How best to find the "best" small solar panel of a particular size?

[NeverDie]

@mfalkvidd
It's a pity this topic hasn't attracted more interest. I guess people are happy with their batteries.

[mfalkvidd]

There seems to be a sweet spot at 6V 1W when it comes to price/performance, at least specified performance. Smaller panels are relatively expensive, and so are larger. I have ordered four of these but they haven't arrived yet.

I have been looking for flexible panels, something that could be folded or rolled up, but haven't found anything with a decent price so far.

[NeverDie]

I think that's because that's where the market demand is. It's hard to find smaller panels generally. I have these on order, but neither has yet arrived:

https://www.aliexpress.com/snapshot/0.html?orderId=83597824322273&productId=32253274314

https://www.aliexpress.com/snapshot/0.html?orderId=83519564882273&productId=32644217312

[gohan]

The problem of adding up panels is that your node would get quite big. The other day I was listening to a podcast and they were talking about a new sensor device with a special paint on the case that acts as a solar panel that charges the internal battery/supercap that needs to be much smaller than usual battery powered sensors

[NeverDie]

@gohan said in How best to find the "best" small solar panel of a particular size?:

The problem of adding up panels is that your node would get quite big. The other day I was listening to a podcast and they were talking about a new sensor device with a special paint on the case that acts as a solar panel that charges the internal battery/supercap that needs to be much smaller than usual battery powered sensors

Yes, that's why I mentioned that Solar Bits (which are just 7mm x 22mm in size but deliver 4.7v) might be a good fit for multiple panels in series. The sum of their areas wouldn't be large. But they come at a price... https://www.digikey.com/product-detail/en/ixys/KXOB22-01X8F/KXOB22-01X8F-ND/4840081

Do you have a link to the podcast? Never before heard of a paint that produces solar power. Is it something I can buy?

[NeverDie]

@mfalkvidd said in How best to find the "best" small solar panel of a particular size?:

@NeverDie Interesting. Let us know what result you get.

I just now tried four 6v panels powered in series by moonlight just to see what would happen, and it produced only 0.06v. So, scratch that idea.

[mfalkvidd]

@NeverDie ths is slightly off topic, but can you recommend any rectifier with low drop voltage to use to prevent the solar panel from discharging the supercap/battery? I'm not comfortable with SMD, so I need to find something that's big enough to hand solder, has low drop and doesn't cost too much.

[gohan]

@NeverDie I have to look at it but it's in Italian and the paint it wasn't just simple paint, but different layers of semiconductor materials printed on the case, so something quite difficult to do at home.

@mfalkvidd I usually seen schottky diodes on the solar panel for that purpose.

[NeverDie]

@mfalkvidd

Not sure about through-hole blocking diodes. I'm using 1PS79SB30, which is SMD, but reverse current leakage is just 10na. Practically all the new good stuff is SMD, so you're severely impaired if you don't make the jump to SMD. SMD is nothing to be afraid of. Presently I'm hand soldering everything, including all SMD parts.

[mfalkvidd]

@NeverDie thanks. How do you handle lead times för pcb manifacturing? Or is it possible to buy prototype pcb cards? I'm worried project lead times will 10x if I need to wait for pcb milling for each iteration.

[NeverDie]

@mfalkvidd said in How best to find the "best" small solar panel of a particular size?:

@NeverDie thanks. How do you handle lead times för pcb manifacturing? Or is it possible to buy prototype pcb cards? I'm worried project lead times will 10x if I need to wait for pcb milling for each iteration.

I'm not sure how others handle it, but what I do is solder each individual SMD part onto pre-made breakout board for SMD parts. Here's an example of one I use frequently: https://www.aliexpress.com/item/100pcs-SOT23-SOP10-MSOP10-Umax-SOP23-to-DIP10-Pinboard-SMD-To-DIP-Adapter-Plate-0-5mm/32769473900.html?spm=2114.13010608.0.0.exgj0t
Then I attach header pins to it and complementary header pins onto a general purpose protoboard, such as: https://www.aliexpress.com/item/10pcs-3X7cm-3-7cm-Double-Side-Prototype-pcb-Breadboard-Universal-for-Arduino-1-6mm2-54mm-Practice/32655345057.html?spm=2114.13010608.0.0.exgj0t
Then I literally plug the SMD part into the protoboard. I then solder wired connections among the various parts. Sometimes I'll use breadboards, but I prefer this method because the wiring on breadboards can get rather loose, which becomes a headache, whereas doing soldered wire connections on protoboard, the connections are great quality and "made" forever until you decide to change them. I do it this way first rather than going straight to PCB because of the long PCB fab time (on average around 2 weeks for me from OSH PARK). Once I verify the design is correct using the protoboard, then I'll translate it to PCB as the final step. Also, while I'm waiting for the PCB to arrive, I have a solid working prototype, so I can continue working on whatever is next and the turnaround time for PCB fabrication isn't quite so painful.

Does that answer your question?

[mfalkvidd]

@NeverDie it answers my question completely. Great stuff. Thanks a lot!

[NeverDie]

Back to the OP. I finally received these smaller solar panels:
https://www.aliexpress.com/item/10Pcs-Lot-5V-30mA-53X30mm-Micro-Mini-Small-Power-Solar-Cells-Panel-For-DIY-Toy-3/32811111041.html?spm=2114.01010208.3.10.2aPJ36&ws_ab_test=searchweb0_0,searchweb201602_5_10152_10065_10151_10130_10068_5160013_436_10136_10157_10137_10060_10138_10155_10062_10156_10154_10056_10055_10054_10059_100032_100033_100031_10099_10103_10102_10096_10147_10052_10053_10050_10107_10142_10051_5150015_10084_10083_10119_10080_10082_10081_10178_10110_10111_10112_10113_10114_10037_10181_10183_10182_10185_10033_10032_10078_10079_10077_10073_10123_10120-10033_10077,searchweb201603_4,ppcSwitch_4&btsid=98b799be-2ff3-401c-aaeb-fa49f82cdc7c&algo_expid=a78e15b1-b89c-40bf-849b-b9f99d840f0d-1&algo_pvid=a78e15b1-b89c-40bf-849b-b9f99d840f0d
However, they're disappointing. They do produce 5v if put into direct sunlight. However, indoors, they produce maybe 1/3 the open-circuit volts of a larger mini panel. Instead, I had expected it would produce the same open circuit voltage, but simply able to produce less current.

So, I think what may be going on is that these panels may have some kind of internal leakage. In direct sunlight, that leakage is relatively small. However, in the less good mini solar panels, that internal leakage starts to dominate as lighting conditions get dimmer.

I've tested even smaller mini solar panels from Digikey, and they do manage to produce much more voltage under dimmer lighting than these cheap ones from China, which I'm now doubting will be of much use to me.

[gohan]

70 cents a panel... doesn't get more Chinese that this

[mfalkvidd]

@NeverDie too bad. Thanks for sharing your findings.

[NeverDie]

I suppose if you were to boost them, then for some inside applications, like a potted plant moisture meter, they would be adequate because for that you don't actually need updates all that often. So, they may not be a complete loss, but they aren't as good as even smaller panels that you can buy on digikey for more money. For instance, in my testing, this micro panel performs much better indoors than the cheap Chinese ones, and it's smaller too: https://www.digikey.com/product-detail/en/panasonic-bsg/AM-5610CAR/869-1009-ND/2165194

[novicit]

I just completed a battery-free outdoor node reporting temp/humid based on the information in this thread. Works great, thanks to @NeverDie for his insights. Node is a sensebender micro, RFM69W, 10F supercapacitor, and 5V solar panel. Reporting every 4 minutes it sends temp, humid, supercap %, & solar V. Uses the simple tiny solar charger schematic from @NeverDie shown above.

Results: The supercapacitor cost $2 at Digikey, is about 2/3 the size of a AAA battery, and lasts 40 hours without the solar panel connected. The 40 hours is when the supercapacitor reaches 1.9v as the low end for operating the node. Easily gets through the longer dark winter nights.

The solar panel is from China, Ebay, 110mmx70mm, 5V, 1.25W. The panel works great also. It reaches 4V in the smallest amount of light, and 5.45V (no load) in modest indirect light. In fact, it will fully recharge the supercapacitor in the morning before the sun even rises above the horizon, just from the sky brightening. (From the ~20% overnight discharged state.) It seems to charge a fully empty supercap in ~10 minutes under average light.

Bottom line: no longer walking through 1ft of snow in 10 degrees (F) to get the node, bring it inside, remove water seals, and change the batteries. This is the solution I was looking for.

Also have a 2nd outside node in testing using the Ceech solar board (A great PCB also!), a 18650 battery and the same solar panel. So far the battery has gone 2 months and is at 3.85V without the panel connected. In testing, the solar panel has no problem charging the battery. In fact so much, I am going to add a FET to disconnect the solar panel periodically to allow the battery to cycle some.

This site is excellent, such great talent, thank you all!

[NeverDie]

@novicit

Cool! That's great info. Thanks for sharing your findings.

[novas]

@novicit Could you share schematic and software of your node?

[gohan]

@novicit if you could make the battery work between 20% and 80% it will help the battery life.

[korttoma]

@novas said in How best to find the "best" small solar panel of a particular size?:

@novicit Could you share schematic and software of your node?

also links to the hardware would be appreciated

[NeverDie]

@novicit said in How best to find the "best" small solar panel of a particular size?:

Also have a 2nd outside node in testing using the Ceech solar board (A great PCB also!), a 18650 battery and the same solar panel. So far the battery has gone 2 months and is at 3.85V without the panel connected. In testing, the solar panel has no problem charging the battery. In fact so much, I am going to add a FET to disconnect the solar panel periodically to allow the battery to cycle some.

I'm not aware of any advantages to cycling a LiPo battery. As mentioned by Gohan, though, charging to just 80% does seem to extend lifespan.

Better yet, use a LiFeP04 battery. They're safer, and you don't need to down convert their voltage because their max charge is 3.6v.

I prefer supercaps mainly because they should last practically forever.

[alexsh1]

@NeverDie Did you look at @ceech solar harvesting boards?
My experience with solar panels is this. There are so many of them with the claiming same spec,. but behaving differently in real life. I think the same reason is quality. I have a weather station using ceech's board with temp/hum/baro/lightning/light sensors and nrf24l01+, 5V 1.2W solar panel and 3.7v 400mA LiPO. This has been running for some time. I was actually surprised that the battery is being charged even on a dark winter day. Some harvesting even is possible without the sun.

Now I have been looking at replacing LiPO with a solar harvesting board and a few supercaps. Do you think 10F would be enough? Or shall I go for something higher like 20F or 50F?

[wallyllama]

@NeverDie were they all at the same angle to the light source? If you laid them out on a table and shone a lamp, theynwill be at different angles and have different results.

[alexsh1]

@wallyllama I would say that in the similar circumstances different solar panels do behave differently due to quality solar cells, i.e. manufacturing quality as well as efficiency.

[NeverDie]

@NeverDie said in How best to find the "best" small solar panel of a particular size?:

@novicit said in How best to find the "best" small solar panel of a particular size?:

Also have a 2nd outside node in testing using the Ceech solar board (A great PCB also!), a 18650 battery and the same solar panel. So far the battery has gone 2 months and is at 3.85V without the panel connected. In testing, the solar panel has no problem charging the battery. In fact so much, I am going to add a FET to disconnect the solar panel periodically to allow the battery to cycle some.

Thanks for the pointer to Ceech. I hadn't been aware of him, but I see that his Ebay store has a lot of interesting boards, including this one:
http://www.ebay.com/itm/Solar-Energy-Harvesting-power-supply-w-LTC3108-and-super-capacitor-storage-/331974364829?hash=item4d4b36ce9d:g:tXwAAOSwmLlX4Osy
which I may order just for fun.

Impossible for me to really answer your question, as it all depends on what your current drains are. A 10F supercap isn't really much compared to a 400mah Lipo.

[alexsh1]

@NeverDie OK, I have a low powered board based on ceech 328p (pro mini footprint) board with a few sensors and nrf24l01+. LiPO 400mA is an overkill - the battery almost never goes below 4V. I just have a spare one from my RPi3 UPS (UPS Plco).

[alexsh1]

This is proposed by Sparkfun - very simple diagram

[gohan]

That's a simple circuit, but it lacks the balancing of the supercaps or at least the individual overvoltage protection

[alexsh1]

@gohan you can go for one large cap. I think the idea here is to keep it very simple

[gohan]

The idea is to put them in series to increase voltage

[NeverDie]

It's a buck converter. Here's the datasheet for the chip that's on it: http://cds.linear.com/docs/en/datasheet/35881fc.pdf

So, it's for the case of having a solar panel that produces more volts than the target device can handle. It's a larger solar panel than the ones we've been discussing.

[gohan]

Even with a 5V panel you still need 2 supercaps in series or get the 5v supercaps but they have less capacitance. I have ordered those 5v 4F too so I'll have to see how they behave

[alexsh1]

@gohan can you ask what is it you are trying to use supercapacitors for? Which board? What's your sleep / TX consumption?

[NeverDie]

@gohan said in How best to find the "best" small solar panel of a particular size?:

Even with a 5V panel you still need 2 supercaps in series or get the 5v supercaps but they have less capacitance. I have ordered those 5v 4F too so I'll have to see how they behave

It depends on how you want to handle the charge termination. If all you want to do is use a blocking diode, as in the sparkfun schematic, then yes. However, remember that if you put two capacitors in series, they have half the capacitance. Furthermore, if you do charge them to 5v, you'll have to step down that voltage one way or another, or else you'll fry the radio on your node. That's why I've gone the direction of using a larger supercap, but terminating the charge when it gets near its rated limit. Yes, you can buy supercaps that are rated up to 5.5v, but they're relatively expensive compared to the 2.7v rated ones.

[gohan]

it is all a matter of what components you want to use: if you go 5v you can still use a LDO to power your devices, otherwise you need a boost converter at least for arduino. As soon as I receive those buck-boost converters I'll put them to the test, but so far they are in an unknown place (tracking shows they left china a month ago and still missing any further info)

[NeverDie]

@gohan said in How best to find the "best" small solar panel of a particular size?:

it is all a matter of what components you want to use: if you go 5v you can still use a LDO to power your devices, otherwise you need a boost converter at least for arduino. As soon as I receive those buck-boost converters I'll put them to the test, but so far they are in an unknown place (tracking shows they left china a month ago and still missing any further info)

Well, yes and no. I think the challenging case is how to make it as small as possible but still low cost and fairly simple.

[alexsh1]

Personally, with exception of a few cases, I dropped 5V boards in favour of 3.3V ones. All transmitters - nrf24l01+ or rfm69 or rfm95 or eap8266 / esp32 are using 3.3V and can be fed by 2.7V

[gohan]

I think nobody is talking about using 5v boards

[NeverDie]

I think I'm trending toward including a 3.3v boost converter with a "pass-through". So, either:
https://www.openhardware.io/view/285/33v-Boost-Converter-with-Pass-Through
or
https://www.openhardware.io/view/279/Adjustable-Boost-Converter-with-Pass-Through
or possibly:
https://www.digikey.com/product-detail/en/microchip-technology/MCP1642D-33I-MS/MCP1642D-33I-MS-ND/5137719

That way, if you have a sensor that needs to be powered at 3.3v, or the voltage on the supercap has fallen below 1.8v, you turn on the boost converter, but only for as long as it's needed. All the other times, you leave it turned off and running your node directly from the supercap voltage, because boost converters are generally quite inefficient.

Thoughts?

[alexsh1]

@gohan perhaps I misunderstood you - why do you need 5V caps?

[gohan]

because if you have a 5v solar panel you can charge it directly from the solar panel and then use an LDO to get the 3.3v or a buck-boost converter to always have 3.3v (in case you are using rfm69hw that requires a little higher voltage than rfm69h). Of course you could have a 5v->2.7v buck converter to charge the supercap and then use it to directly power your boards. My only biggest concern it to avoid overcharging the supercap, so using a 5.5V supercap you have a safe margin when using a 5V solar panel (could be a 6V panel and add 1 or 2 diodes to drop voltage a little )

[alexsh1]

@gohan I suggest looking into this if you want to avoid overcharging supercaps:

Automatic cell balancing prevents overvoltage damage to either supercapacitor while maximizing charge rate.

[gohan]

I have seen theese too, it only gets more complex/expensive to make the circuit and source those special components

[alexsh1]

@gohan Did you see this?
Click here

@ceech is selling a few interesting solar harvesting boards.

[NeverDie]

@alexsh1

Here's the thing: when you consider the "quantity 1" price of either the chip or Ceech's boards, you could boost your voltage more cheaply by just buying a bunch of cheap solar panels and stringing them together in series. For instance, today's Digikey price for the BQ25770 is $8.12: https://www.digikey.com/product-detail/en/texas-instruments/BQ25570RGRT/296-37014-1-ND/4430487

I'm not sure how they price these chips, but it seems like they price themselves out of much of the market to the point where they become almost like novelty items.

BTW, I like what I can see in Ceech's boards, and I think they're fairly priced considering the cost of the components. I'm just trying to be objective about this.

[NeverDie]

Likewise, I don't understand by the LTC3525D-3.3 (above) is about 4.5x the price of the MCP1642D-33 (also above), especially when the latter chip actually appears to be superior, at least on paper. Am I missing something?

[Edit: Answering my own question, it appears that the LTC3525D-3.3 has a much lower operational quiescent current. The MCP1642D-33's appears to be around 40x higher. So, that might definitely matter if you're planning to have the boost converter running all the time. ]

[NeverDie]

I went ahead and made the switch on the node that I'm experimenting with: I'm now using a LTC3235D 3.3v passthrough boost converter. I switch it on when the node wakes up, do various measurements, then switch it off before transmitting the results if the supercap voltage is above threshold. Everything working smoothly.

I think I may next try putting a buck converter on the solar panel so I can get more juice out of it when its voltage is above 3.3v. I'll be trying this one:
https://www.openhardware.io/view/276/33v-Buck-Boost-DC-DC-Converter
because it's the only one I have on hand. If it seems to help, then I'll look for something more on-point (and less expensive!).

[NeverDie]

UPDATE: I finally received these:
https://www.aliexpress.com/snapshot/0.html?orderId=83519564882273&productId=32644217312

Even though they are smaller, under dim lighting, they showed ~4x the open circuit voltage as compared to: https://www.aliexpress.com/snapshot/0.html?orderId=83597824322273&productId=32253274314

That said, the panasonic solar panels (which are even smaller!) are about 4x their voltage (i.e. about 16x the 53mmx30mm ones) under the same dim lighting.

So, regarding the OP: yes, there are definitely quality differences that express themselves in dim lighting, and this is proof.

[mfalkvidd]

@NeverDie Interesting find. Could you post the product links instead of the order links? Order links can't be visited without being logged in as the user who made the order.

[NeverDie]

@mfalkvidd said in How best to find the "best" small solar panel of a particular size?:

@NeverDie Interesting find. Could you post the product links instead of the order links? Order links can't be visited without being logged in as the user who made the order.

OK. The bad ones were these: https://www.aliexpress.com/item/10Pcs-Lot-5V-30mA-53X30mm-Micro-Mini-Small-Power-Solar-Cells-Panel-For-DIY-Toy-3/32811111041.html?spm=2114.01010208.3.9.Qf7wa3&ws_ab_test=searchweb0_0,searchweb201602_5_10152_10065_10151_10130_10068_5010015_10136_10157_10137_10060_10138_10155_10062_10156_437_10154_10056_10055_10054_10059_303_100031_10099_10103_10102_10096_10147_10052_10053_10050_10107_10142_10051_10084_10083_10119_10080_10082_10081_10178_10110_519_10111_10112_10113_10114_10181_10037_10183_10182_10185_10032_10078_10079_10077_10073_10123_10120_142-10077,searchweb201603_9,ppcSwitch_5&btsid=191df259-f87c-4a83-9576-00e59515e886&algo_expid=870c1458-2143-42ea-88c3-cdd4fc63f450-1&algo_pvid=870c1458-2143-42ea-88c3-cdd4fc63f450

I can't post a link now to the good ones, because it looks as though they are no longer being sold. I had posted links before that worked. Believe me, no one is sorrier about this than me, because I want to order more of them.

[gohan]

Maybe somebody else figured too that they are good and got all of them

[NeverDie]

I just today started using a smart diode between the solar panel and the supercap. Voltage drop across it is a mere 0.03v, so I assume that means more power is faster transfered from the panel into the supercap.

[gohan]

@NeverDie said in How best to find the "best" small solar panel of a particular size?:

smart diode

what's that?

[NeverDie]

@gohan said in How best to find the "best" small solar panel of a particular size?:

@NeverDie said in How best to find the "best" small solar panel of a particular size?:

smart diode

what's that?

It's this:

It better approximates an "ideal diode" than an actual diode does. So, it behaves like a diode, but the voltage drop is much less. I measured the reverse current leakage on it, and it's about 1ua, which is higher than some of the best diodes, but not bad.

[NeverDie]

TI also makes a single integrated "smart diode": http://www.ti.com/product/SM74611
which I've also tried. However, it has a bigger voltage drop, so I don't like it as much. Also, TI's chip is more expensive.

[NeverDie]

Back to the OP, here's a screen scrape of what had been the "good" small solar cells:

Here's a closeup photo of one:

It may not show so well in the photo, but what I think may be key is that it wasn't pieced together out of scrap solar cell fragments (like the "bad" one was) but rather appears to be a single piece. So, that may be a clue on what to look for in the future.

[NeverDie]

Here's a schematic for the ideal diode that is pictured above:

[gohan]

Do those solar panels handle outdoor weather?

[NeverDie]

Not sure. There's probably a risk of UV damage to the epoxy, but that's just a guess.

At the moment, I'm focused on indoor use, which is a largely ignored category. I think outdoors is easier because, well, it's a lot brighter. For instance, novicit (above) reported "The solar panel is from China, Ebay, 110mmx70mm, 5V, 1.25W. The panel works great also. It reaches 4V in the smallest amount of light, and 5.45V (no load) in modest indirect light. In fact, it will fully recharge the supercapacitor in the morning before the sun even rises above the horizon, just from the sky brightening. (From the ~20% overnight discharged state.) It seems to charge a fully empty supercap in ~10 minutes under average light."

[NeverDie]

If anyone has a favorite buck converter chip, or one they want to recommend, I could possibly make a PCB for it and post it on mysensors.org. Suggestions anyone?

[NeverDie]

I received the 0.7v to 3.0v boost converter module from Aliexpress whose link I had posted earlier above. Can anyone identify the chip which is on it?

[Nca78]

@NeverDie it is a ME2188 from MicrOne
http://www.sz-hxdz.com/uploadfile/pdf/201605111221584282.pdf

[ceech]

Anyone interested in super capacitors.
I'm working on a backup super capacitor storage board. It connects directly to the power rail and charges and balances super capacitors bank. When and if power fails the board steps in and provides power to the circuit.

It is based on LTC3110 super capacitor bi-directional charging IC.

[alexsh1]

@ceech interesting board. Is it available that order or not yet?

[ceech]

@alexsh1 Boards will be most likely available by the end of the week.

[Tmaster]

why use supercaps on an arduino tha consumes so low current? that caps will not discharge during night,when solar panels are not producing energy.?

[gohan]

They will not discharge during just one night

[NeverDie]

@ceech
Isn't it more economic to buy just one much larger supercap, with say 16x the Farads of the supercaps in your photo? Or does a capacitor bank have some other advantages besides just higher Farads?

By the way, have you found supercaps that you like which also have very low self-discharge? I've been using leakage as a proxy for self-discharge--because self-discharge data is hard to come by--but they are actually defined differently.

[ceech]

@Tmaster They will slowly discharge. And depending on quality discharge times can vary a lot. The ones I'm using have a leaking current of 6uA.
Of course not always is a super capacitor a suitable replacement for a battery but where batteries are not available, suitable or safe, super capacitors can take their place. They also have some advantages - high cycle life of more than 100.000 cycles comes to mind.

[ceech]

@NeverDie Not necessarily. Larger super capacitors tend to be more expensive than smaller ones. Anyway, the board on the photo is a proof of concept. I would like to see how balancing and bi-directional charging/discharging work. Besides, I have some super capacitors left from other projects.

[gohan]

If you could make one the smallest as possible just for 2 supercaps, it would be a nice thing to have to integrate into an existing project to give it a little backup power.

[NeverDie]

@gohan said in How best to find the "best" small solar panel of a particular size?:

If you could make one the smallest as possible just for 2 supercaps, it would be a nice thing to have to integrate into an existing project to give it a little backup power.

But I'm guessing you want them in series, not parallel. Am I right? I suppose that might be useful actually. Ceech's seems to be parallel, though I can't say for sure whether or not it connects two banks of parallel caps in a series configuration or a parallel configuration.

@Ceech What voltage does it supply? Are they 2.7v supercaps, or something else?

[gohan]

of course I want them in series

[NeverDie]

You know what would be interesting? A capacitor bank that accepts charge in parallel but gives it back in series after a voltage threshold has been reached. That way you could charge it with very weak voltage (e.g. from a solar panel in very weak light), and still get a useable voltage out of it without the huge efficiency losses you usually get from a boost converter running at very low currents. Of course, the ESR might be very high (since it adds) if it wasn't very low in the supercaps to begin with. Low ESR supercaps are certainly available though.

@Ceech What do you think?

[wallyllama]

@NeverDie you mean a like battery from a hybrid car? Cells form a pack, but monitored(at least) and [possibly] charged individually. I believe that is how they are set up. It might be a way to find a reference on how to do it.

[ceech]

This is the capacitors schematic:

Two caps are in series, pairs in parallel. IC only allows for up to 5.5V. But voltages can be set - charge voltage, backup voltage, trip voltage and charge current can all be set. @NeverDie There is a solder jumper pad on the board for low current efficiency selection. Some other DC-DC converters also have this option.

[NeverDie]

Charge pumps work on this principle. It's just that they dump their higher voltage from the series configuration into another capacitor, whose voltage eventually rises to the series voltage. I guess that could work too, but you need switches which switch at very low voltages to pull it off if you're starting from a cold-start and, say, a solar panel is only providing you with extremely low voltage. I'm not sure how far below 0.6v you can find switches that still function. Same for the oscillator that typically drives a charge pump. That's why Ceech's board (that I linked earlier above) that works with a startup voltage of as low as 20mv (?) is so impressive. Or is that its lowest operating voltage, and it still requires a higher start-up voltage? I guess I should read the chip spec sheet.

[ceech]

@NeverDie 20mV is stat-up voltage.

[alexsh1]

@ceech very impressive Roman! Keep us posted when the board is ready for orders

[NeverDie]

@ceech Is there a particular brand/source of supercaps that you recommend to fit your PCB? Or do they come pre-installed?

[ceech]

@NeverDie I like those:
http://www.mouser.com/ds/2/40/AVX-SCC-1018831.pdf
Super capacitors will be installed on boards.

[alexsh1]

@NeverDie @Nca78

I must admit, I got totally lost in the middle of this thread with so many links.
Did you manage to find a holy grail? I'm keep to change a 400mA LiPO for supercapacitors on one of my nodes

[NeverDie]

@alexsh1 said in How best to find the "best" small solar panel of a particular size?:

@NeverDie @Nca78

Did you manage to find a holy grail?

I haven't received the supercaps from China, so I can't comment on them as yet. I expect they should arrive by next weekend, or possibly before. However, there are lots of other supercaps that one might use.

I still think that for most people a simple 6v solar panel for around $1 on Aliexpress, together with https://www.openhardware.io/view/382/Tiny-Solar-Charger-for-27v-Mote-Supercap
and a sufficiently large supercap is all you're going to need for most things. If you have a sensor that needs 3.3v, then you will want to add a pass-through boost charger such as:
https://www.openhardware.io/view/285/33v-Boost-Converter-with-Pass-Through
or
https://www.openhardware.io/view/279/Adjustable-Boost-Converter-with-Pass-Through
that you enable when your node wakes up and disable before going back to sleep.

One can get more elaborate to handle more challenging low-light cases, but for 80%+ of the cases, I would expect the above to be enough.

Fortunately, all this stuff is fairly inexpensive and easy to assemble.

[Terrence]

I have this Solar Powered IoT Device Kit from Cypress....but have not started using it yet. though I would post it in case you might find the BOM useful.

http://www.cypress.com/documentation/development-kitsboards/s6sae101a00sa1002-solar-powered-iot-device-kit

[alexsh1]

@NeverDie Thanks for your suggestion. Here is a thing - I cannot find MAX8887EZK27 anywhere.

Meantime, I did manage to find this interesting article - https://www.ti5.tu-harburg.de/publications/2009/fgsn09_lifetime.pdf

And these boost converters on the e-bay:

1. http://www.ebay.co.uk/itm/ClosedCube-TPS61221-Low-Input-Voltage-from-0-7V-Boost-3-3V-Converter-Breakout-/182051630756

2. http://www.ebay.co.uk/itm/ClosedCube-TPS61291-Low-Iq-Boost-Converter-with-Bypass-Operation-Breakout-/182304011227?var=&hash=item2a722a73dbmCSTFZtUluPLYwuCR7RYaeQ

A bit expensive, but based on a very advanced chip which may be used with supercaps.

[NeverDie]

@alexsh1 said in How best to find the "best" small solar panel of a particular size?:

@NeverDie Thanks for your suggestion. Here is a thing - I cannot find MAX8887EZK27 anywhere.

https://www.digikey.com/product-detail/en/maxim-integrated/MAX8887EZK27-/MAX8887EZK27--ND/6227255

[alexsh1]

@NeverDie it is a no go for me at USD 18 delivery charge to the UK. Unfortunately, there are no other sources for this chip.

[NeverDie]

You'll need to substitute a different LDO for it then.

[mfalkvidd]

@alexsh1 in case you don't already know, octopart is great for finding sources. https://octopart.com/search?q=MAX8887EZK27
There seems to be only american sources for this component though

[alexsh1]

@mfalkvidd

@alexsh1 in case you don't already know, octopart is great for finding sources. https://octopart.com/search?q=MAX8887EZK27
There seems to be only american sources for this component though

I did not know about this source
Well, that's the problem. The postage is golden and then if the value is greater than 20 bucks (including postage), I have to pay 20% VAT.
:(((((

[ceech]

@alexsh1 @NeverDie You can substitute the MAX8887 with MIC5365.
This whole thing got me so intrigued that I'm going to make a board myself. With charger, super capacitors and DC-DC converter. I'll use MIC5365 and for DC-DC conversion TPS610986. It has loads of options. And the whole thing will cost less than 10 bucks. I've made a schematic

[alexsh1]

@ceech said in How best to find the "best" small solar panel of a particular size?:

TPS610986

Excellent choice - TPS610986 - The TPS61098x
suits for low power systems very well, especially for the system which
spends the most of time in sleep mode and wakes up periodically to sense or transmit signals.

[alexsh1]

@NeverDie I am sure you saw this brilliant write up by Nick Gammon:

http://www.gammon.com.au/forum/?id=12821

He experimented with 0.47F capacitor. It was enough to run a mote (<10uA in a sleep mode) overnight

[NeverDie]

@alexsh1

I hadn't seen it, but a mote in sleep mode draws practically nothing (<300na for combined draw from both atmega328p and RFM69). It's the self discharge rate of the capacitor more than anything which seems to govern.

[alexsh1]

@NeverDie This is what he experienced! Have a look at the link. Within a few charges though, self discharge stabilised

[NeverDie]

@alexsh1

Regarding the choice of replacement LDO. If the LDO you pick turns out to have unacceptably large reverse current flow, just place the blocking diode after it instead of before it. That particular diode will leak only about 20na, and the entire issue doesn't even come up until the last ~0.3v of charging, where you'll then have a very long charge tail. However, if that were to be an issue, you could just use a larger Farad supercap, and then it wouldn't matter because you'd have enough charge already before encountering the long tail of the charge process.

[alexsh1]

@NeverDie Thanks for heads-up. I'll keep it in mind

I found an interesting cap -
https://www.digikey.co.uk/product-detail/en/vishay-bc-components/MAL219691204E3/4705PHBK-ND/5015889

Its voltage is very close to the solar panel so I think it may be connected directly via diode (0.7v drop = 6 - 0.7 = 5.3V vs 5.6V max rating)

[NeverDie]

@alexsh1 said in How best to find the "best" small solar panel of a particular size?:

@NeverDie Thanks for heads-up. I'll keep it in mind

I found an interesting cap -
https://www.digikey.co.uk/product-detail/en/vishay-bc-components/MAL219691204E3/4705PHBK-ND/5015889

Its voltage is very close to the solar panel so I think it may be connected directly via diode (0.7v drop = 6 - 0.7 = 5.3V vs 5.6V max rating)

I've tried that cap already, and I wouldn't recommend it. It has a fairly high ESR.

[alexsh1]

@NeverDie Given that you have been tinkering with it for a while, what caps can you recommend? With what ESR?

This cap has got 2.4 Ohm @ 1kHz - I think it is low?!