Make a Joule Thief (2002)
bigclive.com
bigclive.com
Big Clive is an electrical engineer from The Isle of Man and has a very unique English accent, complete with local slang and idioms. He’s a very engaging character and passes as entertainment, even without the electronics content. Genuine and occasionally heartfelt and up-lifting. :)
https://www.youtube.com/watch?v=DWRyIShLRO0
They are the Tappet Brothers of the 21st Century.
- photonicinduction (warning, absurd levels of everything up to 11, borderline insane at times, but if you want to see ultra high everything, here's the channel)
- styropyro, spiritual son of the above, but for everything lasers
They were way overselling its benefits, claiming you would get something like 6 times as much runtime out of your batteries, and it got a lot of negative reviews and debunking at EE tech sites.
I wish someone would make one of these with that sleeve approach, not as some sort of miracle energy extended scam but rather as a voltage curve adjuster.
A common non-rechargeable alkaline battery starts at about 1.5 volts and over its useful life drops fairly steadily to a little over 1 volt, and then rapidly drops to near zero.
A NiMH rechargeable starts at around 1.4, fairly quickly drops to about 1.3, then over most of its life drops fairly smoothly to about 1.2, then starts dropping faster to around 1.1, then rapidly to near zero.
This is why you can use NiMH in devices designed for alkaline batteries even though nominally alkaline batteries are higher voltage. The device has to actually be designed to handle 1 to 1.5 volts, and the NiMH is in that range for nearly all of its discharge curve.
This is also why some devices designed for alkaline report low battery on NiMH long before the batteries actually need changing, and still report low but useful battery level right up until the device stops working. They are estimating battery life by looking at the voltage and fitting that to the alkaline discharge curve to estimate how far the battery has discharged, and it makes the batter level meter on many devices close to useless if you use NiMH.
It would be great if there were a sleeve you could put around your NiMH batteries that would dynamically raise or lower the voltage as the battery discharges to make it match the alkaline curve. Then your battery level indicator on your devices that were designed only for alkaline would work.
I'm not sure if an NiMH sleeve was ever practical given the space constraints, but it's kind of moot now that better technology exists.
You can get around this[0] by starting the power converter in series with the battery (so the load pulls current through it to start it up) although that's: a: complicated, b: more conversion overhead when it is running, and c: rather difficult when you only have 1.5 volts (or 1 volt) to work with initially, compared to eg a 9-volt or 12-volt battery, which exacerbates problems a and b.
/-[starter]-+---\
ˍ˖ | | | |
Ξ [regulator] [LOAD]
| | |
\----+----------/
Stay away, they are a well known and debunked scam.
https://www.indiegogo.com/projects/batteroo-reboost-make-rec...
https://en.wikipedia.org/wiki/Batteroo_Boost
https://www.eevblog.com/forum/blog/eevblog-751-how-to-debunk...
Don't miss this one!
Unfortunately, I don't think the kit I built is still available.
Here's some information I could dig up:
- https://www.gigaparts.com/etow-humanalight.html
- https://aa7ee.wordpress.com/2014/08/18/the-humanalight-a-fla...
NiMH is a lot safer, cheaper, and has slightly better capacity. It's been a few years since first available, but any posters even heard of anyone using 1.5V Li-ion cells? Once you know enough to even be aware of them, you'd be using 3.6V Li-ion instead, so really it's the 1.5V Li-ion that was instantly obsolete, yet still somehow hangs on due to extremely specific circumstances of boutique users that want a flat-regulated 1.5 volts.
NiMH is safer, and cheaper for now, but those factors are irrelevant when 1.2V is insufficient for a particular device. If a device accepts 3.6V with overdischarge protection, then regular Li-ion makes sense, but there are a lot of 1.5V-3.0V AA/AAA devices out there.
Great question. Yes. At least I was, until I just found some newish 1000mAh 14500 cells. Last I knew, 800mAh was the highest capacity around. Ironically, I think in a lot of cases the 1.5V Li-ion cells use Wh rather than Ah to obfuscate the fact that the cell has less considerably less capacity than a 30¢ alkaline. So there is now about the same energy in an AA 1.2V NiMH cell as in a top capacity 14500 3.6V Li-ion cell. But this doesn't stay true for the lifetime of the cell as opposed to a single charge, for the most part because NiMH cells can tolerate 4X the recharges before resistance gets too high.
> NiMH is safer, and cheaper for now, but those factors are irrelevant when 1.2V is insufficient for a particular device. If a device accepts 3.6V with overdischarge protection, then regular Li-ion makes sense, but there are a lot of 1.5V-3.0V AA/AAA devices out there.
In hindsight, the observation I made does not apply much beyond flashlights. You're right, a 3.6V cell is not going to help someone whose device takes 2xAA cells without Dr. Frankenstein's assistance. However, if 1.2V is insufficient to power a device that was designed around alkaline chemistry, then it's a problem with the design, not the chemistry (granted, I have heard stories that NiMH doesn't work in some devices, but this wasn't most devices nor devastating to consumers). Alkaline cells aren't 1.5V for very long, it's a pretty steady discharge curve from ~1.65V down to 1V. So a device with such strict power parameters designed around drawing power from alkaline cells, without taking into account the discharge curve of alkaline chemistry, is thus poorly designed and wasting upwards of a third of your batteries' capacity (if 1.2V no longer works, and an alkaline cell is fully charged at 1.65V).
I think the only exception might be in ultra low power devices which last over a year on a single charge. But even then, I am likely to use the more expensive NiMH, because I'm so fed up with leaking alkaline batteries.
I've got a pair of first generation AAA Sanyo Eneloops in my bathroom scale. They've been in there for 1028 days. I have a pair in the display unit for a pair of wireless fridge/freezer thermometers. That went 395 days.
I've got first generation AA Panasonic Eneloops currently at 1079 days in one of the fridge/freezer thermometer sensors and 978 days in the other. 539 days in a humidity meter. 544 days in an analog wall clock.
1: https://www.tindie.com/products/sdp8483/multi-led-joule-thie...
Joule Thief (June 17, 2021) https://en.wikipedia.org/wiki/Joule_thief https://news.ycombinator.com/item?id=27540614
How to steal every last bit of juice from a battery (April 28, 2015) http://aeguana.com/blog/the-joule-thief/ https://news.ycombinator.com/item?id=9451484
Joule thief – A minimal boost converter (January 25, 2015) https://en.wikipedia.org/wiki/Joule_thief https://news.ycombinator.com/item?id=8941971
It might be suitable for miniaturization.
The Joule thief charges an inductor in one half of each cycles, and let’s it discharge into the load in the next half. Inductors like to emit a constant current and will kick up the voltage to maintain the current, so it’s more like a current source. Because it drops to 0 quickly before the transistor starts charging it again, it delivers its output as a series of pulses.
So it’s a pulsed current source. The only way to get a constant voltage out of it is to 1) feed the current into a fixed resistance load 2) filter the output via a capacitor.
I thought about how to make a single transistor voltage source but I gave up. Transistors are so cheap there’s not much point. The world has a bizarre preference for voltage sources (maybe due to the fact that early batteries were DC, or maybe it’s just Edison’s ghost). Most real-world loads have useful outputs proportional to current (LEDs, magnetic fields) and the voltage is just a secondary variable.
Curious now whether some of the little garden path solar cell lights use these to get maximum output from 4-5 square inches of cheap solar cell and a small nicad battery.
I was hoping, from the name, for something that would pick up ambient RF and deliver a milliwatt or so. That would be useful for things that need minimal power and are a pain to connect to a power source.
Such things have been around for decades. There's an Instructable for this.[1] Newer designs try to use more of the available spectrum.[2]
A whole milliwatt is a lot to hope for from ambient RF (you can do it if you get lucky or make the device big enough), but it's easy to get from solar cells. The purple amorphous kind are less efficient in sunlight but provide more power under the dimmer lighting we use indoors.
RIP, Maplin. Glad Big Clive's still alive and kicking.