Inside a Nokero solar light (with schematic)

This isn't a standard solar light. It's got very unique solar panels, and even more unique circuitry that makes it very different from ordinary solar lanterns.
This light was sent by Steve at Nokero (No Kerosene) a few years ago, so hopefully the circuitry is still applicable. The Nokero website appears to be down now, so I'm not sure if they're still making these.
The reason this light was designed is to provide safe illumination to parts of the world with no reliable electrical infrastructure. It uses good quality solar cells and a LiFePO4 battery that potentially offers around 2000 charge cycles with low loss of capacity over time.
The most interesting parts of the circuitry are the current regulator, which took a lot of deducing to identify. The use of a Texas Instruments microcontroller that can monitor voltage accurately for circuit control, the unusual charge control technique of using the LED as a shunt load, and the very efficient use of a MOSFET as a very low drop alternative to the more common solar panel diode.
The solar panel's transistor is only turned on when the panel's voltage exceeds the lithium cell voltage. It's effectively being used as a software controlled zero-loss diode (Active rectifier).
That suggests that charge limiting is mainly based on the LED being used as a shunt load.
The use of a LiFePO4 cell is notable, as it is safer and more rugged than the typical rechargeable lithium cells used in most phones and appliances. Although the technology has lower energy density than the classic 3.6/4.2V lithium cells, it offers greater capacity retention over its lifetime, and a much less volatile electrolyte.
If you enjoy these videos you can help support the channel with a dollar for coffee, cookies and random gadgets for disassembly at:-
www.bigclive.co...
This also keeps the channel independent of KZbin's advertising algorithms allowing it to be a bit more dangerous and naughty.

[@WineScrounger]

I like that it’s easily repairable. That’s really important. When the cells finally go, some enterprising chap can make a living replacing them. Repairability is vital in developing countries because it encourages self reliance rather than fostering dependence on first world suppliers.

[@danytoob]

@@ziplockk You would enjoy Louis Rossmann's Right to Repair quest. Important work to be sure.

[@rexsceleratorum1632]

This is not easily repairable unless you mean swapping out extremely niche proprietary boards no one will ever manage to source in some remote village. Because here in the tropics the board will likely rot away before the cell dies. The hole is good for water drainage but also good for insects to get in. The early Chinesy LED bulbs with ventilation holes were practically insect traps, give it a week and the diffuser dome would be filled with dead insects.

[@rexsceleratorum1632]

@@ziplockk Who is going to do the conformal dipping, remote villagers?

[@rexsceleratorum1632]

@@ziplockk The original argument was that this design, as it exists, is suitable for 3rd world, usually tropical, countries. I pointed out a couple of design flaws for tropical usage. You are saying they can be fixed by the manufacturer, which is not a defense of the original argument.

[@ianleitch9960]

I llike the lack of flashing features; so practical.

[@alexstone691]

I haven't seen a single use for the annoying flashing

[@Pickleriiiiiick]

Noo!! The sos flashing feature would totally save a family in africa!! What are they to do without this completely un annoying feature!!

[@seanb3516]

Who doesn't want a blinking light to read by? Are you pulseshaming the poor defenseless lamp? YOU ARE AN ANTIFLASHIST

[@miklov]

I agree. My bicycle lamps all have blink modes I have to cycle past to turn them off and once oxides build up between the battery connectors they start to change mode when you go over bumps and then you become one of those annoying people with blinking bicycle lights >.< (edit: typo)

[@rogerhargreaves2272]

@@miklov yup, that’s annoying. On/off is my preferred option, but they are hard to come by.

[@keithking1985]

the regulation question could might just be what you said about some places having blazing sun shine. seeing that they are made for such places with more intense sun shine it makes perfect sense. And it seems/is VERY well made. so i wouldn't doubt it at all thats whats its for. your a clever boy Clive ; ) not many would have noticed that, and id bet my gaff your right... im going to get one. actually im going to get two and have the other sent away to some kid that really needs it, like you said at the start of the video..

[@seanb3516]

Hey BCDC, to see the circuit lines on the board you could try freezing it first and then while warming use your BV9900 to take an IR picture. The different heating rates can provide thermal contrast (hopefully). More extensively you could use an inductive field to heat the metal components of the board and again use IR photography. As a bonus we will be entertained watching the PCB melt plus the ensuing fire... XD

[@umxrr8677]

Great idea, but it could potentially damage components

[@seanb3516]

@@umxrr8677 I believe enough thermal differentiation could be developed to use the IR camera simply by cycling the lamp PCB through its operating range. In this way the unit is not stressed beyond its design limitations. Any attempt to move outside the range (just a SWAG here folks), -40°C < unit < 40°C , would undoubtedly cause irreparable harm. Another approach is to provide a gentle heating or cooling source under the board and due to the effects of differing rates of heat transfer and dissipation (Entropy & Enthalpy, might have gotten terms messed up) through the board it would still be possible to get the photo. [TMI Disclosure: I am making these statements having never tried it despite the fact that I also have a BV-9900 Pro, in my housecoat, drinking early morning coffee, vaping nicotine and our completely legal honey oil all while consuming some BCDC vids. NO TOUCHING!! :D Good morning world!

[@seanb3516]

@@umxrr8677 Yup. My point exactly. :D

[@railgap]

One word: x-rays.

[@seanb3516]

@@railgap I was going to suggest a Positron beam from a particle accelerator. We have one 15 minutes from where I live (TRIUMF). Might not work but damn fun to try. [side note, my gf's dad worked there more than 20 years so I have seen things almost no one has. Visited the area contaminated with radioactive Carbon. Also a new Hot-Box room no one was allowed in. Kewl]

[@gabrielribeiro3166]

Me not knowing much about electronics: “Ah yes, that is indeed very clever by the circuit 🤔”

[@sammy61187]

I know nothing about electronics but after watching clive i feel incredibly smart

[@kapegede]

2:28 Aaand: After you get rid of the clip you can screw the lamp into a lightbulb socket with a E27 thread and replace a burnt lightbulb or having light from the ceiling during a blackout.

[@pear7777]

Right, saw that too

[@phils4634]

Aldi occasionally sell pairs of "strip fluorescent equivalent" COB LED lamps. Ideal for hanging off anything, and with magnetic clamps for mounting on anything ferromagnetic.. Obviously not solar, but easily charged (5v USB connection) and an all round excellent emergency / standby light.

[@ssss-df5qz]

Imagine the two metal bars being live from the socket to charge the thing >_< Nothing would surprise me.

[@SytheZN]

Maybe that 1K resistor was chosen to handle the case where the cell is overdischarged below the MCUs cutoff threshold? Would allow the cell to trickle super slow until the voltage comes up to a level where it's safe to start a higher current charge.

[@Ghlargh]

This is almost definitely the case. It's just implemented in an odd way since it's integrated with the voltage measurement and not a separate resistor and diode (the current design does save on components though, if you can trust the ESD diodes)

[@kevinmalec4977]

I had the same thought - that path is necessary if the battery is empty. This might also be why it doesn't ever obviously disable the charge input

[@henryokeeffe5835]

That 1k resistor has no effect at all on this. Clive didn't notice the MOSFET underneath the solar cells has a body diode that would do exactly the same thing, but with a slightly lower voltage drop (as a power device), and without a resistor at all.

[@markb4071]

Once again, Clive highlights a worthwhile cause in his own way. I bought a load of Aspods from that video and i'll go and donate a few of these through Nokero now too - Great idea.

[@K-o-R]

If it wasn't for the hook at the bottom it looks like it can mount into an E27 fitting.

[@Robvdh87]

You beat me to it with this comment. I was wondering why you would have to use this light in a e27 fitting because that indicates mains electricity. But it would make sense in a generator powered house/village (no kerosine use at night) or in a warzone were the power went out.

[@cryptickcryptick2241]

@@Robvdh87 It could also be handy for load shedding. In many countries with poor infrastructure too much demand and not enough supply, they turn off power to an entire section of town at a time and then rotate through them every few hours. In a case like that you might find in easy to unscrew the regular light bulb, then screw this into that lamp to hold the light. Myself, I would think most people would simple put a string on the light and hang it from a nail. But that does not require and additional design up front. (A string can go through the clip.) While giving the option to hang from a e27 does give some additional value at times.

[@WineScrounger]

@@cryptickcryptick2241 exactly, it’s another use option that cost no extra to make, so why not? I can see the utility in it and it’s another indication of how much thought has gone into the device.

[@FroggyMosh]

It also gives the user an out towards manipulative/possessive people. Just screw it into an old desk lamp, play it off as a bit childish, and your bully brother might just leave it alone for a while.

[@johnmorgan1629]

Thanks for leaving a link to the company in the description too, for those who might want one or to donate to them.

[@seannot-telling9806]

This is one of those worthwhile causes. Thanks Clive.

[@stonebranson3364]

you have taught this old bloke a lot over time thank you man

[@collegestudent6071]

We have reached peak enlightenment. Using light to generate electricity to generate light.

[@laurawerner]

I didn't know SunPower made tiny solar panels, but I have big SunPower panels on my house. They're great! When I bought them ~3 years ago they were more efficient than the usual Chinese import panels in watts per square meter. They were also more expensive, both in dollars per panel and dollars per watt. But the higher efficiency let me use only the south-facing parts of my roof, so I'm using the panels as efficiently as possible. (If I'd gone with cheaper panels I'd have had to use a west-facing part of the roof, which is a lot less efficient, and would have needed a lot more panels.) So far, I've been very happy with the SunPower gear. It seems well-made, and the efficiency hasn't dropped off very much as the panels have aged. (They're still fairly young, though.)

[@McTroyd]

Thanks for this Clive!! Just bought one, and threw a donation in the cart.

[@kittenisageek]

I agree with you, Clive, about LiFePO4. I think its an amazing technology. Compared to other Lithium batteries, LiFePO4 are quite resilient and fault tolerant. I have a 15-year-old 100Ah battery (4 cells in series for ~12v) that still has close to 90% of its capacity, even though for a number of years it was continuously charged by a 2A lead acid charger (I do not recommend charging that way, but it was before I owned it). They're extremely efficient when charging, have a very low leakage rate, and do not get hot when used within their design specs. In fact, charging feels like it cools the cell slightly. Their only "real" downside, and the reason you don't see them in cell phones, is because of their energy density. An 18650 LiFePO4 is usually between 1200mAh and 1800mAh, while Lithium Cobalt 18650s often run about 3400mAh in capacity.

[@Cablesmith]

Excellent cause Clive 👌🏼 thanks for making me and many others aware of this 😊

[@smartgorilla]

Love seeing these things and less throw away stuff.

[@smartgorilla]

you are an amazing man

[@OntologicalQuandry]

The two transistor circuit is one I have used many times in my designs, but I put a n-channel MOS-FET in series with the LED and sense resistor so I can get a sharper knee on the current regulation. However, your point about the 0.6V headroom is right except that you also need to take into account the 25mV/°C temperature coefficient of the transistor (if you want to be very precise).

[@stanleydenning]

A lot of love went into this design.

[@reinux]

Not very often that you can say that about an LED light, but I think that's the best possible way to describe it.

[@rpavlik1]

Interesting gizmo. The discussion of cell types reminded me of something. The kiddo got a pretty cool remote control car for Christmas - super light and fast, foam tires. Looked more carefully at it: it's LiFePO4 powered! Made me smile and not worry about it as much, more likely to burn out the motors than the battery. Any kind of lithium ion battery in RC cars is brilliant, sure beats waiting hours for NiCd to charge, and then play for 5 minutes. Charge for 30 min (yeah I'd prefer 1hr to not exceed 1C), play for 15 to 30 minutes, kids these days are spoiled. (Also got one that came with a pair of actually 800mAh 14500 cells. What kind of thing comes with a spare rechargeable battery?)

[@jeffreyallen1290]

I have two of these lights I use when camping. I love them.

[@dcallan812]

interesting light idea idea of buying one for someone who really needs it. Nice.

[@74HC04]

A couple of thoughts. LiFePO4 cells have voltage limits of 2.0V to 3.65V according to Wikipedia. The MSP430G2230 microcontroller has an upper absolute maximum supply voltage of 4.1V. The LED illuminating when fully charged could be there to a) prevent the cell voltage rising above 3.65V but also b) Prevent the MSP430 supply voltage rising above 4.1V in the case of an open circuit LiFePO4 cell.

[@fredflintstone1]

I am fairly sure that the flashing LED is just to let the user in the poor country know that the light is charged, if they leave it out and it doesn't flash during the day then they know it has not charged fully?? I am though very pleased it doesn't flash SOS :-)

[@EasyOne]

Very well explained

[@twocvbloke]

I'd be happy to have big phones again with a battery that lasts, heck, it's tempting to make a new battery pack for one of my 90s Motorola phones with LiFePO4 cells, just for the halibut... :P

[@RobertSzasz]

Hmm. I wonder if there are any projects for making a SDR cell site for very old cell phones (briefcase style etc)

[@Sashazur]

Moto G Power phone, 5000 mAh battery. #1 battery life in Tom’s Guide tests.

[@kardeef33317]

I think the 1k resistor is so the micro can turn on the mosfet below the 1k resistor so it can be access to the negative rail in case the battery protection cuts off the power from the battery so the micro can power itself from the solar cell. Just my amateur guess. As a guess do you think iam on to something?

[@martinwinfield2935]

Thanks Clive always in depth and well explained. Always a good video.

[@perikon]

Its quite a clever design. The solar cell mosfet has a parasitic diode in it and can only cut off the solar cell when the cell voltage is lower than the battery. When battery is fully charged and the cell make the battery voltage +0.6 the diode in S2 will conduct (and the one in the microcontroller too). To prevent overhanging it just burns energy by turning the led on. But that is okay as the battery if full and its bright light at this moment.

[@chrishartley1210]

I have a couple of theories about the constant cycling. Firstly, it might be to keep the battery warm in low temperature applications, low temperature charging is a battery killer. Charge/discharge will keep the temperature up where charge/cut-off would allow battery temperatures to fall below "safe" operating range. Secondly, it might be to prevent the formation of Fe2O3, it can cause micro-shorts which is also a battery killer. Charge/discharge may prevent the formation. I only heard about this one today, so it's just a guess.

[@salmonline]

"it comes on itself, at night", Clive?🤣🤣🤣🤣🤣🤣😉👌

[@fuzzybobbles]

I come on myself at night too. What's the problem?

[@JUANKERR2000]

@@fuzzybobbles Keep some tissues under your pillow.

[@lofman]

@@JUANKERR2000 Just use the pillow. Switch side every other night...

[@BedsitBob]

@@fuzzybobbles Much more fun, if you come on somebody else.

[@badWithComputer]

My cousin used to do that a lot, we didn't invite him to our wedding

[@wafikiri_]

​@bigclivedotcom Clive, at 11:30 you ask what style of drawing MOS-FET transistors we prefer. I suppose you ask whether we prefer an encircling line or not. It is not a question of preferences: the encircling line represents the element's packaging, so that an integrated circuit with three transistors will have one, not three, encircling line in its drawing to include the three of them, and a triode-pentode valve will have also just one, not two, encircling line around the conjoint representation of the triode and the pentode. Individual transistors, whether bijunction, MOS-FET or whatever other possible kind, should each be individually encircled, whereas they should share drawing encircling if integrated in just one physical package such as an integrated circuit.

[@piconano]

I like this light a lot. Great video. For many MCUs, the input impedance to the ADC has to be less than 10k Max. I think the 1K is fine. If you put a 100K instead, the sample and hold time has to be increased and ADC will become out of spec. So instead of 10 bit resolution, you'll get closer to 7 bits. The 2 resistors and one capacitor, were not necessary to debounce the switch, but they might have been short on flash memory to fit more code, or just didn't wanna be bothered with reading the port 50mS apart, to see if the switch has debounced. The watchdog timer makes sure no hanky panky goes on, and will reset the MCU in nanoseconds. However, I'm not clear as to the purpose of the 47K resistor in parallel with the solar cell. Comment if you know.

[@williamsquires3070]

If Bigclive took a Tesla to bits, would he find a capacitive dropper inside? 🤗

[@TheHutchy01]

BigClive could take a mars bar apart and find a capacitive dropper between the chocolate and caramel.

[@frogz]

@@TheHutchy01 ............actually that is a diode junction, A CH-CR junction is a boundary or interface between two types of semiconductor materials, chocolate-type and caramel-type, inside a single bar of mars. The "CR" side contains an excess of holes, while the "CH" side contains an excess of electrons in the outer shells of chocolate. Candyopedia

[@fazergazer]

Clive, I learned more in this one video! Amazing.

[@johnmorgan1629]

You ask how to draw MOSFET's I say your the one doing it. So do as you please and carry on making quality vids that's all that matters to most of us.

[@midnight_perdita]

I'm just here for your relaxing voice

[@6581punk]

I have some mountain bike light batteries from 2008 using LiPo that still work.

[@SproutyPottedPlant]

I use LifePO4 18650s in my mechanical squonker vape mod, no problems with them for 2 years other than the usual torn wrap that I get repaired, the voltage doesn't drop like normal lithium ion and they last all day! I use an MTL coil about 10 watts.

[@peterkis4798]

I done the same type of charge regulation for my outdoor esp8266 weather station. Simply if the voltage is to high then the esp stays a week until it drops(or timeout). Keeps the lipo battery voltage around 3.8v. It's running over 2 years without any problems, sends the data every 10 minutes over espnow.

[@v8mufflerboy84]

I bought a 6 pack. I intend to give them to Trumpers around me as as teaching point. There are no small parts to swallow are there? I will keep the lanyard. I may go back for the floater, it looked nice But I am not aware of a clear plastic that can handle UV.

[@pspkutube]

very nice Prof! That extra R at the MCU input is likely a part of a calibrated RC filter ... interesting to see the LED being used for charge regulation.... I wonder if that constant switching on/off reduces the lifetime/output of the LED...

[@NeilWhelchel]

The reason for the MOSFET between the battery and solar panel is not to cut off the charge current when the battery is full, the body diode of the MOSFET will charge the battery even with the MOSFET off. (As you pointed out , there is another indirect charge path through the input protection in the microcontroller.) It is used as a switch to prevent battery discharge through the leakage in the solar panel when it is dark. When the battery has enough voltage for the microtroller to come alive and the input pin with he 1K from the solar panel is low, it turns the MOSFET on so there is almost zero loss in the charge circuit. It is likely that the software will turn off the MOSFET for a tiny period of time (multiple times per second) to monitor the solar panel voltage and if it is producing power, turn it back on, if not leave it off until it starts producing power again. The reason it is a MOSFET instead of a diode is because it lacks the bandgap voltage drop that even a schottky diode has, it is purely resistive at just a fraction of an ohm. (See also active rectifier.)

[@petert.naujoks1572]

The 1k resistor is there to protect the processor from too much current, right. The main current goes through the parasitic diode of the second MOSFET - so this FET could be used to switch from "fast charging" to "triple charging", it can not turn charging off. So the end of charge solution to turn on the LED periodically is the only way to "stop" the charging. Also it could be used to signal to the user that charging is done. In that way the whole device is not longer than necessary in the full sun and probably heat.

[@ikbendusan]

along with other people who have mentioned the mosfet parasitic diode, there is indeed a diode on the analogue input pin of the microcontroller. it is for ESD protection and is in fact reverse biased (as you have drawn it); only a tiny leakage current (nA) would flow through that, so no such charging event can occur

[@MrEtonmess]

I have no idea what Clive is talking about but these videos are so soothing.

[@MetallicBlade]

They're nice, but I'd still keep the kerosene lamps. These will probably last only max of 5-10 years before either the battery or LED itself craps out. But if they're cheap enough, then that might not be as bad of a compromise.

[@mikerhodes9198]

Good video. Can you please investigate the Yinlong LTO battery which supposedly can rapid charge and discharge 100000 times without issues and can operate at - 40. Would appreciate your take on this cell.

[@stridermt2k]

That seems like a very cool lamp.

[@nmosfet5797]

The idea of the 1K resistor to the MCU is that in case of complete loss of charge from the cell, the MCU is the first to wake up, its voltage goes higher than the battery cell which is fed via the internal protection diode, so the MCU can wake up early and start regulating the charge. And yes, the MCU must regulate the battery charge against overcharging, and thus switch on the LED to drain the battery when it is reaching critical high level. A very clever design, quite bulletproof, recovers from failures such as throwing it into the mud for a week so that it completely draing, and then cleaning it up, it should work fine. Respect.

[@gd2329j]

Interesting little lamp . Could the diode get things running from a totally discharged state ? Like when the mcu can't drive the mosfet or is just not running at all . Wonder what the decompiled microcode looks like . Don't all say ones & zeros …............

[@djctech2586]

I'm a bit surprised to see a fully linear current regulator. Cleaner on the radio bands, but burns a little power a switcher could have avoided.

[@1kreature]

I was hoping for a test of how much the short circuit current of the sunpower based light was vs the old knockoff.

[@bigclivedotcom]

I was lacking sunlight for a proper test.

[@TheEmbeddedHobbyist]

My vote is for the one with the intrinsic diode shown, which might have been the 3rd fet drawn if there was one.🤣

[@RobertSzasz]

Oblique lighting can really help seeing tracks on those boards

[@beargun42]

The 1k resistor may help the circuit come back to life, if the battery is every fully discharged. Since the MCU needs to turn on the mosfet for the solar array to start charging.

[@bigclivedotcom]

That's possible.

[@misterhat5823]

Not true. Another commenter pointed out the intrinsic diode in the MOSFET will allow charging when the MOSFET is off.

[@pum882]

@@misterhat5823 I don't understand how a reverse biased diode can turn on the charging

[@misterhat5823]

@@pum882 Because when the battery voltage is lower than the solar cell voltage it isn't reversed biased.

[@BobMuir100]

Now Clive didn’t follow a fair bit of that (yet) but it seemed both simple and clever in areas?! Good video mate Bob

[@LOTPOR0402]

Yeh but will it last longer than 1 season or end up in the bin ? .Where most of these garden lights end

[@ATMAtim]

This appears to be a very well thought out device. Pretty cool looking design as well. It almost appears have a threaded base to screw into a standard bulb socket. Interesting thought.. Where do I sign up to have a box of these sent to me? My power bills are ridiculously high! 😎

[@uwezimmermann5427]

the clone looks pretty much like a previous generation Nokero. I have been using their buy-one-donate-one out of curiosity during their first two generations. The quality had improved, but was not comparable to the one you showed here - I like it! The ring around transistors is unnecessary and even irritating... don't do it... we even teach it without...

[@jussikuusela7345]

Back in vocational school, I was taught that the ring marks a discrete component, opposed to one inside an IC or a potted module.

[@peterzeboroff7448]

I wish you had the equipment to be able to tap into the controller modules in all the various switching power supplies and l.e.d. drivers and be able to display and analyze the firmware that drives all these type of microcontroller based regulated power supplies and all the switching type battery chargers. The firmware that controls all the various functions may be quite complex and it may even be able to alter or reprogram these devices and put your own parameters in. I have always been very fascinated with switching power supplies and the almost magical way that they work and how they are able to produce such a large amount of current and power given their relatively small size.

[@anthonycampbell1726]

If you leave it in the sun it blinks when it's charged. Could just be the it's set to over charge then discharge directly in the programing? or it's programed to overcharge and activate a separate dump, that's why the input is more precise than the output?

[@alexfedorov1160]

4k7 resistor is probably placed to isolate programming data pin from debouncing cap. And in the "proper" mosfet, as it drawn on schematics, the parasitic p-n junction between substrate and drain would pass charging current.

[@Miasmark]

The N193 lights appear to be the ones that have "automatically turns on and off with the sun" features. If you go under "all products" it looks like they sell in 6 packs. 10 Lumens seems awfully dim though.

[@sayyidsahal4533]

Even after being in electronics tinkering for 5 years I even don't know a simple constant current circuit, tanks, Clive for that little circuit

[@snowdaysrule]

I have a laser pointer driver board that I was trying to reverse engineer and the current regulating ic on the board has a voltage divider on its voltage sense pin and I couldn’t figure out why the divider was there. But after watching this video I believe the divider is present so that the output current could be changed with something like a potentiometer or an external ic pulling the end of the divider down to increase the output current like this circuit is doing. So thanks for helping me solve the mystery :)

[@curiouscrandall1]

I really think that that AP4312 is unnecessary, being as you have a microcontroller with ADC inputs. Surely, you could connect the top of the 3R resistor to an analog input (with a smoothing cap). Now, it's only 70mV, but the ADC is 10-bit, so it has 2.44mv/count, so you get 28 counts at the desired current. The microcontroller output could then be used PWM-stylee to maintain the desire output current.

[@dustysparks]

That "burning off of the top level power" from the cell to keep it from over charging reminds me of the techniques that BMS' use when maintaining packs of many many cells in series. When the highest cells get to the topped off level, resisters kick in to burn off the excess until the lower cells get a chance to catch up (this process is called "top balancing", and only kicks in when the pack reaches near the full charge).

[@bigclivedotcom]

I've been looking at balancing circuitry like that. Often using a DW01 chip to switch on a shunt resistor.

[@dustysparks]

@@bigclivedotcom Have you seen the "capacitive" balancers that rob power from the high cells to bring up the lower cells? I'm really curious to see that circuitry. Those can be used while charging to boost the low cells quicker with the incoming power to the high cells. If you use them without charging, they average out all the cells to an average (minus losses, which are still far less than just burning off the excess deliberately as heat).

[@bigclivedotcom]

@@dustysparks I've not come across that type.

[@dustysparks]

@@bigclivedotcom kzbin.info/www/bejne/Y4PJfn6mrq2eo5Y I have one of these that I'm testing (different brand and slightly different configuration), and... my results are inconclusive as my cells are already pretty balanced in the packs I have (LiFePO4 26650 packs).

[@MrKillswitch88]

You could look into those Sunstream Pro solar panels with power bank as that works better than lights like this though not cheap, the panels don't perform as rated but are pretty solid compared to the usual plastic that is the norm. As for the power bank they come with the light is ok as a flashlight but the cool factor is they use a 18650 cell but good luck getting it open as the body is aluminum and they over tighten the end caps like mad.

[@singlesteve2011]

it is a nice thing to hack also, but could you do a video on the different types of resistor's and capacitors ?

[@Mr.Unacceptable]

I got in the habit of underlining numbers as well as capitalization when writing so I don't confuse them. 1's look like L's. 6's look like G's. 0's look like O's. So I put a line through my zero's. Especially for passwords.

[@matthewbeddow3278]

Nice cleaver construction. It is odd that we don't see more use of the LiFePO4 batteries in products, I am sure there will be some unscrupulous reasons for this.

[@jarno.rajala]

LiFePO's have much lower energy density. In phones, laptops, cars and such, where size & weight matters, the longer lifespan of LiFePO is not a good tradeoff.

[@matthewbeddow3278]

@@jarno.rajala I'm sure if someone invested time & money into the platform it could be improved, this would be then a direct competitor to other battery systems where powerful people have billions invested. I do agree though you are totally correct.

[@jarno.rajala]

@@matthewbeddow3278 Battery chemistries are heavily researched and LiFePO isn't fundamentally different from other Li-ion chemistries. It would be easy to switch to LiFePO, if it made sense to do so.

[@kjur18]

Dave done some explaining that you can power microcontroller via any pin and it's esd diodes. I have a electronic stapler, one that needs 4 D cells. I put inside two AA sized LiFePO4, and it works great. I think this type of battery has different discharge curve, keeps 3.2V for most of capacity.

[@bdf2718]

Acorn found that out with the first development board they put their first ARM CPU in. It ran exactly as expected. They later noticed the power rail wasn't connected to the chip. :) Only works if the power consumption is low enough, otherwise you can fry the ESD diode.

[@alexstone691]

The lazy mosfet drawing is the best, there's also laziest one, the digital symbol which is more logical when you look at it

[@piotr86]

You forgot about the diode hidden in the mosfet, so the current flow through the resistor and microcontroller pin will not occur.

[@FindLiberty]

Correct sir or madam; extra credit points awarded. You are hereby released from the forced discharge wrist-strap handicap. Go forth and work carefully/safely...

[@misterhat5823]

Which means that the MOSFET cannot stop charging... But... It can be turned off to prevent the battery from discharging into the solar cells in low light and not need the usual series diode.

[@ricardolandsman9465]

Hi clive just a question i made a solar light with a cellphone battery. It works but very dim. When i conect the light straight to the battery though its very bright. What have i done wrong? Please let me know

[@jimbothompkins]

Wow I love your videos. This is stuff I've always been interested in and I'm learning a lot. Thanks!

[@davefuss]

Tried ordering one from there site, but minimum shipping cost to Australia is over $58.00 or $191.00 if I wanted FedEx. I will contact them to see if there is another way ....

[@Julian_Bester]

I like the lazy Moffett drawing! It looks clean

[@melanieturner9461]

Like when they changed the zigzag resistor to a block, same information, more efficient to use/draw

[@FindLiberty]

Interesting. Very nice/clear explanation. Don't bother to circle that lazy S2 unless nearby circuit component crowding makes the distinction helpful.

[@shanejohns7901]

I think the question is what does any solar device do with the power when the battery it is using is full? Most have some way to discharge to make more room for the incoming energy. But if you shunt it in a way that creates heat, then you need a way to dissipate that heat. Turning it into light is probably the easiest way to go. And considering how little it appears to stay on, it's probably very never going to get to be too much for that LED in there. I'm trying to imagine this thing in like Africa on the hottest day at noontime, shunting that extra power into light... It'd probably get pretty hot on that circuit board, but not too hot to destroy the LED or harm the solder joint.

[@bigclivedotcom]

You can leave the solar panel open circuit once the battery is charged.

[@Red_Scorpian]

have you thought of have a face came down at the bottom corner while your looking at the products?

[@dotpointer]

I have a suggestion for a modification project not related to this video that I have not found that anyone else has tried yet: convert a digital indoor/outdoor thermometer to a very basic stock price ticker - or display of another value - by replacing one or both sensors with variable resistance controlled by a computer interface, like the GPIO on a Raspberry Pi or maybe something else more simple and even cheaper. These thermometers are easy to come by since they often are scrapped when the outdoor sensor cable has been torn off. The simpler and cheaper solution the better. Maybe this has already been tried before?

[@SimonLanghof]

You call that MOSFET style lazy? The lazy style is just a rectangular kink in the Drain-Source line and a Gate bar, add a circle on the Gate connection for P-channel MOSFETs 😄

[@NeatNit]

I genuinely could not find a difference between the two drawings, care to tell me what he was talking about?

[@panklovatina3329]

@@NeatNit ​ The left one is with circle. Right (lazy) one is without circle.

[@NeatNit]

@@panklovatina3329 do you mean the circle around the whole thing?

[@panklovatina3329]

@@NeatNit Yes, I do. There is also difference in gate "wire" possition - top for the left one, middle for the right one, but this does not change difficulty, so there is no connection with laziness :)

[@NeatNit]

@@panklovatina3329 Thanks. I didn't know the circle was part of the symbol.

[@Mr.Unacceptable]

I have a set of solar light I've been playing with. Trying to figure them out. They have 4 transistors and most I've played with are much simpler. Joule thief types. This helped a lot. Now I understand better why there is 4 transistors. But it also has a photo resistor. I'm trying to figure our where I'd add a motion sensor. To build a few new ones on perf board for a much better light and panels.

[@reinux]

This was a surprisingly fascinating one.

[@sausage5849]

Hi BCDC, I need to get some 12v dc sbc lamps for my solar powered caravan, what wattage and colour would you recommend?

[@xsauce3858]

I have been waiting for a solar light video for ages. Finally

[@MothKeeper]

Can we have a 2kw safety leakage test done please in case I'm holding one when I'm struck by lightning while wearing rubber gloves and welly's, stood on 7 inch rubber mats.

[@Electronics-Rocks]

Wow nice to see some thought has gone into designing this product.

[@fuzzybobbles]

They thought about condensation by making the small hole in the bottom of the bulb part. However a bit they missed would have been a waterproof conformal coating on the circuit board. Only costs a small amount extra, but in a country where day and night temperatures can change by 50c or more it may be a good investment to extend the life of the product. Also making the battery easily replaceable might be worthwhile

[@ikbendusan]

the 4 is included with the PO when talkin about phosphate [PO ₄]³⁻

[@MB-st7be]

Your MOSFET is the IEC version, which is much better than the American one (which you call the proper one!)

[@Skyliner_369]

I'm thinking that the transistor to the solar cell remains closed unless the power coming in from the cell overpowers the ability of the LED to draw power. in short, too much sun and the led can't keep up, the panel is cut out. I'm thinking that yeah the light is a sort of charge regulation. but also a charge indicator of sorts. that case looks like it has a lot of room. I wonder how it might handle a much larger cell or collection of cells... in parallel, mind.

[@flickr4jazz]

They are $46 dollars in the US. Expensive

[@chrislee7817]

Hi Clive, can I send you a picture of a component I found in my sander? Think it's gone bang.

[@digitalmediafan]

Why has it taken you so long to review this ?!

[@springwoodcottage4248]

Another product like Rolls Royce once made, everything optimised to get every joule possible & every second of longevity, such a joy to see & to be taken on a tour of the clever circuitry. Thank you for sharing, deeply enjoyed compared to our throw away world of mediocrity.

[@MetalheadAndNerd]

Why would Rolls Royce optimize anything? They can just throw money on problems.

[@springwoodcottage4248]

In the early days of the uk auto industry the cars made by Royce were of such quality that they attracted the wealthy Rolls & in the by and by they formed Rolls Royce. During the second war a relative sold cars to the scrap dealer who broke them apart by lifting them with a crane & then dropping them. This worked great with all cars except Rolls Royce & these proved do difficult to break apart that the scrap dealer refused to take any more Rolls Royce.

[@WackyT08]

Nokero is located here in Denver, CO

[@kennyselvischannel6216]

I literally just asked myself when you're going to upload! XD

[@davidstaniforth681]

Always love your videos, not that I understand much of them at all, despite your superb explanations. I have a question, A power toll li-ion battery will not charge. I have tried the trick of connecting it to a good batterybut no joy. Can I charge it with a car battery charger, trickle or fast charge? I'd rather not blow the house up.

[@bigclivedotcom]

A slightly over discharged cell can cause a battery management PCB to lock out. Sometimes measuring the individual cells indicates one is low, and you can gently trickle charge it to bring it back up to 3V. I've done this to Ryobi power tool batteries in the past.

[@davidstaniforth681]

@@bigclivedotcom Thanks for your speedy answer. I tested each cell, they all show 3.6 or thereabouts. I wonder if this means it's time to buy a new battery for my drill. I don't know a way of charging each cell separately.