When i was about 17 the power banks with built in solar panels were starting to come out. I thought awesome i better get one. I can keep it in my car and it will always be charged and ready if i need to put it in my pocket for use. Lasted a few weeks on the dash of my car in summer and it pretty much cooked the battery to the point it would only get my phone an extra 2 percent. I didnt know then that lithium batteries should not be charged at 160 degrees on my dash. I dont know if the heat only is what ruined it or if it also got overcharged from the solar panel.
I brought 4 off the bay £16 1broken 1 blew the end of the battery off after being on the side of my caravan 2 still working when they want to , good idea badly made as Clive said saved pennies in wrong places
I would imagine that the engineers decided to substitute an internally-protected 18650 for the more usual 3 AAA NiMH batteries, but the accountants decided to procure unprotected 18650s because they were cheaper than the protected ones.
Clive thank you for schematic and details . I have 6 of those. I disconnected the diode and added the 100K resistor you mentioned and a 18650 charge / discharge control board . Now the light works fine .Tx again.
i wonder if these 18650 lithium batteries have gotten so ubiquitous and common that maybe it was cheaper then using a NiMH battery and they just switched it out.
it is. i bought this recently because I was hoping to extract a NiMH AA battery inside (cheaper than buying a NIMH battery), since old solar garden lights have a single AA Nimh. To my surprise, it is a lithium battery but at the form factor of a double A!
I've mentioned this before, but a number of neighbours here have similar lights, and it seems none of them work any more, gotta love outdoor solar lights that can't handle being in the sun & rain... :P
Would be nice to see some projects built around reusing these failed lamps. I have four failed ones from a friend, which failed after one summer, and kept them around in case I can repurpose some of the parts. All had their Li-Ion batteries fried. My theory was also that there didn't seem to be any over charging protection. But I liked the fact that thy shutdown instantly after they sense dark. All my PIR modules with light sensors shutdown only if it gets dark AND there is no movement, which is annoying if they are used as a nightlight inside the house, and they remain on after turning on the lights.
Yeah all the batteries would instantly discharge, i took one from 2 solar lights that didnt work, one battery was at 2 volts, (deep discharge) and one was at 3 volts (possible rechargeable) and they all do not work and wont recharge anymore
I bought one of these from the local Chinese shop. I used your design to make it run as a day/night switch. (I added my own protection circuit and a stripped down cellphone battery)
I’ve got one of these. I believe there is no overcharging protection as I have to change the 18650 battery every few months. They even saved money by direct soldering the battery into the unit instead of using a spring release clip. Cheap shortcuts leading to inbuilt redundancy.
Pretty much every PIR driven thing I have has gone weird or stopped working after a period of time - except the PIRs on the burglar alarm (seemingly more robust for the application). Never really figured out why - I replaced one on a night-light/emergency torch and it worked again for a bit before packing up. This light is similar to a Li-ion one I also have which has become progressively dimmer - probably the same reason due to no over-charge protection...I have had varying levels of performance from others. The other thing that dies is the lens on the front which usually disintegrates from being exposed to the sun :(
That button design is probably a great feature if you have anklebiters around in the daytime. They are very fond of pressing buttons, and then when you come to use your light at night you find it has been left on and the battery is flat.
I had 4 of these also. 3 exploded in the summer after overcharging. Got mine from Banggood and they weren't interested. I now have 4 small solar panels for a project at somepoint...
Not with a Li-Ion cell. Lithium Polymer batteries are the ones that are violently angry when overcharged; a Li-Ion cell will generally just vent some schmoo no more dangerously than a NiCD/NiMH cell would do. Doubly so for a shitty low-capacity low-output cell like would be used in this device.
No. There has never been an example of a Lithium-Ion or Lithium-Polymer explosion. They'll catch on fire, yes, but detonation requires significant containment (which phones and laptops don't have) Basically, the cell expands, and will shatter the casing. If it happens fast enough, it'll ignite. The Lithium batteries that _will_ explode, and are dangerous as all get out, are the Lithium Primary cells, such as the CR123. You can find information about them on the candlepowerforums.com site. Those will detonate, and they release hydrofluoric acid vapor. (depending on the chemistry, of course)
Yes it is, IF you're talking about a multicell installation, as in an EV. Plenty of YT vids showing it. But this one is the clearest, I would suggest. th-cam.com/video/WdDi1haA71Q/w-d-xo.html One catches fire, and the rest explode in a chain reaction. There's a follow up vid from the guy running the shop, where he admits it was a stupid mistake to have attached a 'dumb' charger to the battery pack. In essence, the same as this solar light.
I had one of these go pop a year or two ago now. If i remember correctly the negative side of the battery popped with enough force to punch a hole through the outer casing and knocking it of the fence it was screwed too.
Mine died with lack of proper sealing on the solar cells, rain got to the back of the cell and fairly quickly rotted the copper strips leading to the solder points.
Sorry that Neville and others have had bad luck with these. I have at least 5 like this in my backyard, and a couple more that are “flat” rather than wedge-shaped so as to shine directly down rather into neighbors’ windows. Most are south-facing. All are still working fine after more than 4 years. The solar panels and/or the LED arrays shows signs of degradation and/or water intrusion, but still work well. If I replace any of them or put up others, I will use Clive’s trick of sealing the seams with nail hardener / gel , or maybe hot glue.
10:11 i think one of our garden solar lamps has a circuit that dumps solar panel current to LEDs when the battery is full. in short, the LEDs light up under the sun. I think I've seen it only once when I placed a (manually) charged battery to the lamp.
The nature of the circuitry means that happens if the battery voltage floats too high and the solar panel voltage can exceed the forward voltage of the LED. It also happens if the battery is removed.
also you can put a depending how sunny it is in your area one or two diodes in series 0.6 diode dropper in series with the solar cell to reduce over charging risk and also put 2 k to 10 k load resistor across battery depending how sunny it is
Would be cool to build an array of these PIR sensors and make a sort of primitive IR detecting camera. Maybe use the ones that have a waveform output and write some code to raster an image based on the array signals.
Well, flaws aside still looks like a decent little light The ones I use are always on Low light then go full bright with motion, I have 2 on the back of the house, 1 on a swingset pointing towards the house, 2 on a pole on both sides and 1 on the awning near the front door. They do a hella great job. Due to how they work with the auto brightness mode they last all night, and if by chance they don't I can charge the batteries with a charger. Strangely enough though they don't use 18650's but a normal AA size rechargeable. They were extremely popular this summer, Walmart kept selling out of them in the stores. Couldn't you have used the FLIR to see what component was heating up to see what wasn't working with the board?
you sure it's a regular NiMH AA cell? there are 14500 lithium batteries which are the size of AA cells. dont ever want to accidentally put it into a circuit meant for regular AA :)
The Pir sensor also reacts to reflections of infrared in all surfaces. The reflection properties for infrared light of surfaces is sometimes not the same as for visible light. So your meter or the roll of tape can be very reflective for this band of light
The PIR input on the MCU may have been damaged by an ESD event when you were tinkering. But otherwise I agree that this design is quite bad - hard to acheive when it is also so simple.
The bulbhead atomic beam sunblast sold in the US has the problem of overcharging or running the battery down too far, I don't know which. I fixed mine by adding a protection pcb from a phone battery.
Thanks Clive . My mom has many garden lights and many failures seem to occur, most I've been able to fix but one triggers once and then thats your lot . It is a nice ornament in it own right , its like a fairy castle type thing, she has two one which works properly and the other that as i said just triggers once. I don't really want to take it apart because they are attractive in their own right and to get to the electronics would mean breaking them pretty much. Thanks for the video interesting .
AS312 fully integrated PIR sensor has a working condition supply voltage input range of 2.7-3.3v with an absolute max of 3.6v. Sensitivity and timing is non adjustable. The 18650 cell *may* have a protection circuit inside though, if so the only issue here is subjecting the MCU to overvoltage.
I have one of those - and it just exploded and the lithium battery was ejected and spewed its guts. The insides were like coiled foil paper and burnt to a cinder. The side of the lamp was 3m away and the battery around a meter away. Thank goodness it was no where near people and was outdoors.
Hi Clive looking at your video on this light my daughter has 13 of these lights and she has had two fail and now I know why due to your video thanks for that information so to save her lights could I purchase the protection circuits and fit them plus the resistor you suggest in your video thanks Bob in the UK
Thanks! I bought 2 of these for parts as an unit was 3 USD . The light output is weak but not bad when powered by (5 - 1.2) VDC; USB dropped by two diodes. Sadly the solar panel and LED array are strongly affixed so it isn't likely I can pull them off - just use with the existing case.
Put a simply shunt regulator, like tl431 with pnp transistor, and mosfet switch (for disconnect solar panel with shunt regulator from battery if no charge) to limit a charge voltage to 4.2V. Or just use a DW01 protection board (or battery with protection).
Well, Clive, there's an idea for a potentially great video, featuring possible flameage and the explosion containment pie dish: What happens if you constantly put 18V to a normally discharged 18650 LiIon-cell (sitting at around 2.8V) and current limit it to 1A? Might be fun! Be safe!
If it's a decent cell (any of the namebrands) likely nothing. It would get really hot, but it wouldn't explode or bulge. These things have vents, so if they are overcharged, stuff vents out. if you look at the datasheet, most are actually rated to take up to 12V of absude without exploding.
The sensor triggers for approx 15 - 30 seconds not 7 then goes off if no movement is detected, maybe that's what that built in circuit on the sensor does i.e. a built in timer.. Just a thought ;-) The booklet that came with my light which is very similar to this one minus the small side panel lights says.. Battery : Li-ion, 1300mAh, 3.7 V DC My light started to play up just as you've described after about a couple of months use, so I thought maybe it's to do with the capacity of the battery holding the charge. I replaced the battery with a TR 18650 3.7 V DC Li-ion 9800mAh and so far so good. Just had a thought after looking at your diagram again where you mention its a square wave from the PIR to the MPU, as the square wave signal is reduced by the batteries capacity being able to hold a decent voltage then it would enter into the "indiscriminate zone" for the signal which could account for the weird operating of the unit.
I bought 6 similar to this in 2019, one was installed on a west facing wall. In Australia, Queensland in Jan 2020 we had many days where the temp was over 40 degC and it went off like a bomb! blew the end of the battery off and tore a great chunk out of the side of the light... pity I can't attach a photo. am very careful where I place them now!
I have one that cooked the battery and triggered the safety on the battery, I wonder if fitting a Tp4056 to protect the battery and the 100K resistor to the MPU would make it safe and useful?
They might be relying on the fact that a protected cell is to be used to manage its own charge. I wonder if the factory cell is though. I have the same unit, will have to peek inside!
I have two of these types of light, but they use Nimh cells & come on at dark at a low level, thereby draining the battery even if there are no triggers. They went through last winter without trouble & are fine now.
I bought 4 of these and was similarly concerned over the charging circuit. In mine they did away with the battery clip and just soldered to it and hot glued it in. I was after a low powered PIR alarm trigger that was solar charged, so it may work ok for that with modification by adding a 1sBMS and resistor+zener clamp and sealing around the panel.
My dad got 2 off eBay one has just gone bang. I opened the other one and the battery voltage was 4.9v. I fixed them with a new battery (an old battery from a laptop) with added protection board bms.
Having had these outside my front door for 3+ years they have worked without issue, and they get full sunlight throughout the summer. These are the only units I have purchased that work in the depths of winter too. Best "cheap" PIR solar lights I have brought with a decent amount of light. Still I agree the cheapness of the circuit design is terrible! but if it aint broke dont fix it!
Pull out the lithium batteries and fit 3 series-wired NiMH cells like Clive mentioned. At least that takes away the safety concern. If one current-limited the solar cell, some NiMH cells will take an endless trickle charge in stride. ("Trickle" charge being somewhere in the range of 1/10th C to 1/40th C. Example: A NiMH cell with a 1000 mA capacity would be charged no more than 100 mA current, ideally closer to 25 mA.)
I actually orderd this two months ago. It came as a set of four, two worked two did not. The set i received, the battery is held loosely with a spring and easily slides out(on its own). These batteries are high capacity. I thought it was best to have a battery that was low capacity on solar chargers since the amount of light use is minimal . And is ideal when skies are cloudy, because it doesnt need hours of direct sun exposure to get charged. I opened it and did not like the set up. It did not look safe and feared something catastrophic could happen and cause a fire. Didnt bother to send back because postage would cost as much as i paid for it.
I also have two of these somewhat similar with eight pin control chip and 18650 battery and it stopped working after about ten months. It only blinks ones after covering solar panel.
i was getting 6.7vdc from the solar panel. mine also had hardening/yellowing on the clear covering , as did the PIR cover. mine was given to me by a friend and had the 18650 cell that was reading about 1.3vdc . i liked the brightness of the LED array , so i removed the PIR and the solar panel wires from the PCB & modified the casing and added a 9vdc battery with soft terminal connector . they worked as expected .i hung the lighting the hallway to the bathroom & would turn the switch on at bedtime and off in the morning
I have some very similar to these and they work excellent, they always recharge in the southern facing sun, and they hold a charge well. I take them camping in Eldorado national Forest.
Got the same lamp installed on a wooden fence in a really sunny place for 4 months summer and it hasn't exploded or burned the entire fence yet, I feel disappointed. Need to check if it has a protected battery.
Clive, you said that the cell was surprisingly light or some similar phrase. That is an indication that it is not an 18650 with the stated capacity. There is a video somewhere of someone determining whether 18650s are genuine or not by weighing them on a standard digital kitchen scale and then measuring their capacities. All the light ones (usually obtained from e-bay) where only about 2/3rd of their stated capacity but the heavy ones where around their stated capacity. If you have bought an 18650 and it is light, then you have probably been ripped off. Sorry, Can't remember the weight that determines it is genuine, but I'm sure you can find the video on TH-cam if you search for it.
I always thought that it's kind of crazy to have a lithium battery in those. Isn't heat supposed to be bad for lithium ion batteries? What do you think Clive?
@@bigclivedotcom thanks Clive. I have a couple cheap ones but never put them outside for a full day to charge. I always charged the batteries up & use them for emergency lights.
I wonder if there is a protection built into the cell. Otherwise, that thing might explode... Hot black plastic on a clear day charging at 5-6V... Yikes. It still might explode...
Clive, unfortunately I found this video only after a similar issue with these solar lights. The brand I got have 30 leds. They worked faultlessly for 2 months and then one by one they all suffered the same issue; flashing at around 1 hertz until they died completely. Ebay were no help as they said it had been after the months 'protection' and the seller didn't want to know. I took one apart and suspected that the 30leds were draining the battery too much and the solar panel couldn't charge them despite living in Australia where we get blistering sunshine. I replaced the non marked 18560s with some LG hG2 cells thinking the extra capacity might solve the problem but sadly not as all 20 of them are starting to fail. I looked at the circuit board and like you was amazed that there is no charge controller to prevent over charging and more importantly over discharging of the cells. I then thought about adding one of the cheap 1s BMS boards but the solar cell doesn't seem to give enough oomph for the BMS to start working. There are other versions of these out there that do have Liion charging circuitry but it is a bit of a hit and miss as to which ones you get. By hook or by crook I will get the original units working but now have 20 cells to try and revive from around 1volt!
@@bigclivedotcom I have had a bit more time on this today on the solar cell kicks out 6.75v in full sun and, after connecting the bms properly (it has common +ve not -ve as the documentation says!) I can get some output from the BMS but I can't measure any current going into the cell. I am going to discharge one later and replace the one in the device and see if it charges. I am still annoyed that I can't recover the 20 dead LG cells, I tried the putting the dead one in parallel with a fully charged cell for 30 seconds but all I can get is 1.38v from 0.6v. In the words of Sherlock, "The game is afoot, Watson!" I had some of the button 18650 protection circuits (the ones that have the connection going along the side of the cell) but they didn't work as they have common terminals that doesn't isolate the connection. I have got some usb-c liion charger modules (www.aliexpress.com/item/1005004654476348.html?spm=a2g0o.order_list.order_list_main.34.338c1802FVVZht) intended for power banks (1.50 Aud) on order from 'Mau's Empire of Dirt'; it has 2 input terminals, terminals for the Liion cell and 2 more as output terminals. I think I will connect the solar panel directly to the input, and the output terminals to where the solar cell was and finally take out the diode and put in a wire. This should allow the 18650 to be charged properly and the output to power the rest of the circuit. I will let you know how it goes.The frankensteining seems to have been a success; I removed the diode and piggy backed the usb-c charging board to the solar panel, connected the liion cell to the 'battery' terminal on the usb-c module and then the 'output' terminals to the battery+ and battery- on the original pcb. They have worked flawlessly for 2 weeks now and the liion cells are charged to 4.2v! Win Win.
I'd love to see some real-world voltage data from one of these garden lights. Open circuit voltage will differ from load voltage, and things like shading (whether from dirt or pollen covering the entire array or from a tree branch shading just one of the cells) will change the output voltage quite significantly.
clive. could you do a series of videos showing how to set up a basic diy solar power set up on a budget? Starting from just charging a laptop or phone charger and allowing a gradual expansion to say lights etc. aim it at people like me who want to be greener but dont have the funds, but want to get the ball rolling and build the system up when finances allow.? Couild we modify bargin store stuff to mae a system? can you gear it to UK weather conditions? thanks
I just received one of these from bhang goot and immediately noticed the same flaw. The battery in mine checked out to just under 1200 mAh but has no charging protection. For six bucks, it was worth taking apart and I suppose when it fails, I can try to put in a protected battery or change over to a few NiMH cells. Too bad they skimped a little too much.
@@bigclivedotcom I have been looking at one on eBay that kind looks like a UFO with a hook on top to hang it, interestingly it also has USB ports to charge a phone! 3 different types of USB ports! Solar panels on top but it doesn't look like they might have the capacity to charge the internal lithium battery.. possibly batteries if it was intended to actually charge a phone! Might be a worthy contender for a peek inside...
This is not the first time I have seen a device with an 18650 battery and no charge control. The other was a speaker which had a diode and 2r2 resistor from 5v usb. This is a worrying trend.
it is possible that. 1. the battery is lifepo4 and not marked as such. 2. at the few milliamps the cell can self clamp due to leakage current. 3. an internal shunt resistor that effectively shorts or loads the solar panel causing it to not be able to charge anymore. 4. the low capacity of 600mah combined with the low c rating and the low current of the solar panel even if the cell could be destroyed the internal short that usually fires up most lithium cells would just discharge it. when most battery fires happen it is because the charger is dumping several amps into the cell.
Watched this last night, realised I had the same one in my conservatory as a motion sensor when it was dark. Checked the voltage and the cell was at 5.2v (still working) thank you for flagging this up before it set my conservatory on fire.
Panic over Clive, I contacted them and this is genuinely the reply I got ……New message from: trybest6699 Top Rated Seller(13,481YellowShooting Star) 5.3V has no harm to the human body. There is no danger below 36V. do not worry.
I have two similar lights without the side LEDs. They worked OK for a few days, but after that they are extremely problematic. After a day of full light they may work for a few times (2 or 3 or so) at night when the PIR detects movement, then die. Had a look at the PCB and though similar, this one has an 8 pin controller chip.
i think a lot of this stuff is made to crap out after a short time hoping you'll go buy more. I bought one of those cheap LED xmas light strings that used 3 AAs, it had the fade blink and constant on. So and i added a 18650 battery instead, it worked great for 6-7 months only ran them down to 3V. then it started totally discharging my 18650s, so much they were ruined (below 1v)and they wont even charge. Then it started acting up wouldnt even turn on unless i fiddled with the switch for 5 mins. After taking it apart I noticed the chip inside was getting so hot you couldnt touch it! I added a heatsink but that didnt help. And it appeared one of the SM resistors had exploded off the board, but they're so tiny its hard to tell if it did actually have one on there. Im glad i only spent $1 on it otherwise i might have taken it back and complained. lol
Bought 4 (without the side lighting) a while back to stick on the shed. Better take one apart to see if it's similar. If so I'll try replacing the battery with a protected phone battery.
George, did you have any success with protected 18650's, in my attempts so far the solar cell won't power the 1S BMS (those round ones usually put on top of a 'naked' cell.
We where given a set as a Crimbo present which have been installed and I now have nightmares of the tree catching fire and setting the hedge alight also doesn't bare thinking about.
Could you stick in a zenner for protection and a couple of diodes from the solar panel to the micro to reduce that voltage (or even another zenner to clamp that down too)?
Hey big Clive, over at the eevblog forum there is a guide on how to hack your particular clamp multimeter to have the screen go to 9999 counts and other interesting things about it.
@@paulstafIt's considered uncouth to directly link to other fora/ videos. If you search uni-t ut210e hacks you'll find quite a few long threads about the matter. Take care though, you can easily make it useless.
I’ve had 3 of these all work for a month or two and stop. They are shite. I wonder if I did add a protection circuit and s current limit resistor if it would work. Also thought of a 10K pull-up resistor on the MPU to give the voltage needed for trigger?
One thing they could have done for battery protection is to put in a dump load - the light controller could potentially switch on the load if the battery nears full charge.
Just turning the LED's on would do that without too much extra cost. Ok, so the light on in the daytime is not ideal but it would make it live longer - but that isn't what they want is it?
What is the point of 3.3V LDO? If light present, the voltage on the microcontroller rises to the battery voltage (solar panel diode +0.6V under battery voltage and MCU clamp diode -0.6V). If battery is unplugged, the microcontroller (and probably PIR) will be damaged by high voltage from solar panel (about 6 volts for unloaded solar panel).
My work sells these. I was very surprised when I opened it up to find not very much at all. Ironic how it says not to use it in full sunlight when it's being sold in Australia, the country with one of the highest levels of solar radiation in the world.
That epoxy coated solar array isn't going to last a single summer Down Under. Almost worth siliconing some glass over the cell (but even then the cell's not going to last more than TWO summers!! :-)
i have 3 of these and one sit outside window and 2 inside and dont have leds on the sides but still working nearly a year , but a thing is the 2 inside i kept off and the one outside on... stil works (i didnt check any charge current) but just seel them with clear silicone around ...stil works... do you have any improvement to the circuits?
The Aldi one looked a lot better. I brought it across with me in the belief that the chip was doing some clever sensing from a standard PIR sensor, but this unit made me realise the PIR is probably digital.
I've got a few of the versions where it only comes on temporarily when it detects movement, they've been there a couple years now and have been working fine. No idea if the battery is charge regulated.
The maximum voltage to the MCU's input pin is 4.2V PLUS the diode drop, not minus the diode drop. About 4.4V or higher. If you use these in somewhere like Arizona, you will soon have a lawn with a series of scorch marks where the lights used to be.
Gracias.desde colombia quisiera que me ayudaras como dejo que la luz quede permanente o que el tiempo sea muho mayor al que tiene. Ocomo anulo el sensor de movimiento un diagrama por fa.esplicas muy bien
Something I've discovered is that if you use a protected cell, or add a protection circuit to an 18650, you can directly charge the cell from a 6V solar panel. The protection circuit seems to do a good job of preventing the input voltage from pushing the cell voltage too high. So if that's acually a cell with an integrated protection circuit, it's probably okay. If it's an unprotected cell, they may be relying on the idea that the solar cells can't fully charge the lithium cell over a full day's sunshine with their limited current-supply capacity, which as you pointed out is a bad idea when the light isn't left on continuously overnight. Summary: You're right on all counts. It's a bad circuit design.
Solar light that requires "do not use under direct sunlight" warning
When i was about 17 the power banks with built in solar panels were starting to come out. I thought awesome i better get one. I can keep it in my car and it will always be charged and ready if i need to put it in my pocket for use. Lasted a few weeks on the dash of my car in summer and it pretty much cooked the battery to the point it would only get my phone an extra 2 percent. I didnt know then that lithium batteries should not be charged at 160 degrees on my dash. I dont know if the heat only is what ruined it or if it also got overcharged from the solar panel.
@@poellot you're lucky it didn't immolate your car 🤨
@@CsykKrit and i was top of my class in college. I can see why this world needs so many warnings. A smart kid like me even needed them.
I brought 4 off the bay £16 1broken 1 blew the end of the battery off after being on the side of my caravan 2 still working when they want to , good idea badly made as Clive said saved pennies in wrong places
@@poellot jesus
I would imagine that the engineers decided to substitute an internally-protected 18650 for the more usual 3 AAA NiMH batteries, but the accountants decided to procure unprotected 18650s because they were cheaper than the protected ones.
I was thinking the same thing some cells have a BMS board on the end of the cell.
so If I put 3 aaa nimh in then it will work well? I have a dozen of them.
Clive thank you for schematic and details . I have 6 of those. I disconnected the diode and added the 100K resistor you mentioned and a 18650 charge / discharge control board . Now the light works fine .Tx again.
i wonder if these 18650 lithium batteries have gotten so ubiquitous and common that maybe it was cheaper then using a NiMH battery and they just switched it out.
this can very well be the case, clive made a video about lithium batterys being used in one time phone chargers and e-cigs
it is. i bought this recently because I was hoping to extract a NiMH AA battery inside (cheaper than buying a NIMH battery), since old solar garden lights have a single AA Nimh. To my surprise, it is a lithium battery but at the form factor of a double A!
It's very well designed to be a cheap product that breaks really quickly.
I've mentioned this before, but a number of neighbours here have similar lights, and it seems none of them work any more, gotta love outdoor solar lights that can't handle being in the sun & rain... :P
Would be nice to see some projects built around reusing these failed lamps.
I have four failed ones from a friend, which failed after one summer, and kept them around in case I can repurpose some of the parts.
All had their Li-Ion batteries fried.
My theory was also that there didn't seem to be any over charging protection.
But I liked the fact that thy shutdown instantly after they sense dark.
All my PIR modules with light sensors shutdown only if it gets dark AND there is no movement, which is annoying if they are used as a nightlight inside the house, and they remain on after turning on the lights.
This. This is exactly what im trying to bypass
Yeah all the batteries would instantly discharge, i took one from 2 solar lights that didnt work, one battery was at 2 volts, (deep discharge) and one was at 3 volts (possible rechargeable) and they all do not work and wont recharge anymore
I bought one of these from the local Chinese shop. I used your design to make it run as a day/night switch. (I added my own protection circuit and a stripped down cellphone battery)
I’ve got one of these. I believe there is no overcharging protection as I have to change the 18650 battery every few months. They even saved money by direct soldering the battery into the unit instead of using a spring release clip. Cheap shortcuts leading to inbuilt redundancy.
Pretty much every PIR driven thing I have has gone weird or stopped working after a period of time - except the PIRs on the burglar alarm (seemingly more robust for the application). Never really figured out why - I replaced one on a night-light/emergency torch and it worked again for a bit before packing up. This light is similar to a Li-ion one I also have which has become progressively dimmer - probably the same reason due to no over-charge protection...I have had varying levels of performance from others. The other thing that dies is the lens on the front which usually disintegrates from being exposed to the sun :(
The lower case l in "light" on the branding sticker was the first red flag before it was even cracked open!
That button design is probably a great feature if you have anklebiters around in the daytime. They are very fond of pressing buttons, and then when you come to use your light at night you find it has been left on and the battery is flat.
Perfect timing Clive. Only last night I was wondering why my relatively new light was very dim despite excellent sunshine all day. "Frying tonight?"
Have previously seen a few pictures of those types of lights, same exact design with blown cells that violently erupt from the casing.
I had 4 of these also. 3 exploded in the summer after overcharging. Got mine from Banggood and they weren't interested. I now have 4 small solar panels for a project at somepoint...
Isn't "no lithium charge control" a long phrase for "bomb"?
Not with a Li-Ion cell. Lithium Polymer batteries are the ones that are violently angry when overcharged; a Li-Ion cell will generally just vent some schmoo no more dangerously than a NiCD/NiMH cell would do. Doubly so for a shitty low-capacity low-output cell like would be used in this device.
No. There has never been an example of a Lithium-Ion or Lithium-Polymer explosion. They'll catch on fire, yes, but detonation requires significant containment (which phones and laptops don't have) Basically, the cell expands, and will shatter the casing. If it happens fast enough, it'll ignite. The Lithium batteries that _will_ explode, and are dangerous as all get out, are the Lithium Primary cells, such as the CR123. You can find information about them on the candlepowerforums.com site. Those will detonate, and they release hydrofluoric acid vapor. (depending on the chemistry, of course)
The current from the solar panel is too low for the battery to be a safety risk - it will degrade though.
@@tbelding christ i was just reading about that and I have a sudden urge to never use them again.
Yes it is, IF you're talking about a multicell installation, as in an EV.
Plenty of YT vids showing it. But this one is the clearest, I would suggest.
th-cam.com/video/WdDi1haA71Q/w-d-xo.html
One catches fire, and the rest explode in a chain reaction.
There's a follow up vid from the guy running the shop, where he admits it was a stupid mistake to have attached a 'dumb' charger to the battery pack. In essence, the same as this solar light.
Shame it's a nice case. worth saving? Maybe a salvage series in the future?
maybe slap a proper charging board in and your gucci.
or just it as a dumb nightlight
I meant a series of videos on repair, recycling, reusing some of the things Clive shows us. This light is a prime candidate.
I was thinking the same thing
I had one of these go pop a year or two ago now. If i remember correctly the negative side of the battery popped with enough force to punch a hole through the outer casing and knocking it of the fence it was screwed too.
Mine died with lack of proper sealing on the solar cells, rain got to the back of the cell and fairly quickly rotted the copper strips leading to the solder points.
Sorry that Neville and others have had bad luck with these. I have at least 5 like this in my backyard, and a couple more that are “flat” rather than wedge-shaped so as to shine directly down rather into neighbors’ windows. Most are south-facing. All are still working fine after more than 4 years. The solar panels and/or the LED arrays shows signs of degradation and/or water intrusion, but still work well.
If I replace any of them or put up others, I will use Clive’s trick of sealing the seams with nail hardener / gel , or maybe hot glue.
10:11 i think one of our garden solar lamps has a circuit that dumps solar panel current to LEDs when the battery is full. in short, the LEDs light up under the sun. I think I've seen it only once when I placed a (manually) charged battery to the lamp.
The nature of the circuitry means that happens if the battery voltage floats too high and the solar panel voltage can exceed the forward voltage of the LED. It also happens if the battery is removed.
@@bigclivedotcom thanks! I can recall it also happened without the battery!
also you can put a depending how sunny it is in your area one or two diodes in series 0.6 diode dropper in series with the solar cell to reduce over charging risk and also put 2 k to 10 k load resistor across battery depending how sunny it is
Would be cool to build an array of these PIR sensors and make a sort of primitive IR detecting camera. Maybe use the ones that have a waveform output and write some code to raster an image based on the array signals.
Well, flaws aside still looks like a decent little light
The ones I use are always on Low light then go full bright with motion, I have 2 on the back of the house, 1 on a swingset pointing towards the house, 2 on a pole on both sides and 1 on the awning near the front door. They do a hella great job. Due to how they work with the auto brightness mode they last all night, and if by chance they don't I can charge the batteries with a charger. Strangely enough though they don't use 18650's but a normal AA size rechargeable. They were extremely popular this summer, Walmart kept selling out of them in the stores.
Couldn't you have used the FLIR to see what component was heating up to see what wasn't working with the board?
you sure it's a regular NiMH AA cell? there are 14500 lithium batteries which are the size of AA cells. dont ever want to accidentally put it into a circuit meant for regular AA :)
@@Polite_Cat Yup, its a normal size AA rechargeable, I do have 2 lights that use 2 bigish 14500 cells that you're talking about though.
I wonder if it originally was meant to use a protected button top flashlight battery.
The Pir sensor also reacts to reflections of infrared in all surfaces. The reflection properties for infrared light of surfaces is sometimes not the same as for visible light. So your meter or the roll of tape can be very reflective for this band of light
The PIR input on the MCU may have been damaged by an ESD event when you were tinkering. But otherwise I agree that this design is quite bad - hard to acheive when it is also so simple.
That did go through my mind, although I do try to keep a low static risk work area.
After all, some PIR sensors have a high impedance output. They are quite sensitive to ESD.
The bulbhead atomic beam sunblast sold in the US has the problem of overcharging or running the battery down too far, I don't know which. I fixed mine by adding a protection pcb from a phone battery.
Thanks Clive . My mom has many garden lights and many failures seem to occur, most I've been able to fix but one triggers once and then thats your lot . It is a nice ornament in it own right , its like a fairy castle type thing, she has two one which works properly and the other that as i said just triggers once. I don't really want to take it apart because they are attractive in their own right and to get to the electronics would mean breaking them pretty much.
Thanks for the video interesting .
Hmm... As mentioned b4, I've got a similar light, it's dead too... among many non-working solar lights. They seem to just die for no apparent reason.
AS312 fully integrated PIR sensor has a working condition supply voltage input range of 2.7-3.3v with an absolute max of 3.6v.
Sensitivity and timing is non adjustable.
The 18650 cell *may* have a protection circuit inside though, if so the only issue here is subjecting the MCU to overvoltage.
Thanks. I just ordered some AM312 modules from eBay.
Battery looks like a Ultrafire special.
I have one of those - and it just exploded and the lithium battery was ejected and spewed its guts. The insides were like coiled foil paper and burnt to a cinder. The side of the lamp was 3m away and the battery around a meter away. Thank goodness it was no where near people and was outdoors.
This is happening a lot. There appears to be no charge control on some of the lithium batteries.
Can you make a simple circuit to put it in the casing so that you could use the parts well ? And... would a protected litihuim-cell work?
Hi Clive looking at your video on this light my daughter has 13 of these lights and she has had two fail and now I know why due to your video thanks for that information so to save her lights could I purchase the protection circuits and fit them plus the resistor you suggest in your video thanks Bob in the UK
The ones of these I have packed up. Now work very nicely now that I have modified them by adding battery charging/protection and new cells.
I have one of these in the bathroom currently the solar bit is in the window its perfect at night, no need to turn the lights on.
do you experience any blackout moments in the middle of your toilet activities because of its 30-seconds power-off feature?
@@paulhayhurst lol 😂
Thanks! I bought 2 of these for parts as an unit was 3 USD . The light output is weak but not bad when powered by (5 - 1.2) VDC; USB dropped by two diodes. Sadly the solar panel and LED array are strongly affixed so it isn't likely I can pull them off - just use with the existing case.
Put a simply shunt regulator, like tl431 with pnp transistor, and mosfet switch (for disconnect solar panel with shunt regulator from battery if no charge) to limit a charge voltage to 4.2V. Or just use a DW01 protection board (or battery with protection).
6:08 Oh lovely, it will charge the Lithium cell to 6V under daylight/sunshine! Fantastic!
This is going to make the battery blow up!
Thank you Big Clive! I appreciate what you do, you are one of the few that I actually learn from, your AWESOME!
Thanks again.
My name is Guy Garrett.
Well, Clive, there's an idea for a potentially great video, featuring possible flameage and the explosion containment pie dish: What happens if you constantly put 18V to a normally discharged 18650 LiIon-cell (sitting at around 2.8V) and current limit it to 1A? Might be fun! Be safe!
It will fly out of the pie dish like a rocket and burn down his workshop.
@@BenjaminEsposti We wouldn't want that though.
If it's a decent cell (any of the namebrands) likely nothing. It would get really hot, but it wouldn't explode or bulge. These things have vents, so if they are overcharged, stuff vents out.
if you look at the datasheet, most are actually rated to take up to 12V of absude without exploding.
The sensor triggers for approx 15 - 30 seconds not 7 then goes off if no movement is detected, maybe that's what that built in circuit on the sensor does i.e. a built in timer.. Just a thought ;-)
The booklet that came with my light which is very similar to this one minus the small side panel lights says.. Battery : Li-ion, 1300mAh, 3.7 V DC
My light started to play up just as you've described after about a couple of months use, so I thought maybe it's to do with the capacity of the battery holding the charge.
I replaced the battery with a TR 18650 3.7 V DC Li-ion 9800mAh and so far so good.
Just had a thought after looking at your diagram again where you mention its a square wave from the PIR to the MPU, as the square wave signal is reduced by the batteries capacity being able to hold a decent voltage then it would enter into the "indiscriminate zone" for the signal which could account for the weird operating of the unit.
Can you please teardown the PIR, we might be seeing a whole new all in one PIR sensor.
I bought 6 similar to this in 2019, one was installed on a west facing wall. In Australia, Queensland in Jan 2020 we had many days where the temp was over 40 degC and it went off like a bomb! blew the end of the battery off and tore a great chunk out of the side of the light... pity I can't attach a photo. am very careful where I place them now!
I have one that cooked the battery and triggered the safety on the battery, I wonder if fitting a Tp4056 to protect the battery and the 100K resistor to the MPU would make it safe and useful?
They might be relying on the fact that a protected cell is to be used to manage its own charge. I wonder if the factory cell is though. I have the same unit, will have to peek inside!
holy shit i was freaking out like how on earth did his fingers get that tiny then i was like oh thats a blown up printout.....
I have two of these types of light, but they use Nimh cells & come on at dark at a low level, thereby draining the battery even if there are no triggers. They went through last winter without trouble & are fine now.
I bought 4 of these and was similarly concerned over the charging circuit. In mine they did away with the battery clip and just soldered to it and hot glued it in. I was after a low powered PIR alarm trigger that was solar charged, so it may work ok for that with modification by adding a 1sBMS and resistor+zener clamp and sealing around the panel.
My dad got 2 off eBay one has just gone bang. I opened the other one and the battery voltage was 4.9v. I fixed them with a new battery (an old battery from a laptop) with added protection board bms.
Having had these outside my front door for 3+ years they have worked without issue, and they get full sunlight throughout the summer. These are the only units I have purchased that work in the depths of winter too. Best "cheap" PIR solar lights I have brought with a decent amount of light. Still I agree the cheapness of the circuit design is terrible! but if it aint broke dont fix it!
had some of these delivered last month but not fitted them yet.
Won't hold my breath after this.
Pull out the lithium batteries and fit 3 series-wired NiMH cells like Clive mentioned. At least that takes away the safety concern. If one current-limited the solar cell, some NiMH cells will take an endless trickle charge in stride. ("Trickle" charge being somewhere in the range of 1/10th C to 1/40th C. Example: A NiMH cell with a 1000 mA capacity would be charged no more than 100 mA current, ideally closer to 25 mA.)
You can change the lithium cell for a protected one.
@@McTroyd will do that. Many thanks
I actually orderd this two months ago. It came as a set of four, two worked two did not. The set i received, the battery is held loosely with a spring and easily slides out(on its own). These batteries are high capacity. I thought it was best to have a battery that was low capacity on solar chargers since the amount of light use is minimal . And is ideal when skies are cloudy, because it doesnt need hours of direct sun exposure to get charged. I opened it and did not like the set up. It did not look safe and feared something catastrophic could happen and cause a fire. Didnt bother to send back because postage would cost as much as i paid for it.
I also have two of these somewhat similar with eight pin control chip and 18650 battery and it stopped working after about ten months. It only blinks ones after covering solar panel.
They could have compensated by just having a charging control chip on the battery itself, but I guess that's too expensive for them.
Was thinking along the same lines. Perhaps the circuit was designed with that in mind, and then corners were cut.
i was getting 6.7vdc from the solar panel. mine also had hardening/yellowing on the clear covering , as did the PIR cover. mine was given to me by a friend and had the 18650 cell that was reading about 1.3vdc . i liked the brightness of the LED array , so i removed the PIR and the solar panel wires from the PCB & modified the casing and added a 9vdc battery with soft terminal connector . they worked as expected .i hung the lighting the hallway to the bathroom & would turn the switch on at bedtime and off in the morning
I have some very similar to these and they work excellent, they always recharge in the southern facing sun, and they hold a charge well.
I take them camping in Eldorado national Forest.
Got the same lamp installed on a wooden fence in a really sunny place for 4 months summer and it hasn't exploded or burned the entire fence yet, I feel disappointed.
Need to check if it has a protected battery.
Clive, you said that the cell was surprisingly light or some similar phrase. That is an indication that it is not an 18650 with the stated capacity. There is a video somewhere of someone determining whether 18650s are genuine or not by weighing them on a standard digital kitchen scale and then measuring their capacities. All the light ones (usually obtained from e-bay) where only about 2/3rd of their stated capacity but the heavy ones where around their stated capacity. If you have bought an 18650 and it is light, then you have probably been ripped off. Sorry, Can't remember the weight that determines it is genuine, but I'm sure you can find the video on TH-cam if you search for it.
I always thought that it's kind of crazy to have a lithium battery in those. Isn't heat supposed to be bad for lithium ion batteries? What do you think Clive?
If it's mounted at the base of the unit it shouldn't be too bad as it's not getting heated directly.
@@bigclivedotcom thanks Clive.
I have a couple cheap ones but never put them outside for a full day to charge. I always charged the batteries up & use them for emergency lights.
@@johncoops6897 I don't have any protected cells but thank you for the information!
I wonder if there is a protection built into the cell. Otherwise, that thing might explode... Hot black plastic on a clear day charging at 5-6V... Yikes. It still might explode...
Clive, unfortunately I found this video only after a similar issue with these solar lights. The brand I got have 30 leds. They worked faultlessly for 2 months and then one by one they all suffered the same issue; flashing at around 1 hertz until they died completely. Ebay were no help as they said it had been after the months 'protection' and the seller didn't want to know. I took one apart and suspected that the 30leds were draining the battery too much and the solar panel couldn't charge them despite living in Australia where we get blistering sunshine. I replaced the non marked 18560s with some LG hG2 cells thinking the extra capacity might solve the problem but sadly not as all 20 of them are starting to fail. I looked at the circuit board and like you was amazed that there is no charge controller to prevent over charging and more importantly over discharging of the cells. I then thought about adding one of the cheap 1s BMS boards but the solar cell doesn't seem to give enough oomph for the BMS to start working. There are other versions of these out there that do have Liion charging circuitry but it is a bit of a hit and miss as to which ones you get. By hook or by crook I will get the original units working but now have 20 cells to try and revive from around 1volt!
I wonder if the solar panel is putting out decent current.
@@bigclivedotcom I have had a bit more time on this today on the solar cell kicks out 6.75v in full sun and, after connecting the bms properly (it has common +ve not -ve as the documentation says!) I can get some output from the BMS but I can't measure any current going into the cell. I am going to discharge one later and replace the one in the device and see if it charges. I am still annoyed that I can't recover the 20 dead LG cells, I tried the putting the dead one in parallel with a fully charged cell for 30 seconds but all I can get is 1.38v from 0.6v. In the words of Sherlock, "The game is afoot, Watson!"
I had some of the button 18650 protection circuits (the ones that have the connection going along the side of the cell) but they didn't work as they have common terminals that doesn't isolate the connection.
I have got some usb-c liion charger modules (www.aliexpress.com/item/1005004654476348.html?spm=a2g0o.order_list.order_list_main.34.338c1802FVVZht) intended for power banks (1.50 Aud) on order from 'Mau's Empire of Dirt'; it has 2 input terminals, terminals for the Liion cell and 2 more as output terminals. I think I will connect the solar panel directly to the input, and the output terminals to where the solar cell was and finally take out the diode and put in a wire. This should allow the 18650 to be charged properly and the output to power the rest of the circuit. I will let you know how it goes.The frankensteining seems to have been a success; I removed the diode and piggy backed the usb-c charging board to the solar panel, connected the liion cell to the 'battery' terminal on the usb-c module and then the 'output' terminals to the battery+ and battery- on the original pcb. They have worked flawlessly for 2 weeks now and the liion cells are charged to 4.2v! Win Win.
I'd love to see some real-world voltage data from one of these garden lights. Open circuit voltage will differ from load voltage, and things like shading (whether from dirt or pollen covering the entire array or from a tree branch shading just one of the cells) will change the output voltage quite significantly.
clive. could you do a series of videos showing how to set up a basic diy solar power set up on a budget? Starting from just charging a laptop or phone charger and allowing a gradual expansion to say lights etc. aim it at people like me who want to be greener but dont have the funds, but want to get the ball rolling and build the system up when finances allow.? Couild we modify bargin store stuff to mae a system? can you gear it to UK weather conditions? thanks
I just received one of these from bhang goot and immediately noticed the same flaw. The battery in mine checked out to just under 1200 mAh but has no charging protection. For six bucks, it was worth taking apart and I suppose when it fails, I can try to put in a protected battery or change over to a few NiMH cells. Too bad they skimped a little too much.
Does the 18650 have a protection circuit built into it? Some button top cells have the protection built in. Or is that only for under voltage?
This video is a year old now and these things are all over eBay still.
There's even more of them, and all different inside.
@@bigclivedotcom I have been looking at one on eBay that kind looks like a UFO with a hook on top to hang it, interestingly it also has USB ports to charge a phone! 3 different types of USB ports! Solar panels on top but it doesn't look like they might have the capacity to charge the internal lithium battery.. possibly batteries if it was intended to actually charge a phone!
Might be a worthy contender for a peek inside...
Leave the lens off pyrosensors when doing basic bench testing.
(Says someone who used to design PIRs.)
Too bad, there is a decent product that could have had a long life with a few components and a little care. Thanks Clive.
@jerril42
That's most of the cheap chinese products ordered from china.
This is not the first time I have seen a device with an 18650 battery and no charge control. The other was a speaker which had a diode and 2r2 resistor from 5v usb. This is a worrying trend.
it is possible that.
1. the battery is lifepo4 and not marked as such.
2. at the few milliamps the cell can self clamp due to leakage current.
3. an internal shunt resistor that effectively shorts or loads the solar panel causing it to not be able to charge anymore.
4. the low capacity of 600mah combined with the low c rating and the low current of the solar panel even if the cell could be destroyed the internal short that usually fires up most lithium cells would just discharge it.
when most battery fires happen it is because the charger is dumping several amps into the cell.
Watched this last night, realised I had the same one in my conservatory as a motion sensor when it was dark. Checked the voltage and the cell was at 5.2v (still working) thank you for flagging this up before it set my conservatory on fire.
Panic over Clive, I contacted them and this is genuinely the reply I got ……New message from: trybest6699 Top Rated Seller(13,481YellowShooting Star)
5.3V has no harm to the human body.
There is no danger below 36V.
do not worry.
@@andywalker1388LOL
mine was at 4.9, i saw it was overcharged so thank god
I have two similar lights without the side LEDs. They worked OK for a few days, but after that they are extremely problematic. After a day of full light they may work for a few times (2 or 3 or so) at night when the PIR detects movement, then die.
Had a look at the PCB and though similar, this one has an 8 pin controller chip.
Is it getting a decent amount of direct sunlight every day?
@@bigclivedotcom yes very much, totally free and pointed almost straight south. It’s cold here now though...not sure if that might be the problem.
i think a lot of this stuff is made to crap out after a short time hoping you'll go buy more. I bought one of those cheap LED xmas light strings that used 3 AAs, it had the fade blink and constant on. So and i added a 18650 battery instead, it worked great for 6-7 months only ran them down to 3V. then it started totally discharging my 18650s, so much they were ruined (below 1v)and they wont even charge. Then it started acting up wouldnt even turn on unless i fiddled with the switch for 5 mins. After taking it apart I noticed the chip inside was getting so hot you couldnt touch it! I added a heatsink but that didnt help. And it appeared one of the SM resistors had exploded off the board, but they're so tiny its hard to tell if it did actually have one on there. Im glad i only spent $1 on it otherwise i might have taken it back and complained. lol
perhaps this was a protected 18650 cell... some of them do have build in protection...
Bought 4 (without the side lighting) a while back to stick on the shed. Better take one apart to see if it's similar. If so I'll try replacing the battery with a protected phone battery.
You can also get protected 18650 cells.
George, did you have any success with protected 18650's, in my attempts so far the solar cell won't power the 1S BMS (those round ones usually put on top of a 'naked' cell.
So technically, can I just add a tp4056 board for battery protection?
Yes you can.
We where given a set as a Crimbo present which have been installed and I now have nightmares of the tree catching fire and setting the hedge alight also doesn't bare thinking about.
Could you stick in a zenner for protection and a couple of diodes from the solar panel to the micro to reduce that voltage (or even another zenner to clamp that down too)?
Hey big Clive, over at the eevblog forum there is a guide on how to hack your particular clamp multimeter to have the screen go to 9999 counts and other interesting things about it.
I have toyed with updating the memory chip. I think it was a group of Russian guys who discovered it could be done.
I have one of those, can you post a link? Thanks...
@@paulstafIt's considered uncouth to directly link to other fora/ videos. If you search uni-t ut210e hacks you'll find quite a few long threads about the matter. Take care though, you can easily make it useless.
@@ΚώσταςΜπέκος-ι4ι I doubt it as it brings extra traffic. It's called "flattery" :-)).
So if you were to design such a circuit, (as you have so well on the solar garden light), how would the schematic and parts list differ?
If I buy one of these will tape off some of the panels. would that drop the charge down to a safe level?
I've got this same light and its lasted us well over 6 months. No complaints here.
The same case often has wildly different circuitry in it.
I’ve had 3 of these all work for a month or two and stop. They are shite. I wonder if I did add a protection circuit and s current limit resistor if it would work. Also thought of a 10K pull-up resistor on the MPU to give the voltage needed for trigger?
One thing they could have done for battery protection is to put in a dump load - the light controller could potentially switch on the load if the battery nears full charge.
Just turning the LED's on would do that without too much extra cost. Ok, so the light on in the daytime is not ideal but it would make it live longer - but that isn't what they want is it?
What is the point of 3.3V LDO? If light present, the voltage on the microcontroller rises to the battery voltage (solar panel diode +0.6V under battery voltage and MCU clamp diode -0.6V). If battery is unplugged, the microcontroller (and probably PIR) will be damaged by high voltage from solar panel (about 6 volts for unloaded solar panel).
The regulator was to provide a stable voltage to the PIR module to avoid it self-triggering again as the LEDs were turned off.
Always a plesure when i see a new video from clive
My work sells these. I was very surprised when I opened it up to find not very much at all.
Ironic how it says not to use it in full sunlight when it's being sold in Australia, the country with one of the highest levels of solar radiation in the world.
That epoxy coated solar array isn't going to last a single summer Down Under. Almost worth siliconing some glass over the cell (but even then the cell's not going to last more than TWO summers!! :-)
i have 3 of these and one sit outside window and 2 inside and dont have leds on the sides but still working nearly a year , but a thing is the 2 inside i kept off and the one outside on... stil works (i didnt check any charge current) but just seel them with clear silicone around ...stil works... do you have any improvement to the circuits?
ALDI do a similar unit with or without backlit LEDs set in the back of of the unit
Clive's reviewed that previously. Looks like a much nicer light. PCB is certainly far better engineered in the Aldi light.
The Aldi one looked a lot better. I brought it across with me in the belief that the chip was doing some clever sensing from a standard PIR sensor, but this unit made me realise the PIR is probably digital.
Its only screwed to the side of your house. What could possibly go wrong?
A good bang for your money and a quick call to the firebrigade?
I've got a few of the versions where it only comes on temporarily when it detects movement, they've been there a couple years now and have been working fine. No idea if the battery is charge regulated.
The maximum voltage to the MCU's input pin is 4.2V PLUS the diode drop, not minus the diode drop. About 4.4V or higher.
If you use these in somewhere like Arizona, you will soon have a lawn with a series of scorch marks where the lights used to be.
No charge circuitry but it has a rechargeable 18650 that's just genius
I have a couple of versions of these lights around, one screwed to a wooden shed. I haven't looked at the circuit (I know, I know) maybe I should....
Gracias.desde colombia quisiera que me ayudaras como dejo que la luz quede permanente o que el tiempo sea muho mayor al que tiene. Ocomo anulo el sensor de movimiento un diagrama por fa.esplicas muy bien
Hi there, i know this video is quite old. But what would you think will happen if i replace the batteries with lifepo4? Would it be better?
Something I've discovered is that if you use a protected cell, or add a protection circuit to an 18650, you can directly charge the cell from a 6V solar panel. The protection circuit seems to do a good job of preventing the input voltage from pushing the cell voltage too high.
So if that's acually a cell with an integrated protection circuit, it's probably okay.
If it's an unprotected cell, they may be relying on the idea that the solar cells can't fully charge the lithium cell over a full day's sunshine with their limited current-supply capacity, which as you pointed out is a bad idea when the light isn't left on continuously overnight.
Summary: You're right on all counts. It's a bad circuit design.
Great, where was this video a week ago? I just ordered one of these off Ebay.