursue this stuff.

@spoon from last blog entry but fits this blog well too: There are likely a ton of email questions being sent to Lepht that could all be answered by having a website, wiki, or forum. I think it's a good idea. I'm willing to help set something up.

While a website would require some $ to setup and run, it definitely has the most potential. A website could have it's own forum and wiki, as well as be expanded. It could also potentially make money if some adds are put up and there is enough traffic.

A cheap and easy thing would be to start a forum or wiki on one of the various sites that have them for free. Starting one of these is low risk, and if desired a website could be created at a later time.

Alexey said...

Development space - what I'm using is Redmine. You can set it up on any machine/VPS. It's better suited for software needs, but I think you can adapt it to handle hardware development as well. It has wiki and news modules with it, can store code and documents, allows multiple users and OpenID registration/login (doesn't work with Google OpenID though, Google seems to have incomplete implementation). It's easiest to install in Debian, but you can also install it on Windows or Mac machine. For more info check redmine.org.

spoon said...

Mmmmm Redmine looks really good!

I'll go see about setting it up on selfmodifier.org

Alexey said...

You can use some free solution as Ellume suggested, or if you're more of a control freak like me you can setup either home server with DDNS (if your internet connection gives you real IP address or you can negotiate with provider to setup port redirection through NAT) or use one of the cheap VPS options based on OpenVZ like, for example, virpus.com that have 0.5 GB RAM, two IPs, bunch of bandwidth and disk space for only 5 USD a month - more than enough to handle Redmine or Trac or similar project tracking software.

spoon said...

I already have a host! I'm just going to go set it up, do some housework on the server since it's going to have a jump in traffic.

What happened here is typical behaviour for me:
"I really need X!"
Then I start implementing X, instead of going on google and searching for it.

Lepht said...

Unq - fuck yes we should. mail incoming.

Sebastian - because it just makes me depressed when i see the really negative ones, and makes me want to jump in to the thread. that's not healthy.

re. wiki: i've no ££ right now, but spoon, if you are still setting it up, Redmine does look delicious. thanks Alexey and Ellume for the input, i appreciate guidance there.

L

spoon said...

Aye, I'm in the process of setting up redmine now.

It's taken a wee while because I've set up a couple of little tools to process logs,had another look at the iptables, sorted out dns, stuff like that.

I'll let you know on here when it's done.

Stiles said...

Lepht - is there a mechanism for donating a few quid to help things along?

Tilka said...

hi Lepht

- SA is getting a lot of traffic lately
- you are flat shit broke
- you don't like ads
- people are asking how to donate
- ???
- PROFIT

stay alive,
Tilka

mydnight said...

Was going to wait to post this until selfmodifier.org was up and running, since this is basically unrelated to your post, but before I forget:

Have you considered using supercapacitors instead of lithium batteries in the southpaw? Supercaps don't tend to degrade so much with time, and (very important for something implanted) don't tend to catch fire/explode spectacularly when damaged.

Unqualified to speak said...

Capacitors not explode? Um...
I've been poking the depths of the lithium-coin-cell lists; some of them are rechargeable, and are less unstable than li-ions.
I get the feeling this project is going to require buying several types of battery, building a charging circuit for each, and utterly abusing them to see what battery survives what torture, and what are the failure modes...

mydnight said...

Unq:
Electrolytic capacitors explode (but not unless you put way more current through them than an inductive charging coil -- they're only unstable while they are being charged), but supercapacitors (i.e. double layer capacitors), ceramic caps, etc. don't. I agree, they're not ideal -- and adding them will involve adding extra circuitry -- however, the primary advantage that they have over any type of rechargeable battery is that Lepht wouldn't have to cut the implant out and replace them at least once a year.

Unqualified to speak said...

I was thinking of electrolytics, yes; they're the most capacitance-per-pound, and I'm fundamentally a cheap bastard. ",)
I will admit to knowing nothing about double-layer caps, but aren't ceramics prone to cracking a layer or two under torsion or bending, though?

Dave said...

Just a random guy here who saw the article in Wired about your work. Loved it...best wishes. Hope the normals don't get you down to much ;)

In my own world I'm involved in the tropical vivaria hobby. Our little glass boxes full of exotic plants and brightly colored animals are beautiful by day but rather dull at night. I introduced artificial firefly lighting to one vivarium, glowing calci sand to another. Even working on ways to incorporate or maybe even embed fiber optics into living plants. Use electro luminescent wiring or fiber-optic cloth to create a glowing moss effect. Thundstorm/lightning fx using sound loops and plasma luminglas, 2d and 3d video background incorporation, star light fx, etc..etc..

I've even been researching naturally bio-luminescent animals/fungi that would wo well in our vivaria. Will probably try the 2 artificial options too, GFP glofish and axolotls. I'm even currently looking into what it would take to create a GFP dart frog in my kitchen ;)

Anyways there has been some resistance to my ideas. Questions like..."what is the point?". Sometimes I don't even know where to begin to address the question for someone if there thinking is so limited to have to ask it in the first place.

Anyways with my efforts to bring a synthesis of art, biology, and technology to my hobby and its community...I sympathize. Again best wishes to you and for your work. Be careful (but not to careful) and don't die...you'll make a random stranger sad ;)

Anonymous said...

i'd also vote for the supercaps. safer or not, they are easier to deal with. liion cells need a very careful battery management, especially deep-discharge is very bad and i expect that to happen quite often in this application. supercaps can also be charged ways faster than li-cells. as already mention, caps win over li-cells because they can withstand virtually unlimited ammounts of charge/discharges. on top of that caps are widely available in all sizes, in industry and military grade quality for a budget (i cant claim the same for my ipod,battery which broke trice in 5 years or so). and you can easily set up a grid or chain of several smaller ones, so you'll get a more flexible structure which doesnt break under torsion.

i also recommend looking into atmega8 microcontrollers. they have special versions with li-battery-management built in(in case you need it), which are very small yet still human-solderable (with some lab equip). i think those verions are labaled 8H or 8HA or so, you can find them right away on the atmel page. those also offer extremly low-power modes.

mydnight said...

Unq:
Sorry if I was unclear -- ceramic caps and double-layer caps are actually two separate things. Yes, ceramics crack under stress; that, combined with their low power capacity, make them unsuitable to power something on their own -- I was just using them as an example of something that doesn't blow up. Electrolytics would be right out, even if there was no explosion danger -- they're designed to store power over time and then release it all at once (the opposite of what's needed). Double layer caps are actually two layers of activated charcoal inside a metal "can" -- they actually look like button cell batteries, except with leads on the top and bottom.

Unqualified to speak said...

I know double-layers and ceramics are different things; thus why I criticised the ceramics but said nothing about the supercaps. ",)
I don't know, I really don't. Order of miscellaneous power-storage components and necessary test-harness gear seems to be looming in the future. First, I worry about the coil.
@Anonymous: ceramic capacitors break inside the package. It's up to the manufacturer to fix the issue, we can only mitigate it.

Anonymous said...

@Unqualified to speak said...
hm. well since the package would have to be entirely coated in biocompatible material anyway that shouldnt be the problem even if it craks, unless it destroys the coating.

i spend the last few minutes searching for a hand full of suited parts and did some rough power estimations. with a hmc5843 compas module and a atmega8hva (both nicely connect via i²c so no mess with analog circuitry)one can get operating voltages between 2.5 and 3.3 V with a active power consumtion of 800uA for the atmega and about 200uA for the compass chip. with idle-currents beeing as low as 150uA in total. put onto a thin pcb it fits into a mere 4x10x2mm space, so only slightly bigger than an rfid chip.

well. bad news is, even with that extreme low power consumtion the device would only last 3 to 5 hours on a full super-cap with 2.5F (and those are almost a 10mm thick)

with the tiny lipo battery from an ipod shuffle (a mere 3mm thick and a long 25mm on each side) the device would already run 5 to 10 days without the need to recharge. with the battery management already build into the atmega that may even be an acceptable sollution. since the body keeps the battery at 37°C and the load is quite weak, it may be well possible to reach the "usual" 300 to 1000 charges. depending on the final implenentation that'd be 5years or more under normal operation.

given you put the charging coil on a flat,flexible surface, the whole package would be about 30mm wide 30mm long and 5mm thick. with the coil beeeing folded around the package untill it's beneath skin level where it's unwrapped in place. only thing's missing on the bill would be some suited neuroelectrodes.

anyone knows li-cells smaller than the ones used in the shuffle? atmega8ha also allows 2cell operation, so 2 smaller ones may even be better.

spoon said...

If you want, you guys can have a wiki now.

http://selfmodifier.org

Kuro said...

When it comes to dev space, you could use my server. If I recall correctly, I mailed you a few months ago, though either you never replied or your reply was buried under the spam in my account; dig through your trioptimum/gmail account enough for "Kuro"; contact me there for specs.

Unqualified to speak said...

@Spoon: woo!
@Anonymous: cracked ceramic cap = lessened capacitance, up to failure of the component. Have a white paper: http://www.syfer.com/category_docs/AN0005_Mechanical_Cracking.pdf

Hmm. The problem with that idea is - how do you signal the battery could do with charging?
I suppose a most-daily trickle charge designed to keep the cell around half-charge (the long-life point of a Li-ion) might work.

Anonymous said...

@unqualified: true that, even if failure of a cap is not endangering health or life (my definition of safety). unfortunately your link didnt work for me.

easiest way to charge would be to depend on inductive charging entirely. requires no critical parts and very little circuitry, and (if my rough power estimation is not completely off limits) it should be enough to charge it for about an hour or two to get going for a week.

bout that trickle charge (aka "energy harvesting"). the most promising option , to use a dynamo (like a magnet in a plastic tube which has a coil on it's outside.. unfortunately are out of question when you deal with magnetic sensors which are supposed to pick up earth's magnetic field. if you seperate the powerunit and the sensor far enough it may work but i dont know to what extend. certainly worth to build it for testing purpose and tape it to the leg.
bio-batteries are too big and not yet available for purchase. there is no temperature gradient worth to be mentioned so peltier elements are out,too. that pretty much leaves you with piezo-electric. those do exist in labs but i wouldnt know a good source to get one. well solar wouldnt work well either.
whatever the choice, the atmega provides a coulomb counter, so during the entire testing it should be possible to get detailed power-statistics.

Unqualified to speak said...

Nah, trickle charge as in "slow charge over the course of the night". We want to charge slowly and not fill the battery to avoid the battery heating up.

Anonymous said...

ah. thx for explaining. well that shouldnt be a problem at all. you can freely choose charging-current by using an approriate Resistor used for sensing the charge-current. once soldered, you can still change the maximum charge-current up to a factor/fraction of 10 by changing a few values in the microcontroller's register.

i dont think heat will be a problem. currents are small, li-cells are rather efficient. since a body is able to deal with drinking hot coffee there should be no problem even if there would be quite some heat to get rid off.

adding a charging-indication led might be a nice touch and would satify the led-demands of the people out there ;)

Anonymous said...

You should totally check out the movie Tetsuo the iron man

Unqualified to speak said...

Ak! I was thinking of NiMH cells when I was talking about temperature. My brain needs a disclaimer. ",)

Anonymous said...

hm. nimh cells are difficlut to charge, especially when charing with low currents the deltaV method becomes hard to get right. you'd also need 2 or 3 of them in series to get a useable voltage without step-ups. and they are quite big.

i just spotted those li-coin-cells someone mentioned earlier. the bigger models range from 50 to 100mAh (24Øx3mm). enough to keep the thing running for 2 or 4 days without any optimisations. with some more agressive power-saving on the microcontroller side it may last up to 2 weeks.
drawback for those nice coin-cells is charging speed. they almost require 50h to be fully charged.
with low enough power consumtion those are a nice choice. giving about 10 hours of operation for each hour charging. with a total reserve of 12 to 14 days. and if charged every few days the cell wont be deep-cycled so it could last many many years.

Kuro said...

Anonymous: charging speed would be rather an issue, unless the user of the tech plans on carrying around some sort of external pow supply with a charger fed into it around, but I'm unsure exactly how viable that'd be; the most obvious concerns of lugging or wheeling something around aside, if you're carrying power around, might as well not have a battery. Some sort of internal auxiliary power supply other than the main battery might be viable, but as far as I can think (though don't trust me to think; I haven't slept in about 20h) the device itself wouyld have to be programmable to switch batteries like that.

Anonymous said...

@kuro: current isnt picky where to flow so the switching between supplys happens all by itself. if the user is ok with wearing something like a charging-sock while sleeping it'd be totaly fine. the internal battery should be able to keep the device running for a week or two without any charging. charging it 3 hours should be enough for one day of operation. so during an average-short-programmer-night of 6 hours you could get 2 days worth of operation.

the charger itself would be rather small and leightweight and can be build as wearable device with a battery. the implant's battery has a capacity of just 100mAh, even a cheap li-ion battery which would be suited for that case ranges up to 10 to 20 times as much. i'd make that wearable charger having a micro-usb port so you can charge the chargers battery on any pc.
so it would look like:
pc/usbhost -> wearable mobile charger (may hold between 3 and 10 full charges for the implant) -> implant's internal battery (which runs about a week when fully charged).
so much about the concept.
with a good,thin silicone coating the implant itself would be something around 4x28x35mm in size. or Ø28x6mm. depending on the arrangement of parts.

adding a small orientation-dependent switch or accelerometer it would be possible to turn off the device when the device is in certian positions (like horizontal for sleeping).

so much about power supply and sensors. what's really missing are some good electrodes to give a feedback to the user. anyone ?

a forum of some sort sure would be a better place to discuss this topic.

Ellume said...

I've suggested that a forum be put up on spoon's website. So I suspect one might pop up there in the near future, otherwise one can be opened on a free forums website.

Anonymous said...

@Spoon:
You can get a free valid SSL cert for selfmodifier.org from StartSSL.

Kuro said...

Anon: Regarding the current, ah, obviously. Note to self: sleep, then internet.

Regarding the charger, perhaps a car cigarette-lighter plug would be useful for those who travel for long distances in the car, but generally USB would be of more use.

As for electrodes, I've seen some quite high prices for so-called high-grade electrodes, but the matter would more be one of implementation than of obtaining parts.

spoon said...

Thanks anon, I'll set that up.

Anonymous said...

@kuro there are tons of cigarette->usb adapters out there. first hit on amazon was just 3$. many modern car-radio even feature a usb connector by default. could you post sources for the electrodes you found?

Max said...

So many comments, awesome. :D
Just a small question: Are we trying to assemble the thing into one coated package or rather into a number of wired packages?

Having a single package gives us the advantage of only having to get one thing plus electrodes under Lepht's skin, but it means this thing will be big. The other option means a lot of wires, but very small components.

How viable would it be to just use multiple smaller batteries/caps instead of one big one?

Then, regarding the mobile charger: the charging coils will have to be kept lined up. (Previous suggestion here was small magnets keeping everything aligned, because screw you, magsafe patent)
My question is: how much of a problem could the magnetic interference of these magnets be for the accuracy of the compass?

Anonymous said...

@max: the package wont be that big. if we use only the smallest components available (a bit nasty to assemble but there sure are ways to get it done). rough size estimations can be found in my earlier posts. i highly object to the idea of several independant parts since connectors are always an issue. if anything, then it'd be one part made from several smaller rigid components with a flexible connection (but still all in one sealed package). multiple batteries are possible, but i would suggest not more than 2. caps are out since they simply dont store enough energy.

alligning charging coils is not really critical, it wouldnt matter if they are offest a bit. you could simply make the charger coil a lot bigger than the implant's one (the implant's coil could be integrated into the PCB) so even if the charger is off by 5 cm or so, it would still charge very well.

like i already said. magnets would be terrible, the earth magnetic flux ranges between 30 and 50 µT. a cheap neodym magnet can easily reach 1T near the magnet's surface. which is 20000 times more. you can easily try the effects with a cheap compas.

also, patents are not an issue here since we'r not going to sell anyting.

what worries me most are getting hands on some electrodes and getting them hooked up.

Mazzalupi said...

@spoon @Ellume I've been writing back and forth with some existing personalities in the grindboat about putting up a forum. We get a good mix of solid admins and the revolution's off to a good start. Kickass graphic/web designer onboard too.

We just need to oust the squatter first. I'll see about whipping up a temporary space by tonight as it looks like we need a board now rather than later. Then we can transition to the new domain when its ready. Trust me folks, it'll be worth the wait.

Ellume said...

@Mazzalupi: Cool, sounds good by me. Let us know when you got something going.

Mazzalupi said...

@Ellume Just asked our design extraordinaire to shove a temporary forum out while we track down the domain squatter.

It's 5am where he is; give the man time for coffee and we'll be set. ;)

Mazzalupi said...

Temporary domain's up biohack.me. This will be the forum's home until I buy out the squatter.

We're installing the forum software now. I'll update here when that's done.

Unqualified to speak said...

Experiments with coils, function generator, and oscilloscope are conclusive: I need to build a power amp. _Another_ delay; what a shock...

spoon said...

@Mazzalupi marvelous!

Mazzalupi said...

And we're up. Go register. http://biohack.me/

Anonymous said...

A thing which came to my mind is, if you want to spare with the hassle of an implanted power source, it would suffice to just implant the coil and provide the current by a powerpack and coil strapped to your leg. It would be a bit of cheating, but in the same time reduce the number of components to develop.

I guess the biggest hassle will be making sure the holes where the electrodes leave the bioproof compartement stay tight, because I can't help but think of them as cracks (which they are, but shouldn't!) (;

Mazzalupi said...

@Lepht Maybe you could endorse the community resources we have in a post so everyone knows where to go without digging through comments. You can always email me back via my Re: Grinding the future email if there's any logistics or issues you want to address.

Anonymous said...

@anonymous (2 posts above)

that wont work. the implant is supposed to messure the weak magnetic field of the earth. remember inductive charging also uses magnetic fields (lot stronger than that of the earth). during operation it conflicts so it's only good for charging at times where it's not in use (sleep / sitting around)

Unqualified to speak said...

@Anonymous: also, induction charging requires a changing electric current. These are hard to get out of batteries with much efficiency, whereas one comes out of the wall with no added complexity. ",)

Anonymous said...

as electrical engineer i have to object. creating AC from batteries is really easy and efficiency is quite good, the efficiency-breaker is the magnetic coupling between charger and implant. but since the implant is very low power and the charger battery can easily be many times more powerful, no problem on that end.

Unqualified to speak said...

@Anon: I was thinking of induction coils at the time. Sorry.
How would you do it? All that comes to mind is variations on 555s switching current.
And it's nice to have someone who knows their stuff weigh in, as opposed to me, who's named what I am for a very good reason. ",)

Anonymous said...

you can use a 555 and an LC series to get an a somewhat ok sine (with a a few jumps), simmilar is used for cheap 128kHz rfid equipment. would be good enough for charging already... even if the guys from the EMC lab will curse you. using op-amps you can get a lot cleaner sine-waves.

Unqualified to speak said...

I really need to learn more about op-amps; I only know enough to know I haven't scratched the surface.

Ben said...

About the electrodes:
I'm not quite sure who of you ever worked with measuring and inducing of neural signals, so if there is someone more experienced than I am please ignore.
I hat one course in Neurobiology plus several in biochemistry and so on, so I am well aware of the working principle of neurons. What we used when recording the signals of neurons were simple platinum wires. The whole electrode was a small wire. Most of it was coated so it wouldn't pick up currents from an area but only the one that was touching the very top of the wire. Think of a wire completely coated but then you cut one end, it still is coated but at the end there is one spot of blank metal.
We had to hook it up directly to a neuron, of course.
As in this case the electrode should induce a signal rather than record one this kind of accuracy is not necessary. I would think of a small ring (diameter of maybe 2mm? need to look up the neuron-density in the target tissue) which is not coated. You can get platinum wires as lab supply, for what I know. Never tried though.
One problem I can think of is not to induce the signal to too many neurons, or too strong of a signal (shocks...) and not to hit the wrong ones. If hitting the motor-neurons they will trigger musclemovement.
Also it would be interesting if the electrodes have to be set up all around the ankle like the motors of the northpaw are or if they just have to trigger different areas. My guess would be: different areas should suffice, but an experiment with an actual northpaw could clarify that quite easily...
Hope I could help a little...
Ben

Anonymous said...

that sounds like it would already work if we could insulate a few platinum/iridium wires with a small loop left blank, and insert a bunch of them into the skin? sounds a bit too easy to be true. what kind of sensation would that result in, how about neccessary voltages/currents to generate a sensation without causing any pain?

it sounds easy enough tho *glares at a bunch of needles*

any information about pulse shapes/duration etc would be useful.

Ben said...

The kind of sensation should depend on the voltage and on the kind of neurons you induce the signal in.
I WILL look it up, not now had an awfully short night and am down now. (explains all the bad writing, too)
What I recall is a current of 60 to 100 mV. One AP (action-potential) takes around 50ms (or was it 50Hz? damn, really have to look up all the stuff again)
But you don't want only one. Depending on (again) the kind of neuron it will follow a different pattern. Most of them will simply stop reacting after a certain time. Plus you have to wait until the neuron can regenerate the ability to send a signal.
Guess it makes more sense to look it up tomorrow and back up my claims with some more information ;) don't want you to have to trust in my word
But yeah, should be not too difficult actually. And do NOT play with ac... It operates on the same frequency as our nerves, that is why it is so dangerous for us..
Ben

Mazzalupi AKA Sovereign Bleak said...

I'd like to request to move discussion of Southpaw development to http://biohack.me so we can help people interested in its development find the relevant info without digging through comments.

Ultraviolet said...

=) you're so full of awesome! i am a bit worried about your health but someone has to do these things first.
And i know you're careful as much as you can be :) *shudders* self surgeries are a nightmare thing though, i totally couldn't... although i am contemplating LEDs on some level of my consciousness :) vanity.

anyway just wanted to say we are so lucky to have you - transhumanist myself i feel totally like honoured to witness the early experiments.
Take care and don't worry about the idiots posting rubbish ;)

xx
UV

mysite said...

Nice Post!
I would like to know more about your blog. You may also visit my blog and tell me what you think.

wireghost said...

@Unqualified: An H-bridge could also be used in place of the 555 and amplifier.

Anonymous said...

Тебе следует изучить бионику и кибернетику, изучив эти науки,
ты можеш творить все что угодно...
Я тоже этим занимаюсь, но ничего вживлять не надо...
Главное знание!!!

Kevin said...

I live by rule #1. It preserves sanity.

Anonymous said...

Hey lepht, really liked your talk at ccc. First time I had heard of you and have been poking around your blog since then. Had noticed you had not posted recently and was just a bit worried about you as a fellow medicated individual. Any how just a shout out from down here in Australia, love your passion and found your talk inspirational. Hope your well catcha l8r...

NastyCyborg

Ian said...

@SovereignBleak: It would probably be fruitful to have L link everyone to biohack.me, or to put a link on its FAQ, or something like that. L's probably busy with university like it was last semester, but if we could get it active on biohack.me that would be awesome :-)

Anonymous said...

Is Lepht alright ?
It's been a while since she posted.

Ian said...

@anon--it's probably busy with university like it was last semester; it already has to repeat the year.

Lepht said...

evening, all. many apologies for recent lack of life. full post about new resources coming. owe you all, many times more than i could ever repay, for your input on this.

L

Post a Comment

[pls no ask about the vodka. debate is always welcome. remember, Tramadol fucks you up]