Interesting solar garden lamp

M

manuka

Senior Member
Yes-sorry- solar garden lamps are now so tedious a technology that the mere mention of them can bring on disdain. Hardware shops globally are stuffed with them, & it seems every season a variation on their theme arises, only to end up the bargain bins.

However I've just noted a "4 for US$10" offering here in NZ that could suit elevated PICAXE projects. For starters it's PV is slanted,which sheds rain & bird droppings,& also better matches non tropical solar angles. The AAA NiCd & (switched) electronics are normal,but space exists beside then for extra compact circuitry. The ~50mmm diameter suits easy elevation on a suitable plastic pipe, bamboo etc. Perhaps best of all, once aloft the unit would be almost unobtrusive.

Request- to suit a possible "Si Chip" application with a solar powered HopeRF 433MHz mesh data network. Can any Australian Forum readers please cast their eye over local hardware outlet garden lamp offerings, & mention if this style is now available in Oz. TIA- Stan.
 

Attachments

G

Grant Fleming

Senior Member
Hi Stan,
Since I am in the hardware 'warehouse' far too much these days, I can look out for these style of lights. The business end of these lights certainly are compact.

Cheers,

Grant
 
M

manuka

Senior Member
Have just spotted similar in Bunnings, c/w CdS sensor, at ~US$2.50 each (8 for ~US$20). They're even narrower, with the aptly named business end just ~35mm diameter. However the top PV slices ( 4 polyX giving ~ 2V @ 50mA) are flat mounted,& water proofing may thus be an issue.
 
D

Dippy

Moderator
You're fortunate. I can't find anything anywhere near that cheap in UK.

So, have I calculated correctly? To give me a 50W 12V panel eqiv it would cost me over $1000? (i.e. 500 off garden solar lights)
 
B

BeanieBots

Moderator
I agree Dippy, a factor of 20~25 times that of buying the equivalent panel.
However, you would also get 500 (or maybe a multiple of) LEDs for free and even if the battery is only 1 off at 500mAhr you would be able to store about 300Whr of charge.

Have a look in the petrol station. They often have bargain solar garden lamps at give-away prices.
 
B

BeanieBots

Moderator
Dippy, I actually got one of those about a year ago because its W/£ was very good when it was on special offer.
It works VERY well in overcast conditions. The integral blue flashing LED even flashes when indoors:) Unfortunately, the voltage dips signifcantly with each flash:(
It now sits on the roof of my shed keeping a 38Ahr Pb topped up.
Works during the summer but is not quite enough to overcome self-leakage during the winter with a max terminal voltage (fully charged battery) of 12.8v over the winter.
Will actually overvolt a 17Ahr battery without regulation.
 
D

Dippy

Moderator
Hi BB. I got one last year to keep my car bat topped while I was away.
It's not bad , never got it to over-volt a rice pudding though :)
(No, my car wasn't in the Garage!!)

You can't use your batt much if it can keep it topped up. 1.5W abs max downhill with the wind behind it. Mine checked out much less at midday/June/Sunny/perp-sun. Off-axis loss was huge too, something like 60% at 45 degrees from memory. But that goes with the territory so I'm NOT disappointed and reasonable value.
 
B

BeanieBots

Moderator
Agree with your findings Dippy. The 38Ahr battery is NOT USED AT ALL.
I "obtained" quite a few and was looking for a way of keeping them topped up.
People often forget that bigger batteries have bigger leakage currents so it doesn't always pay to oversize the battery.
 
D

Dippy

Moderator
Absolutely...padded out to 10 chars.

PS. Geez, you poor UK teachers. Just seen an Ad for job as headmaster of Swanage First School. i.e. Ankle-biters (4 to 10 yo ?) £52,000 ($100,000 US) per year. No wonder they go on strike for more pay. My taxes... aaargh.
 
Last edited:
M

manuka

Senior Member
Splutter- things rapidly got OT! This has got little do do with large battery charging arrays (which of course are increasingly abundant), or UK teachers $tre$$.

My post simply pointed out that dirt cheap solar garden lamps are now becoming available in a "stick" form, & that these may suit compact housing (& powering) of PICAXE circuitry. It's well known (thanks Dr_A etc) that the innards of classic solar lamps can make a nifty low current 5V solar supply. Given the flexible supply voltage needs of 08Ms,a self contained, cheap & very compact datalogger/433MHz/whatever setup could now evolve. Stan
 

Attachments

Last edited:
G

Grant Fleming

Senior Member
manuka said:
Splutter- things rapidly got OT! This has got little do do with large battery charging arrays (which of course are increasingly abundant), or UK teachers $tre$$.

I too was actually wondering where your thread was heading!
It is sometimes interesting and amusing to watch how these things evolve!

Cheers,

Grant
Click to expand...
 
M

Mycroft2152

Senior Member
I've got a few of them. I bought them to make some solar powered "Jar of Fireflies".

It is a very interesting design and circuit. The solar cell is only 4 thin strips, which generates barely 2.4 volts on a good sunny day.

Unfortunately the voltage output is very unstable and hashy. Great for white LEDs, but lousy for microprocessors.

Myc
 
M

moxhamj

New Member
Some email brainstorming is happening between Oz and NZ but just to summarise some thoughts.

Watts are very small but that is ok if you use them wisely. I've done some worst case calcs and running a picaxe continuously and allowing for a month of cloudy days you need at least a 5W panel. These solar light cells produce 11mW in bright sunlight, so the only solution is to run things intermittently.

But many applications work fine intermittently, eg sensing a tank level, measuring temperature, measuing soil moisture. The trick therefore is to get the quiescent current down.

We are brainstorming the schematics in post #11, particularly the one that uses a CDS lightcell. If that is replaced with a BC547 transistor with a very high resistor on its base (eg 100k to 1meg), and that is connected to a biggish capacitor (100uF?), then the oscillator will be kept off most of the time (it will think it is daytime). When the cap finally discharges, the circuit will come on, the picaxe will come to life, and it can measure things, talk to base etc. After 10 secs it can turn itself off by recharging the "turn off" capacitor.

I agree with mycroft - the output is hashy. I'd run it into a 1000uF cap, with a 7V zener across the cap, and then into a LP2950 5V reg. I've found that will run a micro. An extension of this circuit http://www.talkingelectronics.com/projects/PowerSupply5vSolar/PowerSupply5vSolar-2.html

Might still be a bit of a problem with that hash upsetting RF comms. If the storage cap is big enough, the picaxe could even turn off the oscillator, turn on the radio module, send the data, then turn the oscillator back on. Based on experiments I've done, the supply cap might need to go up to 4700uF.
 
Last edited:
M

Mycroft2152

Senior Member
Dr_Acula said:
Some email brainstorming is happening between Oz and NZ but just to summarise some thoughts.

Watts are very small but that is ok if you use them wisely. I've done some worst case calcs and running a picaxe continuously and allowing for a month of cloudy days you need at least a 5W panel. These solar light cells produce 11mW in bright sunlight, so the only solution is to run things intermittently.

But many applications work fine intermittently, eg sensing a tank level, measuring temperature, measuing soil moisture. The trick therefore is to get the quiescent current down.

We are brainstorming the schematics in post #11, particularly the one that uses a CDS lightcell. If that is replaced with a BC547 transistor with a very high resistor on its base (eg 100k to 1meg), and that is connected to a biggish capacitor (100uF?), then the oscillator will be kept off most of the time (it will think it is daytime). When the cap finally discharges, the circuit will come on, the picaxe will come to life, and it can measure things, talk to base etc. After 10 secs it can turn itself off by recharging the "turn off" capacitor.

I agree with mycroft - the output is hashy. I'd run it into a 1000uF cap, with a 7V zener across the cap, and then into a LP2950 5V reg. I've found that will run a micro. An extension of this circuit http://www.talkingelectronics.com/projects/PowerSupply5vSolar/PowerSupply5vSolar-2.html

Might still be a bit of a problem with that hash upsetting RF comms. If the storage cap is big enough, the picaxe could even turn off the oscillator, turn on the radio module, send the data, then turn the oscillator back on. Based on experiments I've done, the supply cap might need to go up to 4700uF.
Click to expand...
That sounds like a good idea. the 5.1 volt zener / 3 LED circuit is marginal.

I agree, you can't pull many milliamps from this.

The 4700 uFd cap is rather large and defeats the small size. Maybe an aero gel cap would be better.

There are some 60 mm square solar cells that will put out a reasonable voltage/miliiamp. They might be a better choice, They will directly charge a pair of AA cells.

Myc

These solar lights are perfect for making small solar powered "Sun Jars". They fit inside the rim of a small Mason/canning jar. A little epoxy does the trick
 
M

moxhamj

New Member
See this link http://www.talkingelectronics.com/projects/PowerSupply5vSolar/PowerSupply5vSolar-2.html

All solar lights have a stepup converter in them as it is not possible to run a led off a 1.2V nicad. If you hack the stepup converter you can get more volts out. See http://picasaweb.google.com/picaxe/SolarPowerPICAXE#5134088355783007314

There is an efficiency peak at 7V which is just where you want it!

Some more musings. With a 6-7V output (and thence into a low dropout, low quiescent 5V reg), the max current output is only 0.23mA. So when this turns on and wakes up a picaxe, the first thing the picaxe will need to do is to go back to sleep for a while, as the picaxe draws 3mA when awake. So it will need to power down for a while to let the volts build up on the charge capacitor. There may be scope for the picaxe to go from wake to sleep to wake at least when the sun is shining and keep the stepup converter running all the time. Or it might be better to shut down the stepup converter completely. This is a work in progress and I suspect in the next few days a number of breadboards are going to get a workout!

Re small size, it might end up being a long thin PCB that sits in the post of the solar light. Or maybe it might sit in the clear part where the led is. There should be enough room for a power storage cap though.
 
Last edited:
M

moxhamj

New Member
It probably doesn't need a clamp but I did play with one design with no load where the volts just went up and up and I chickened out and turned it off at 30V. That would zap the input stage of a 5V reg.

Also if you keep the volts low, you can use lower voltage caps which are smaller.

But there could be some smarter optimisations. While the picaxe is on, it could cycle between running and sleeping such that it draws the right amount of current to keep the volts at <7V. Or if the volts only rise slowly, it could turn on at 2V, go to sleep for (say) 10 secs, by which time the volts are 7V, send some data (radio drops volts to 6V) and then turn off the oscillator.

My gut feeling is to put the zener in for safety while debugging but set the software up so that the volts never get to the zener voltage. So maybe a 10V zener. Or a 10V zener and a led in series - and that led should never come on.
 
Last edited:
B

BeanieBots

Moderator
Good point about it running away over 30v. It just seems sacralidge to CLAMP the output of a small PV?
Without looking at the diagram, I'm sure it wouldn't be to difficult to use a zener from the output to switch off the oscillator and at least let the PV voltage go up. This would at least let the input cap store a bit more ready for when light levels (or increased load) bring it back down again.
Probably not much in real gain, it just doesn't seem right to dump PV power.
 
M

Mycroft2152

Senior Member
Dr A has it corrrect.

These solar lights only provide a very small amount of power. They are "optimized" to run a white LED - barely. It is really amazing how cheap nd minimalistic these are.

The amount of power form a solar cell depends on the amount of light and also the size of the solar cell. Yes, there are other factors, but these are tne main ones. These solar lights use 4 very thin strips of solar cells. I don't thnk they can be much thinner. Each 'stripe' produces abut a half volt.

The whole idea of storing the small amounts of power until there is enough to be used has been implimented very successfully in BEAM Robots by using a "Solar Engine".

Definitely worth a look before reinventing the wheel.

Myc

Norte; i get an error on the page message on the PICASSA link
 
M

moxhamj

New Member
I agree it is not good to dump PV power. Ideally, one would detect an overvoltage and turn off the stepup converter for a bit. I think in reality the turn on times are going to be so short - seconds at most, that a bit of lost power is not going to matter. Most of the time the picaxe will be off and the stepup converter disabled. Unless there is another cunning solution.

Re the picassa link - I too got errors when using Opera but it worked ok on Firefox.

I like the BEAM concept - both the idea of minimalist power, and also the idea of hacking consumer objects for different uses.
 
M

Mycroft2152

Senior Member
Dippy said:
A desperate acronym if ever :)
Click to expand...
Actually not bad, considering the blatant self promotion of one of the founders. He could have given P T Barnum lessons :) Though he was able to get a lot of government research grant money by wrapping it in CHAOS theory.

BEAM had its heyday just before inexpensive microprocessors became widely availible. Very complex "life like" actions could be generated by abusing digital logic gates in an analog manner.

BEAM was a great way to get into electronics at a very low cost by using recycled components.

BEAM robots had a very distinct style -- insect like in both their appearance and response.

Solar was a major part of the BEAM philosophy though not limited to it. There is a certain challenge to create a small robot that will run on a couple of square inches of sunlight. Sometimes watching them move was like watching paint dry.

Of course, there were also flashing LED circuits called Pummers.

One of the earliest BEAM competitions included the "Solar Roller"; a timed event that covered a 1 meter distance. The first "races" lasted 10 minutes or more. But that was reduced over the years by some very creative designs to under 10 seconds.

Very creative and uniquie circuits, and designs were the result of the minimalistic approach and many are still very useful. Solar Engines and low power motor drivers were optimized and worth a look for today's projects. There are also some very clever sensors that are easily interfaced to a micorrocessor in a "horse and rider" technique.

In fact, due to the ease of use and low cost of the PICAXE, a combination of BEAM circuits and PICAXE brain has been christened SMARTBEAM.

Yes, it was a fun time.

Myc

There is still some interest and vendors.
For more info check out the www.solarbotics.com website.
 
Last edited:
M

manuka

Senior Member
Have just opened a Golden Sun "stick" solar lamp that I'd initially mentioned, & taken macro pix of it's innards-below. Given the era (& it's sleek design),the minimal parts count & lack of SMDs is suprising. Minimal = 3 parts on the PCB (with one the usual 1N5819 Schottky), & the 4 leg ZE002 IC takes the oddball prize.

This ZE002 IC is no doubt acting as a LED driver -anyone run across such a beast before? Google mostly returns the likes of distracting Russian bride profiles,although a Hungarian site includes a snippet that mentions-
LED driver : ZE002 IC-vel 470uH-s tekerccsel => 0,9V-t&#258;&#322;l m&#258;&#711;r vil&#258;&#711;g&#258;*t a LED, 4,2V-n&#258;&#711;l nem akartam t&#258;¶bbet adni neki, mert a LED m&#258;&#711;r szinte szikr&#258;&#711;zott (Ibe=120mA)
Click to expand...
Phew- tantalizing figures (both Russian & Hungarian)! Stan
 

Attachments

Last edited:
M

moxhamj

New Member
I just picked up 4 solar lights today - will do some hacking in the next few days. I went for ones that had cds light sensors as I think they will be easier to hack by shorting out the light sensor with a 547 to simulate daylight. Then understanding the schematic won't be so critical.
 
B

BeanieBots

Moderator
Found a little more about it from "made-in-china.com". Where else?
Certainly looks like an all-in-one hybrid boost-driver with everything including inductor on-board. Works down to 0.3v. Max input 2.6v.

Here's the exact text I found.


&#21319;&#22739;&#39493;&#21205;&#25511;&#21046;&#38651;&#36335;&#65288;ZE002&#65292;ANA608&#65289; Boost-driven control circuit (ZE002, ANA608)

&#21319;&#22739;&#39493;&#21205;&#25511;&#21046;&#38651;&#36335;&#65288;ZE002&#65292;ANA608&#65289; Boost-driven control circuit (ZE002, ANA608)
&#29986;&#21697;&#31777;&#20171;» Overview »

&#20379;&#25033;&#21319;&#22739;&#39493;&#21205;&#25511;&#21046;&#38651;&#36335;&#20418;&#21015;&#65292;&#22823;&#37327;&#25033;&#29992;&#26044;&#33609;&#22378;&#29128;&#65292;&#25163;&#38651;&#31570;&#31561;&#12290; Boost the supply-driven control circuit series, a lot for Caoping Deng, flashlight, etc.. &#38651;&#36335;&#24615;&#20729;&#27604;&#39640;&#65292;&#24037;&#20316;&#31337;&#23450;&#65292;0.3V&#20197;&#19979;&#21855;&#21205;&#12290; Circuit cost-effective, work stability, 0.3 V the following start. IC&#26412;&#36523;&#30340;&#21151;&#32791;&#38750;&#24120;&#23567;&#65292;&#33287;&#20998;&#38626;&#22120;&#20214;&#30340;&#38651;&#36335;&#30456;&#27604;&#65292;&#25928;&#29575;&#25552;&#39640;20&#65285;&#20197;&#19978;&#12290; IC's own power is very small, and separation of the circuit devices compared to improve the efficiency of more than 20 percent. &#22826;&#38525;&#33021;&#38651;&#27744;&#38651;&#22739;&#38477;&#21040;0.3V&#20197;&#19979;&#26178;&#21855;&#21205;&#12290; Solar battery voltage down to 0.3 V below start. &#24037;&#20316;&#38651;&#22739;&#31684;&#22285;&#65306;0.9&#65293;2.6V&#65288;1-2&#31680;&#20805;&#38651;&#38651;&#27744;&#65289;&#12290; Operating voltage range :0.9-2 .6 V (1-2 section rechargeable batteries). &#21487;&#22312;5--100mA&#30340;&#31684;&#22285;&#20839;&#35519;&#25972;&#24037;&#20316;&#38651;&#27969;&#65292;&#35492;&#24046;10&#65285;&#20197;&#20839;&#12290; In 5 - 100 mA within the scope of adjustment of the current, less than 10 percent error.
 
M

moxhamj

New Member
I pulled a light to pieces and scoped the output. Unfortunately the output volts only peak at 3V - just enough for a led but not enough for a picaxe. I am wondering if it may be necessary to salvage bits from the light and build an oscillator from scratch?

I tried building this one http://www.talkingelectronics.com/projects/PowerSupply5vSolar/PowerSupply5vSolar-2.html and it sort of works - I get brief flashes and sometimes the led comes on for a few seconds. I suspect the components are not very well matched and are not oscillating properly, as I was reusing the 150uH coil from the light rather than the 0.082mH coil specified. I have had that one working in the past with the correct coil though.

Things that could be recycled though are the solar panel, blocking diode, battery and the case, and maybe the coil if this one is the same value in most lights (150uH does seem a common value).

That talking electronics circuit can go up to 15mA output as it is using a BC337. So this means there would not be a problem with needing to sleep the picaxe on startup.

I did manage to get the turn-off circuit working very well. A BC547 replaces the LDR. (collector to top, emitter to ground). 330k on the base. Cap between the other end of the resistor and ground. With a 4u7 charged to 1.2V it kept the circuit off for about 5 seconds. So multiply that by 100 to 470uF and it should stay off for 100 times longer. And even longer again if the cap is charged to a picaxe 5V output.
 
2

212

Senior Member
I hesitate to even ask, but do they not offer the lights with yellow LEDs in them there??? The ones I have use a 70ma panel and charge two nicad 500ma batteries.
 
M

manuka

Senior Member
I haven't seen anything other than single AA/AAA white LED types for years. Back in the 90s (& thus pre WHITE LEDs - which only arrived ~y2k), ~2V orange ones running from 2 AA NiCds were common however.

Here's the layout of that ZE002 "stick" version I'd initially mentioned. It's almost at the no brain level-you could hardly ask for a more straightforward approach. Brief testing under bright sun confirmed the 30mA PV output,with white LED current drain ~12mA in the dark.

EXTRA: I've also added a few ZE002 specs. -as much as my limited Chinglish can deduce. At just ~US$3 all up, the lamp innards look handy for low budget & tight timeframe "Joule Thief" style hacks & educational explorations, especially since they'd boost learners confidence by working right out of the box!. Yah!! Educators who're only too familiar with youthful "It's not working" circuitry will no doubt appreciate such immense benefits...

Of course, thanks to the near passing of small filament lamps,there are now numerous cheap LED driver ICs available from the likes of DigiKey,with Infineon's 4 lead offering no doubt similar to the ZE002 (See => http://www.infineon.com/dgdl/An101.pdf?folderId=db3a304313b8b5a60113d4239297042f&fileId=db3a3043163797a601167b0da6c21136 ), but these are usually hard to handle SMDs,with bulk orders needed. Stan
 

Attachments

Last edited:
M

moxhamj

New Member
Here is another issue to consider. I just pulled to bits a 3 year old solar lamp. It had stopped working last year, and on the inside it had been colonised by ants, and the solder had turned to oxide. The copper in the wires was more copper oxide than copper.

I am not sure how well these lights really work long term. No rain could get directly into the unit, so the oxidation would be more from mist and from condensation.

Also the epoxy that the solar cells are embedded in has gone opaque and dull and clearly is not as well UV rated as glass and silicon.

And the plastic case had gone brittle in the sun.

This would probably be expected for a 30 year old device, but not 3 years.

As I'm currently in the process of building self contained radio picaxe modules for a wireless mesh, I'm seriously considering how materials age and how waterproof they are. I'm looking at glass solar panels, aluminium frames, painted and treated wood poles, galvanized steel, and even sacrificial plastic covers (eg milk crates) that can be easily replaced but protect standard waterproof plastic cases underneath from UV. I'm also looking at potting the electronics in epoxy to protect the solder from oxidation.

Having had a good look through my local hardware store yesterday, I'm not quite sure whether any solar light on the market could survive 5 years in the elements.
 
M

manuka

Senior Member
Good point -although koala's & kangaroos are not a local issue (!), we have similar high UV solar & environmental conditions here in NZ. I've had a range of PVs up for ~10 years however, & thanks to an occasional wash & brush up, they're still performing well. At the solar PV level however the stainless steel versions around may help.

But gentlemen - we're not talking USAF $1000 hammers in this Forum- keep the nature of the quest in mind. Aside from the tight budget educational benefits (which those of you with deep pockets may be incredulous to learn often stretch to a just couple of US$ per student per year....), electrotech's ruthless rate of change may well mean that almost everything we're discussing may be considered quaint in 5 years.
 
Last edited:
D

Dippy

Moderator
Thats an important point Drac.
I've seen component legs turn to powder too from long term moisture. And those plated iron legs on many LEDS go rusty.

But, ask yourself, why are they so cheap?
They will be made from the cheapest materials on God's Earth and designed to the standard of what-we-can-get-away-with.
They will last the 12 month warranty and then, when they conk, you (the normal customer) will buy some replacements. That's how it works - with slight tongue in cheek factor.

Quality panels from proper manufacturers will last longer obv. hence price. It's not just the PV material cost but also paying some geezer to assemble it into a good enclosure.
Equally obviously you could buy properly encapsulated small products for a few dollars more - but the point of this thread was how to do it on the cheap as usual.

Or, I guess, you could rip the important bits out of the cheap chod and put it in a good enclosure maybe spray the electronics. At which point if you include your labour @ £20 / hour you may as well have bought quality in the first place.
You get what you pay for 99% times- all pretty obvious I'd have thought?
 
M

manuka

Senior Member
Dippy: This may be OT, but suggest you take a trip to 1.2 Billion population China- you may well be gobsmacked.

China's cost & income levels overall are ~1/20th that of UK (with even skilled workers only earning 50p = ~1US$ per hour), yet their facilities & technologies are often now far superior to the West. Many Chinese feel we "few" living in the West are the foolish ones, as we often pay too much for our tariff protected locally made manufactured goods, when they can make superior items much cheaper.

Henry Fords mass production benefits of course compare- "you get what you pay for" never applied to the mega popular 1908-27 Model T, even though it was (& almost by an order of magnitude) by far the cheapest & most durable car then available.
 
Last edited:
D

Dippy

Moderator
Briefly OT.

Stan, I would love to go to China. To see the history and culture and meet the people.


For once it wasn't me that mentioned China and I don't denegrate all products made in that neck of the woods. Especially as more and more of the 'proper' manufacturers have things made there; but made to the standards and q/a of the 'proper' manufacturer. It just depends who is making it. Many products are very good and excellent value for money.
Mass produced electronics, made to Western standards in proper factories, are amazingly good value and quality. But there is cheap and cheap. Cheap can be made to high standards in high volume e.g. branded consumer electronics. Cheap can be also be 'back street' factories where the major proportion of unit cost is labour.e.g. LED testers made using brown PCB with upright resistors. I will modify my YGWYPF algorithm slightly :)

Comparing Model T under the YGWYPF is invalid. It was a revolutionary step in manufacturing and nothing contemporary could be compared on a direct like-for-like. But if you are telling me that a Model T was better quality than a Rolls-Royce then I may be forced to disagree in a light-hearted way.
Engineering tools are another great example. Things will change obviously. I visited a couple of engineering works fairly recently. They won't touch stuff made in certain places. They buy Euro/North American or Japanese.

If I arrange for pcbs to be made low volume will be made in UK but higher volume gets made in China by reputable and checked manufacturers only (along with a painful 8 week lead time).

I also speak to UK importers who have visited China a few times to visit factories who have tendered to supply. The stories I hear.... but a good manufacturer is a good manufacturer wherever it is made. But a good manufacturer will always cost more on an exact like-for-like basis.
 
M

moxhamj

New Member
Well I never mentioned where my old solar lights were made. Truth is, I didn't look. And it doesn't matter where they were made in this instance, it is where they were designed that is important.

Any electronics out in the elements needs protection from UV, from moisture and from insects. That means it needs to be sealed properly - almost to an airtight standard. Or if not, it needs to sit in a shed or similar dry environment.

Tonight I had a bit of a eureka moment. I have my wireless nodes finally debugged and working on the bench. 4 Analog inputs, 4 analog outputs, 4 relays, PC interface, radio interface and wired interface using the pulsout protocol. And a 20x4 display so the data is easily visible.

These have taken too long to make to risk putting them in a solar light. Besides, I don't believe it is possible to build a wireless mesh unless all nodes are listening all the time, and that means either specialised "wake on message" modules, or cheaper modules with a 5mA draw. And that is way too much for a solar light - indeed after many tests in the field, I'm going for a 7aH SLA and a 10W panel.

Next step is robustly protecting everything from the elements.

And that isn't necessarily about cheap vs expensive. It is about thinking smart about the problem. Some of the plastic cases that are rated to a ridiculous IP weather rating have decomposed with UV. But cheap black poly pipe can handle the UV, as can 90mm downpipe.

There is a weather transpnder near my house that I'm going to copy. A thick wooden pole about 6 foot tall. Glass solar panel with aluminium supports on the side near the top. And the electronics are in a piece of 150mm white PVC pipe 30cm long, with two end caps.

I was kind of thinking of some satellite nodes that only transmit data that are sitting in solar lights. Units that only transmit can certainly run on the flea power of a solar light. But I think the electronics would need to be potted, and the battery encased in something waterproof, but also something that could enable the battery to be replaced easily. ? neutral cure silocone.
 
D

Dippy

Moderator
Sounds good Drac.

As you are a scientific man you might find this document from good old Zeus quite interesting re: plastics.
http://www.zeusinc.com/pdf/Zeus_Weathering_of_Plastics.pdf

I agree, many IP6* boxes are awful under pronlonged sunlight and high temps.

However, some specialised polymer enclosures include UV Stabilizers and some plastics (e.g. Acrylics, PTFE) are inherently more resistant to UV.
Guess what, they cost more.... so unless DIY then it WILL be more expensive. But nothing wrong with DIY.

What sort of plastic do they use for burglar alarm bell-boxes? The one on the front of my house is almost good as new and has been there 20 years. (OK it's made in UK for UK sun).
Poss a loaded polycarbonate?
And, of course, there is PVCu or uPVC.
I've found 'pure' ABS, Poly(ethy)lene and Nylon to be awful.
Black Acetyl Copolymer seems quite good but its a painful polymer. Machines fantastic but can't stick it.

I saw the info tabulated somewhere but blowed if I can find it.
 
You must log in or register to reply here.
Top