Foldable 50 W ultra-portable solar system for $150

comparatively low budget device for charging your camera/laptop/GPS/cell phone/sat phone/ham radio/flashlight batteries while backpacking

Similar projects worth following
- flexible, light weight, super low efficiency (6-10%) amorphous silicon solar cells, 60 W
- cells fully encapsulated in EVA with waterproof TPE backing and fluoropolymer (FEP) front sheet, sewn onto a foldable water-resistant tarp backing
- total weight only several pounds, compact package stored in a rigid contractor's clipboard
- 12 V output and battery
- 5 V output
- 18 or 19 V output for a laptop (actually, has variable output: 3V to 30V)

Goal is to be under commercial price and last 20+ years.

AFAIK the low end commercial price for a low weight foldable 50-60 W panel with charging electronics combined is near $300.

AFAIK other flexible panels can be bought for cheaper (e.g. >100 W for <$150) but they are much heavier and will only roll up into what appears to be a 4-6" diameter cylinder. DC-DC converters not included.

Thus, this project aims to be under $150, not counting labor (or fun) of course.

breakdown of expenses:

  • $49 – 10x 6.2 W amorphous silicon solar cells (total 3 pounds)
  • $75 – encapsulating polymer materials: FEP front sheet film ($40), EVA encapsulant ($15, but TPU or TPO might be better if you can find them), TPE back sheet ($20)
  • $17 – voltage conversion parts, misc electronic bits and pieces (under a pound)
  • $15 – 12V 5A-hr sealed lead acid battery (60 W-hr, 3 pounds)
  • $25 – tarp and contractor's clipboard (to store the solar panel)

total cost: $181, but the tarp and contractor's clipboard could fairly easily be replaced with free or nearly free parts, bringing the total cost down to only $156. This compares very favorably to commercial options ($300+) if you have the time to spend on this kind of project.

total weight: well under 10 pounds? not so bad!

  • 10 × 6.2 W Uni-Solar L-Strip amorphous silicon solar cells 62 W (ideal) for $49 shipped. Vmmp - 1.6 V, Immp - 3.9 A. eff.: 6-7%; 10"x14"
  • 1 × 2' x 7' x .001" FEP front sheet $40 shipped from McMaster-Carr
  • 1 × about 2 square meters of .5mm thick STR Photocap EVA (ethylene vinyl acetate) encapsulant usually, each cell gets sandwiched in two layers of this; half as much was needed because the cell backing is stainless steel. Tedlar (PVF) backsheet was also not needed for the same reason.
  • 1 × about 2 square meters of TPE backsheet Tedlar - Polyethylene - EVA
  • 20 × bendy wires (fine multi-stranded copper, insulated) to solder the cells together and solder them to the power controller (these I can get free)

View all 8 components

  • update

    Dylan Bleier05/23/2015 at 02:52 0 comments

    all future updates for this project will probably be posted here:

  • put together the voltage converter thingie

    Dylan Bleier05/17/2015 at 12:53 0 comments

    I put two blue led voltmeter displays on the voltage converter board and a fan on the input. It seems to run pretty cool (at least without a load). More pics soon.

  • Found a source for the films

    Dylan Bleier05/03/2015 at 03:01 0 comments

    I finally found a reasonably priced source. Who would have guessed it - McMaster Carr has something at a comparatively reasonable price, for once.

  • Power management

    Dylan Bleier04/30/2015 at 20:59 0 comments

    I got a combination buck / boost DC-DC converter for $13 shipped on ebay.

    It might need a cooling fan, we'll see. I will use it to convert the 15-20V from the panel to 12V to charge a standard 12V battery, or to charge a 18V or 19V laptop (there will be a way to switch between the two modes and connect the battery or the laptop. If the output is not stable and goes over 18 or 19 V, I will add an 18 - 20 V adjustable regulator circuit to the output to protect the laptop. When the main converter is in 18-19V mode, I will presumably be using it with a laptop, so I can charge USB devices from the powered ports.

    When charging a 12V battery, I'll use a $4 12V-to-5V module to charge USB devices from the battery.

    I'll also put $2 LED voltmeters on everything.

  • Status update

    Dylan Bleier04/28/2015 at 16:16 0 comments

    With $49 spent on the cells, I still need the following to finish the panel:

    • about 2.5 m^2 EVA (STR 15420P) encapsulant (ebay)
    • about 1.25 m^2 of TPT (Tedlar-polysomething-Tedlar) backing (ebay)
    • about 1.25 m^2 of fluoropolymer (or other suitable) front sheet (ETFE, PFA, FEP, etc)

    This should also be done for under about $150 total, since otherwise I might as well just buy a UniSolar 68 W panel which uses the same cells and ETFE coating, for under $200. However that panel only rolls up (does not fold) and probably weighs about 10 pounds alone. The purpose of the encapsulant film is to keep moisture out of the cells. The front and back films also help seal oxygen and moisture out.

    Research on fluoropolymer front sheets (the most expensive sheet part AFAIK):

    to complete the project, I also need:

    • power controller or DC/DC converters, to output 5V 1.0A (USB), 5V 2.0A (USB), 12V, 19 V for charging 18 and 19V laptop
    • small battery or supercapacitor to smooth out power output?

  • Did research and ordered amorphous solar cells

    Dylan Bleier04/25/2015 at 23:54 1 comment

    I learned that you need the right encapsulating polymers or your solar cells will fail after a year or two or three instead of lasting 20 or 30 or even 50 years. The wrong materials will corrode your cells, introduce water vapor, acetic acid, etc. which will kill your cells. It's also important to seal each panel or module properly to make sure there's no water vapor that gets trapped inside. Also, I decided that it doesn't make sense to use super brittle, non-flexible solar cells in a foldable, portable device for backpacking. That left me with CIGS or amorphous Si cells; CIGS cells were very attractive but the necessary sealing films are very hard to find in low quantity and CIGS cells are very sensitive to moisture. So, I ordered some a-Si cells and some EVA encapsulating film.

    semi-flexible thin-film solar cells (CIGS, CdTe, etc):

    • - cheap-ish option ($0.65/W at 50 W scale) with no bus wire needed
    • - pretty damage resistant (there are videos on YT of CIGS cells taking bullets in every single cell and continuing to function at 90+% power....)
    • - SUPER LIGHT WEIGHT (about half a pound for 50 W) - If I get my hands on the right encapsulating materials, I will have to grab some of these in the future.
    • - take up a lot of room because they are somewhat inefficient (around 10-11%)
    • - example product: 25 pieces NanoSolar "NanoCell" 2.6 W CIGS cells sold on ebay for $43 total shipped for a total of 65 W according to spec (since they "may vary slightly from stated specs" as they are liquidation surplus - and I saw one complaint saying the power was off). It has two contacts on the back, so no bus wire is needed, but the aluminum contacts may be a bit tricky to solder to.)
    • - degrades a little bit faster (about 1.0% decrease in power/year)
    • - not compatible with EVA sealing film! (need PVB film or ionomer products from Dupont such as PV5412 or maybe even UV-resistant Surlyn - these seem hard to find at low quantity!)
    • - very sensitive to the use of the wrong encapsulating material!
    • - very sensitive to moisture!

    semi-flexible amorphous Si:

    • - slightly higher cost?: about $0.8/Watt (at 50 W scale)
    • - somewhat damage resistant; flexible
    • - somewhat light-weight
    • - take up more room because they are pretty inefficient (around 6-7%)
    • - example product: 10 pieces Uni-Solar L-Strip 6.2 W - ebay - $50 shipped (I got this)
    • - non-standard cell voltage, but I can deal with that no problem
    • - degrade a little bit faster (about 0.9% decrease in power/year)
    • - you can cut them between the lines, parallel to them, but you'll need to solder new contacts on
    • - can encapsulate with well known EVA material which is field-proven to last at least 20-30 years when applied correctly

    (poly- or) mono-crystalline Si:

    • - low cost: about $0.7/Watt (at 50 W scale)
    • - most efficient (17-22%)
    • - somewhat light weight?
    • - easily damaged/cracked/broken (this is just the wrong application!!)
    • - degrade the slowest (mono - about 0.4%/yr, poly - about 0.6%/yr; use mono Si if you are building a residential or commercial solar power system...)
    • - uses standard .5 V cell voltage, but this can be a disadvantage if you want to get the 18 to 20 V required to charge most laptops from fewer than 40 or 50 solar cells
    • - require soldering bus wire to front & back – requires bus wire, flux pen, solder, & labor!
    • - example product: 20 pieces 5"x5" 2.6 W mono solar cells from China for $35 shipped, PLUS a bus wire, flux pen, and solder kit for $10-20
    • - can encapsulate with EVA

    I decided to go for the flexible amorphous silicon cells, since they will probably last much longer than the CIGS cells, and because a compatible encapsulating material (EVA) is available for them.

    The reason I need to encapsulate the cells in EVA is that it is non-corrosive, UV-resistant, flexible, clear, non-scratch, anti-reflective, and thus seals everything from the elements. The basics of the process are that you melt it on while pulling a vacuum on the cells to eliminate any air bubbles, but I will have to learn.

    Once I've wired together the cells and encapsulated...

    Read more »

View all 6 project logs

Enjoy this project?



helge wrote 05/04/2015 at 21:36 point

I see you found some cheap chinese DCDC components - something you simply cannot compete with on a budget. Moving on...

  Are you sure? yes | no

Dylan Bleier wrote 05/04/2015 at 22:39 point

well the goal is to get it to work adequately at first.  once I learn how to arduino, I will try to create my own DC-DC converter / charge controller that is more efficient and "smart"

  Are you sure? yes | no

helge wrote 05/05/2015 at 17:55 point

being on the working side of switch-mode power supply designs I cannot help but smile... "to arduino". I know you won't even try. Good luck, I'm outta here.

  Are you sure? yes | no

Dylan Bleier wrote 05/05/2015 at 18:15 point

almost any noun can be a verb if you want it to be.

"anything is a dildo if you are brave enough" - Abe Lincoln

  Are you sure? yes | no

helge wrote 04/30/2015 at 18:20 point

Obtained the SM72442MT ($10), not sure if it's worth the hassle to design a MPPT solar inverter with flexible / low profile magnetics (yes, there are such things as ferrite foil - but also somewhat expensive)

  Are you sure? yes | no

Dylan Bleier wrote 04/30/2015 at 18:49 point

Nice.  Looks a little too advanced for me.  I think I'm going to buy a simple buck/boost converter that's tunable with a pot, and perhaps manually determine the settings to get maximum power for charging a battery.

  Are you sure? yes | no

helge wrote 05/02/2015 at 21:22 point

did a quick test with my laptop (a lenovo): when too little current is supplied, the battery charging cycle will not begin - it just beeps and cycles every 1-2 seconds. LiIon and LiPo cells are limited in the amount of charge current, prompting the use of e.g. Li-Ion capacitors (see but they are amazingly overpriced)

gave me an interesting idea for my own laptop power related project. 

For your project I'd still recommend the laptop power bank with slight modifications for solar charging.

  Are you sure? yes | no

helge wrote 04/27/2015 at 06:44 point

well done so far, appears I only covered the mechanical aspects of the construction, good point regarding the humidity and contaminants issues.

  Are you sure? yes | no

Dylan Bleier wrote 04/27/2015 at 12:00 point

Do you know anything about dc/dc converters or power controller solutions?   

I do not...

  Are you sure? yes | no

helge wrote 04/27/2015 at 19:50 point

So far I had to implement flyback, boost, buck and SEPIC converters. The tricky steps are 1) selection of the appropriate topology 2) determining power components and feedback loop component values 3) actual  board layout. MPPT converters are somewhat more specialized.

That should give you an idea (synchronous buck-boost converter with dedicated controller). High side shunt monitor could be INA168 while the low side can be amplified by an opamp (e.g. TLV271, AD820, ...). A clever solution would also  be to just build a charger for a laptop power bank (however they are a bit pricey). 

Before going into more detail, you may want to do a market search for off-the-shelf solar charger circuits. Designing the electronics from scratch is rewarded with optimal integration at the expense of LOTS of work.

If you find a working circuit design, DFRobot and SeeedStudio offer PCB production at a budget. 

  Are you sure? yes | no

mr.jb wrote 04/30/2015 at 19:35 point

  Are you sure? yes | no

Dylan Bleier wrote 04/30/2015 at 21:10 point

cool.  I might go in that kind of direction with arduino or atmega microcontroller setup but first I want to get it working pretty cheaply and quickly.  If I try to make my own MPPT converter with a microcontroller, it will be a long project since I don't have that kind of experience.  But if I eventually do do that, I'll have something more primitive to compare it to.

  Are you sure? yes | no

mr.jb wrote 05/01/2015 at 06:04 point

Select your solar-panel voltage with care:

poor mans MPPT,  keep it simple stupid...
Estimate min. Vin needed for: step-down = battery float voltage (Vout) + 1 V
Suggested open voltage: solar panels    = 1.25xVin (MPPT)

I use like 5.5-6V for charging single cell LiFePO4

Note, you can apply my simple approach to "any" buck  converter

For "12v" systems is easy to find  Mppt:

It is HARD to find a MPPT for single cell LiFePO4 for large currents >3A.

I prefer to charge single cell since you don't need balancing ..and gps, led light is my primary target.

  Are you sure? yes | no

helge wrote 04/25/2015 at 19:22 point

have thin titanium tiles cut to PV cell dimensions + 5mm, attach PV cells and connect with thin metal strips, glue polycarbonate tiles on top of PV cell with soft polyurethane adhesive, then assemble array of tiles onto fabric, spray polyurethane adhesive all over it, cover with transparent plastic foil.

Yeah, it'll work but seriously, please pick a decent MPPT DC/DC converter solution instead of a linear dropout regulator for 50W!

  Are you sure? yes | no

Similar Projects

Does this project spark your interest?

Become a member to follow this project and never miss any updates