No electrical engineer worth his salt would laugh at rotating a potentiometer and gathering individual data points. It's a time-honored tradition amongst us. It's true that it's just not efficient for automated production, but for laboratory work, it's more than acceptable!
raffitz Doctorate student in EE here; I still do most of my data capture by spinning knobs and recording numbers. It's more fun than setting up data acquisition mainframes!
Furthermore, the circuit you set up to both account for voltmeter resistance and sweep two quadrants of your I-V curve was a better design than a lot of undergrads' work. Well done.
Came here to say the same thing. When I was in school (2 year polytechnic), we did all of the measurements like that manually. Besides the learning aspect, setting up an automated rig is just not worth it for one-off experiments. With an automated rig, you could end up spending more time trouble-shooting it than it would take to simply do the measurements manually.
You could use it to power up an extremely low power microcontroller unit to do work from time to time and sleep in the intervals in-between. Since the nuclear battery is always producing current, you will want to design your circuit with energy accumulators that are slowly charged by the battery output, to be used when the microcontroller wakes up. That way you could drive a much more energy intensive device for a short period, if only you have the time to wait for the energy to be generated.
You shouldn't be so hard on yourself. I'm impressed by your I-V setup as an electrical engineering student. It's awesome that you've taken the time to learn so much about a completely separate field. Your knowledge of voltage, current, and equivalent shunt and series resistance in better than many of my peers'. Great work. I hope you keep learning and sharing your knowledge with all of us.
@Iridium this is about the experiment. He's trying to explain the power output of the device, and do it thoroughly. If you want entertaining videos, stick to videos that give no explanation and are entirely about the action. That way you can just see results while learning absolutely nothing.
electrical engineer here: you're circuit is not laughable. you took the measurement error from the voltmeters into account.I was thinking 'oh no a voltmeter next to a 1MOhm' but then you explained it. even the two ways of measuring the thing. not laughable but laudable !
Meanwhile, English majors are laughing at your comment, mainly because of using "you're" instead of "your" Then again, at least electrical engineers have jobs, all English majors can do is enter the highly competitive field of writing books
Some random person here: it's your circuit. Not "you're". Learn how to talk then I can read the rest. Seeing you can't even meet the extremely simple conditions to create a simple dialogue I can't imagine you're good enough to think complex matters like nuclear physics. And having a small store filled with scrap hardware doesn't make you an electrical engineer you know..
I see these tritium vials are available in different colours. Green phosphor around 500nm should give the best efficiency with amorphous silicon cells.
Nice one Dave :) I just asked what the optimal wavelength of light would be best for ASC's, so knowing this would save +NurdRage time & expense in trying different colour tritium vials
Good to hear from our old friend EEVblog. Now we need a good application for a pocket sized nuclear power plant! Any ideas? Please! Maybe a 10 year timer with no display? Maybe the power source for the timekeeping part of a watch with no display. You push a pizoelectric button once in a while to power a display for a moment as you are reading the time?
Outta curiosity would there be interest in me testing different colors? It would be expensive though since i have to buy them all. But i'll consider it if there is enough interest.
I'd love to see you make a solar cell, because i'm sure there is some interesting chemistry there. Not to mention just interesting things you can talk about and show.
Are you planning on looking at the absorption wavelengths of your photorecpetor to have an ideal wavelength you're aiming for/can use as the baseline? Likewise, would you take into account the efficiency of the electronic transitions for the colored tubes? Red-yellow phosphors are slightly more rare due to their weak outputs and the phosphors using Bi^2+ as the phosphor for red have yet to really leave the lab, similarly for other non-rare earth based red phosphors like Mn^4+.
Great video! I think I might have to make one for kicks. For reference though, a single CR2032 battery at 1uW output would give you a theoretical 600,000 hours operation, or 68 years, so the shelf life of the battery in practice (maybe 10 years). So as you mention, only useful for applications were a battery can't be used, or for longer periods than 10 years.
Whoa! Awesome to see you here! Thanks for stopping by. Ya know... if you want... I can send you my battery. Maybe you can insert it into an interesting circuit or something.
EEVblog could you send this nuclear battery to me? it was my idea, anyway americium from a old smoke detector and phosphorus phosphates is better it lasts for 420years use neodymium magnetic induction and lead shield to power a mobile phone 📱
a casio scientific calculator is in the nanowatts range and runs on 1,5 to 0,8volts. you ciyld easily power that and have a nuclear powered scientific calculator !
djteac yeah but the display is just an lcd with a reflective sticker behind which you can peel off. if you add a difuser and the tubes behind it could would as a backlight although a dim one.
I'm an engineering student and I'm not laughing at your circuit. It shows a very basic, often overlooked principle, the Wheatstone bridge, combined with current measurement. You even went to the lengths of explaining the effect of internal resistance of the meters and how it would affect your measurements.
NurdRage you stole my idea this was based off my idea of using bioluminescent glow in the dark mushrooms as a alternative source of sunlight to power electronic circuitry. try a simple experiment of placing the isotopes into Durex Play vibrating cock ring when the isotopes vibrate this might increase current flow 🤔
GREAT VIDEO!, small error on the parallel resistance it should be (1M^-1 +10M^-1)^-1 when you measured with an aditional multimeter you measured (1M^-1 +10M^-1+10M^-1)^-1!
its better than them becoming some kind of toxic crusader who goes around saving people by smashing villian's heads in with hyper violence and gore for a 4 part series that later turns into a in house cross over with some drunken foo fighter and they team up to save the day only later having it spin off into a non hyper violent teenage mutant ninja turtle clone kid cartoon.
Very interesting video, I bought a tritium light years ago and it's fascinating. I'm an electrical engineer and I assure you I'm not laughing at your circuit. We are all amateurs and specialists at different things. As for what you could use such a battery for, some embedded low power microcontroller for data logging comes to mind. It would sleep most of the time, only sipping power when necessary. Similar idea to energy harvesting applications. Additionally you could slowly charge a low self discharge supercap when the device is sleeping.
sjm4306 - what could be used to power a small drone? since tritium seems to not be powerful enough. havent began my education towards becoming an electrical engineer. just about to get my job as a custody assistant with the sheriff's dept. so i can afford it.
Congrats on the job. A drone requires quite a lot of power to generate enough thrust to fly so they pack lithium ion batteries onboard. I've seen researchers mount things like solar panels and wireless charging to drones to help recharge them during flight.
***** - not impossible at all. There are tritium powered microchips and controllers, as well as batteries powered by tritium and promethium-147. The list continues exceedingly, but you get the point. Using energy from the decay of a radioactive isotope to generate electricity
Few things are strictly impossible, but I think a fair statement would be to say that with current technology tritium batteries cannot directly power anything that requires more than microwatts or so without charging an intermediary battery. However tritium would be useful for something that just sips a tiny bit of power every once and awhile. For instance some sort of wireless sensor, beacon, or watch. Think Internet of things, wearables, or remotes.
5+ years ago there was a patent for tritium paint. The cost was estimated around $12/sqft for the stuff. I was totally going to paint that stuff onto a box full of sandwiched amorphous thin film cells and build a much larger nuclear generator. Though the patent was real, the paint never sold. I briefly looked for it again and looked for my notes to no avail. This brings back memories :)
nah in the lab they would have to take H and He and then inducing fussion just to use the results for a fission process if they want to produce hugh amounts of electricity available. On the other hand the vid says that tritium is to release more electrons in order to generate power batteries, well thats the theory but there are still other materials available for the same purpose.
No laughing here. I am an electrical engineer and your method is great! This is a one-off experiment. So sweeping by hand and taking measurements and entering into a spreadsheet is the way to go. After you have taken all your measurements with your setup, the ones using ADCs and Arduinos are still writing the code and building it. Now, if you were going into production on your Tritium Power Source, an automated tester would help. In fact, I am going to build these in the STEM class I teach and use your test circuit. Great job!
I believe there are amorphous solar panels which are catered to absorb blue light, and blue is an available color Tritium light. Also, there exist panels catered to absorb green - I might suggest getting a small panel which matches your existing Tritium lights would be worth experimenting with first... But from what I gather, blue light power generation is higher in amorphous panels, as the light is higher energy. You might also research what efficiency the phosphors coating have at converting B- to photons, and matching that to known panels with specialized absorption, and seeing where from the most power can be drawn. Also, maybe a single panel with designed parabolic reflectors for the lights might be a design worth making.
Would splitting the vials between two cells produce more power than just sandwiching alone? You put solar cell with half the vials and a thick reflector to reflect the light energy and then repeat with the other cell and vials and then sandwich together and see if you have any improvement on the odd 20% over a straight forward sandwich build.
That''s a fantastic bit of science and one of your best ever videos. Thanks! By the way, I estimate the total efficiency like this. You achieved a performance comparable to commercially available tritium beta voltaic batteries. You got a peak power at 820 nanoamps and 1.5 volts or 1.23 microwatts. Let’s calculate the efficiency of converting tritium decay energy into electricity for this device. I estimate the quantity of tritium in his vials like this: the tritium vials you used were 3 x 22.5 mm. I found some data for larger vials on wikipedia [ 1.8-curie (67 GBq) (152.4 mm × 5.1 mm) ]. Scaling by total volume that's 21.5 MBq/cubic mm, so the small vials might be ~3.4 GBq. At 18.6 KeV/decay, that's 10 microwatts of decay power per vial. You used 14 vials, for a total of perhaps 140 microwatts. So the conversion efficiency he achieved is on the order of 1%. Amazing!!!!
Is there any reason to why is it so inefficient? I mean, it's amazing to achieve efficiency compared to professional equipment but I still find it very inefficient
@@0Arcoverde probably because you lose some on the conversion to light and some on the conversion of light to electricity. If both steps are 10% efficient, you would get 1%.
The concept of it is awesome but the amount of space and money you do need to have for the pro version is just rediculous and that only for 12 years, now if it could last for 12000 years ,even if it would cost me $30 billion dollars just to get 6 volt or 12 watt out of it just to be able to power a transformer to convert 12 volt into 120 just to power up a toaster, i still will not get that money back, i may have to live with bills for generations to come but i atleast will have free energy.
If Im not mistaken, the dim light efficiency difference between monocrystaline and amorphous cells has to do with minority carrier current. In a monocrystaline cell you start with a uniformly doped (usually N) chemical vapor deposition grown boule and the junction is formed by creating a P doped layer by adding P dopant atoms. The N dopant is still present, however, and there are a significant (in this case at least) number of minority carriers in the P doped region. The amorphous silicon cell is grown on a glass substrate and the P and N layers are grown independently, not doped afterwards, leading to fewer (but never zero) free minority carriers conducting current in the wrong direction.
Nick Smith I could probably reasonably simplify the explanation of the difference between minority carrier current between the two types of cell, but I think I would be doing the subject a disservice and not explain why monocrystaline cells are _more_ efficient at _higher_ luminous flux, which I felt a bit bad about not doing in the original comment. So, unfortunately, I could (maybe?) explain it in lay terms, but I don't think I could do that as easily with the bulk difference between monocrystaline and amorphous silicon.
Having something continuosly running from this is possible, but hard, especially if it needs to be useful. First of all, I think a MPPT unit in front would be very hard, if not impossible to do without getting any gains, since the active devices and leakage currents would swarm any benifit from it. I think the key here is as low of a component count as possible. First of, the voltage of the output should be buffered by a ultra low leakage capacitor (most often tantalum is used). A random ultra low leakage cap (Kemet T489 series, 10uF) gives around 0.0075*C*V leackage in uA, or around 0.0075*C*V^2 in uW. For 10uF that is around 7.5e-8*V^2 uW, or for the maximum solar cell output voltage of around 2.4V that would be negligable. We of course assume the capacitor is almost always fully charged, so over the curve of the maximum power point. A typical ultra low power arm microprocessor (SL EFM32 gecko series) is around 2V @ 20nA in shutdown mode, and around 180uA per Mhz. So converting to power gives a power draw of about 40nW power down or 360uW running. This gives a charging power for the capacitor of around 600nW or so on average (ball park figure). This would mean that the device would have to charge for around 60 seconds to give 0.1 second of operation of the microcontroller at 1mHz. Or speaking in a different way, 600 seconds of inactive period for 1.000.000 instructions. So about 1666 instructions per second on average. With which you could do a whole lot. Of course this is neglecting any other circuitry needed to do useful stuff.
I think we wouldn't need a classical MPPT because light intensity is more or less constant. You can just assume fixed input voltage. This would probably be at around 80% of open circuit voltage.
That is not true what you say, the input voltage will not be constant, since it is a function of load, and that is the entire concept of the MPPT, that you vary the load to draw as much power as possible.
Yeah but he is right if your load is constant too and such that it makes the solar cell work at MPPT then you don't need active circuitry, the MPPT unit you were talking about. At least for some time because the MPPT would drift caused by the light intensity being not really constant and going down with time.
Or you could use one of those LT energy harvester IC's. They take ultra low power input and use it to charge a couple of 100uf to say 3.3 Volt and tell you when the power is available. Then your micro does something and goes to deep sleep again until power is available once more. If you use something like a tiny Gecko proc, you can do a lot with a 100uF if you use it right.
This was excellent. Every time I started to think ahead, you were right there with an analysis. Totally worth the wait to see what you did with the decapsulated tritium vials!
A CMOS latch to capture some rare event of some kind over a long period of time. Well within the scope of beginner electronics and possibly one of the lowest power circuits you can build with common parts, though you'll probably want to double the voltage for that.
This video is amazing--and not just because "oo cool, radioactive battery" (although, I will admit that part of science is pretty awesome). But because you used a variety of aspects in physics/science that aren't commonly brought up in the videos/experiments: the math behind it. The electric circuits mostly went over my head, having never studied them. However, the usage of the graphs to interpret the data, explaining how to compare the data when they were different sizes, etc.--those things were really awesome and I'm super happy that we have a video covering them. I recognize chemistry doesn't lend itself to mathematical comparisons, as most of it that is shown is either conceptual or hands-on stuff, but as a physics nerd, this video made me happy. :D
@NurdRage I found suitable chip that can be powered by these! MCP7940N works with as little as 1,3V and 890nA and it's enough for it to keep counting time. Of course to read it and do something with that you need external power, but you can always connect that battery to it and know, that you won't run out of time for years :)
I was curious and decided to find out how much energy this Tritium Nuclear Battery contains, so here goes: From the video, we know that this battery outputs a maximum of 1.23μW Additionally, 1) The half life of Tritium is approximately 4500 days or 3.888 x 10^8 seconds 2) Decay constant: λ = ln(2)/(half life) ≈ 1.783 x 10 ^-9 Assuming that the amount of Tritium is directly proportional to Power, Energy = ∫ Power(t) dt = ∫ 1.23 x 10^6 x e^(-λt) dt Integrating from t = 0 to ∞ , Total Energy ≈ 690J (if you had an infinite number of seconds to wait for every single Tritium atom to decay) It seems like NR was right about not expecting free energy from this project :P A typical AA battery contains 20 to 30 times the amount of energy in this Tritium battery, and I didn't even account for the phosphor decay!
I've been trying to do this using cs137 and blue phosphor powder. Blue phosphor powder gives the brightest light in lumens. If you get ahold of me on email, i can share my results and give you the nuclear lab i get my cs 137 from.
you'd also have to take into consideration at what wave lengths of light the solar cells are most efficient, I think that would have quite a big impact as well.
Psh, all that work. I just drink a Radium-water solution every morning. I, myself, am a light source that will last a thousand or several thousand years.
can the nuclear lab produce heavy water I need some for my "project" so far I just got some nitro glycerine and some plutonium of un named scource as well as a excessive amount of brake fluid and chlorine
You can snag a ZnS(Ag) screen which is a scintillator that glows blue when struck by ionizing radiation, on Ebay for $5. Try that. It works very well for my strong-ish Am-241 sources that I'm trying to make a battery out of for the last few months.
Nicely done! I'm impressed with your knowledge of EE for only being a chemist. ;) Just messing, you know more about EE than most engineers I went to school with. On the subject of lights choices, if you are able to get them, white would be the best, but if it's only specific colors I would recommend the one that can produce the most photon energy, which would be blue of ~ 2.62eV. What I would be curious on is using different types of solar cells. I know you briefly talked about thin-film amorphous, poly-cry, and multi junction, but have what about actual material. Curious on the difference between a Si (band gap = 1.12eV) or a Ge (band gap = 0.66eV) based cell.
Interestingly enough. I read that Indium Gallium Phosphide solar cells are the best so far for this purpose. Their shunt resistance is exceptionally high and their band gap is perfectly suited for green/blue. the problem is they're ludicrously expensive and custom made in laboratories. If you're a military or space contractor you might be able to get them. But not even I have procurement channels for something that exotic.
Personally, I consider my mind blown by this even if it can't power a phone. (Yet) I didn't even consider such a thing as nuclear batteries even existed, I can now look at my tritium keychain light with a new kind of awe. Thanks for a great post. :)
Very nice bit of Applied Science! That whole experiment sounds like those we have done at university(being MS in Applied Physics), with designing proper circut to measure nanoamps and math going there, gathering data point by point, looking for a way to make it more efficuent... Should I mention I loved lab physics classes the most in University?:)
One thing to keep in mind with the city labs spec is the 20+ year listed lifetime of the product. With regards to the halflife of tritium being a little over 12 years, I would expect their battery to offer nearly 4x-ish their spec at the beginning life of the product.
as an electrical engineer I was not laughing at you but instead commend you for getting the data any way. We get so caught up sometimes we spend more time making a circuit to take measurements for us we could have done it by hand and been done by the time we find the spool of op-amps that we know "should be right here."
I don't know how or why your channel slipped away from my radar, but it's bloody well back on it now! I somehow forgot just how cool the stuff you come up with is, this is absolutely amazing, and I honestly bet that someone who saw this has by now set up a bulk purchase of tritium vials and small amorphous solar sells and created a company selling nuclear photovoltaic batteries at 1/4 the price of betavoltaic ones. Just astounding that you saw something that should have been obvious to anyone in that field of study. daym.
What you could do with the device is listen to the atoms. What I mean is get some op-amps involved, do some signal conditioning, and output the signal from the device to a speaker. It would probably just sound like white-noise, but it's something to try if there's nothing else you can think of to do with it.
NurdRage, I think you missed that they have the device guaranteed to 20+ years. Now if by that they mean they can run circuits off their stated voltages up to those timescales they are putting a safety factor on their voltages, amperages, etc. that factors in the half life and technology degradation. Since these are critical power devices I would expect as much. So a lot of cost there to back up the claims and in the small package, may even technically be capable of matching your battery. This video was unreal by the way, very fun and creative; dropping a few squid on ebay now for tritium sources, just to mess around!
Excellent Work, very inspirational! Just a note, and most EE's will deny this, but only because the new knowledge hasn't flowed down yet: A solar cell actually does emit photons during normal operation. In fact, the best way to design a solar cell is to design it like you would an LED. At the cutting edge, scientists are now trying to design vastly superior solar cells using this technique/approach, and I am sure they will succeed in short order. Improving the reflectivity of inside walls of the enclosure will result in disproportionately more output power. Now, design me one of these batteries that makes a couple of milliamps at 3 volts for 20 dollars each and I will buy millions of them from you!
P.S. How to use this device: Connect it to a super-capacitor. This will create a "battery" with a huge discharge capacity that is always charged up when you need it (assuming intermittent operation). Here on the coast, it would be a fantastic battery for a hurricane light. We have to buy new batteries every season, and they just sit in our lamps and self discharge. All that the cell needs to generate is an incremental amount more than the super-cap's self-discharge rate.
A small LCD clock might work. I have no idea how much power they use; but it shouldn't be much. If you can power those off a potato then Dr.Lithium's device might work.
It is about output power, not voltage. A joule thief is a very inefficient boost converter that steps up the voltage at the cost of power. It would not work, you get even less power out of the battery that way.
Thanks for a very interesting video. I particularly like the two photo cell approach. Since you may be planning to use a small metal box to house your power cell, you might consider gluing mirrors around the perimeter to maximize output and minimize outside light sources which would vary the output of the cell. Also, have you considered other color options like yellow or white as a light source? I'm subscribed and urge others to subscribe too. Thanks again.
This battery could be used for keeping real time clock working and retaining some RAM memory over long periods. For example very old chip DS1307 uses about 300nA of current at 3V in order to keep its clock running. Newer chips could be more economical. There are extremely low powered chips for making various measurements. Excessive power can be used to recharge a capacitor to power up a small microcontroller for a short time to make measurements. This is especially useful in environments with very low temperatures like Antarctica, where sun light is not available for about 6 months and battery chemistry just freezes.
Philip Polkovnikov Sorry, an Rb oscillator needs several watts of power. That said, I don't know the power requirements for the new Chip Scale Atomic Clocks. I suspect micro watts is still not nearly enough, though.
You could look into low-power microcontrollers. like the ATtiny43U for example. It might be a cool project to use it as some sort of remote environmental monitoring station.
The ATtiny43U is a microcontroller that typically requires more power than 0.84 microwatts to operate. The operating current for an ATtiny43U can range from a few microamps to tens of microamps, depending on its operating mode (active, sleep, idle, etc.). To power an ATtiny43U reliably, you'd typically need a power source that can supply at least a few microamps to several milliamps, depending on the operational requirements of the microcontroller and any additional components connected to it. Given that 0.84 microwatts is an extremely low power, it might not be sufficient to power an ATtiny43U directly. However, if you're using energy harvesting techniques or intermittent power sources combined with energy storage components (like supercapacitors or small batteries), you might accumulate enough energy over time to power the microcontroller intermittently or in low-power modes. Always ensure you have enough power for the microcontroller's basic functionalities and consider strategies to optimize its power consumption or employ sleep modes to make the most of the available power.
I'm an EE and I'm not laughing. Sometimes you just need data. If you had gone through all that arduino junk just make a few load curves then I would have laughed at you. I work in high temperature thermoelectrics (~550C) and I have built several automated endurance testers that produce a ton of data. I still end up slapping together little circuits like this on occasion.
I'm no electronic engineer either, but as far as a circuit that you could build to make use of this technology goes I think it could be as simple as connecting the RPG in parallel to a decently sized low-voltage capacitor and then using it to power a very low power microcontroller like the EFM32 Gecko series from Silicon Labs. www.silabs.com/products/mcu/32-bit/efm32-gecko/pages/efm32-gecko.aspx I'm not trying to advertise the product or anything, I just found this page and realized that the chips have a low-power "stop mode" that requires only 0.6 µA at 1.98V and provides full RAM and CPU retention. I can't be sure because you didn't show a graph for the two cells hooked up in series but it looks like that would be just slightly below what the battery could provide. So with two batteries or just a slightly bigger battery or maybe just a more efficient microcontroller you could simply program it to put itself into a sleep or "stop" mode every so often to allow the capacitor to charge and then power the application from the capacitor which would allow for a much higher current to be drawn for brief periods of time. Just a thought. I know you can get development boards for those microcontrollers fairly inexpensively.
There are also some Atmel chips which can operate down to 1.62V and even an ATTiny one that has a built-in boost converter which allows it to operate on only 0.7V. Seems like the way to go if you really want to build that "eat a dick" device mentioned in another comment. ;) Edit: The EFM32 would probably be overkill, now that I think of it considering it's a 32-bit ARM core. I didn't look into it a whole bunch but I think those Atmel chips can operate at something like 0.35 µA per MHz at 1.62V so at 1-2MHz you should be well within spec for the double-sided cell.
I am just a tinkerer, and when I saw the title of this video I had to watch. Yes, you totally crushed my ideas of powering something with this technology. Thanks for the video, I found it very interesting.
I don't have anything super constructive to add to the discussion, but I wanted to say that I love this video and you should continue making more like it. As much as I love chemistry, watching the same "mix product and solvent, add this other thing, filter, distill, bam you have your slightly off white powder" gets kinda old after you've binged 3 channels worth in about a week. It's hard to find high quality science videos like this on RU-vid. Well explained, HD videos with understandable audio and well thought out processes are a godsend and I would love it if you would expand your channel a little more beyond strictly chemistry.
Hey Nurdrage, its good to see you actually trying to do things like the electronics, rather than just ignoring that bit. I'm certainly not laughing at you, and I think it's awesome. And keep up the good work!
[unsolicited counterargument ahead] I think that one of the major things causing the CityLabs batteries to cost so much must surely be quality control. Their batteries are intended for applications that absolutely cannot fail. Part of it might be that betavoltaic cells are more reliable than a slowly decaying phosphorus coating, but more broadly speaking I bet that every single component needs to be subjected to expensive testing to guarantee absolute reliability. At that price, they better be robust.
oh absolutely. They're also selling to military contractors. Meeting certifications for reliability in Marine, Aerospace and Military sectors is not cheap. We cheat in that we're building the device itself and not certifying it for anything. To physically build the city Labs device might be just as cheap, but it's all regulatory stuff and certifications around it that make it expensive.
Photovoltaic cells do work like LEDs and produce some light when powered but not in the visible spectrum. It's usually a (very weak) near-infrared glow.
Ever since reading about a "beta cell" in a 60s era chemistry book I've wanted to build one, this is probably about the closest I could get. The schematic of the cell I saw was effectively a capacitor that charged itself from a beta source, and supposedly generated very high voltages at negligible current.
+Sci-Twi don't think that would ever happen at my public charter school... had a teacher joke that these strang things i bring to school are bomb-making material.
Atomic pacemakers used to be a big thing but then they fell into disfavor when long life lithium batteries came to market. You likely don't have an atomic pacemaker. Nonetheless, there is some resurgence of the idea among medical device makers because they are extremely reliable and last the lifetime of the user. So the nuclear battery concept is still being explored.
NurdRage phew. Guess going nuclear will just be rage rather then literally. Thanks for replying. By the way. I like your explanation process and great in depth research. Keep up the great vids man!
In the kind of specialist devices that chip gets used in I'm guessing 3 grand is chump change anyway. I used to work for a reseller selling expensive mini computers - whatever space wasn't used in the full rack by the computer, disk arrays etc you could fill with blanking bars. These were just bits of plastic that were like £100 for a pack of 30. This one customer bought 30 packs by accident - so they spent £3000 on blanking plates. When we informed them of their error they didn't care, they said, "meh, keep em", didn't want a refund or anything.
I'm not an electrical engineer, but from what I gather from the video for a little more cost you could boost your power output by sandwiching a double sided mirror in between the Tritium cells and then use the solar cells to each side. Completing the project by placing completed cell into a dark case of some kind to finish it out. In theory, you would get the power gain from the reflection of the light as from the use of one solar surrounded by foil. Therefore, in theory, it should double the output.
