avatar Traveling light: solar power

By Barry Kauler
Page updated: September 22, 2014

This is one page of a series that I am writing on "traveling light", whether it be hiking in the wilderness or wandering the world by boat, bus, train or air.

If I wasn't going to be away from a power-socket for more than a day or two, it would be far simpler (and cheaper) just to bring the AC mains charger for my smartphone, and maybe just change the batteries for my torch. However, the geeky electronic engineer in me wants to be setup to recharge off-grid, to be totally self-contained indefinitely.

A note about the evolution of this page: when I first started to write it, I was not "up to speed" with modern technology. My knowledge was for charging of lead-acid batteries, nor did I fully appreciate some of the issues with the peculiar current practice of using USB ports for charging.
Some of my early solar panel and battery purchases were not very compatible. Anyway, as my knowledge got refined, I updated this page accordingly.

Mini-tutorial:
mAh and energy
This is a topic that needs to be explained before reading this page. The storage capacity, or energy, of a battery is usually specified in mAh, or so many milliamps of current over one hour. However, this is an awful way to specify energy. A 12V battery with a rating of 1000mAh, actually stores over three times as much energy as a 3.7V battery rated at 1000mAh.
We run into this difficulty with the batteries tested on this page, as the USB voltage is 5V (or thereabouts), whereas the internal battery is typically 3.7V.
Energy should really be specified in Wh, or Joules. "Wh" is volts*amps for one hour. I will try to do meaningful energy tests based on Wh readings.

So, making a start, looking at what is for sale in the online marketplace...

Portable 5V folding panels

These are some solar panels found at dx.com, for 5V USB charging. Higher voltage panels, such as those used to charge 9V and 12V batteries, are excluded. Note dx.com is a China/HongKong-based online retailer. This is a sample only, not a complete list of what they sell...

Product
Power (watts)
Battery (mAh)
Weight (grams)
Price (AU$)
Notes
miniisw SW-5W80 Foldable 5W 8000mAh Solar Powered Panel
5
8000
697
61.59
Folding cloth.
KP-8000 Foldable Solar Powered 5W 5V 8000mAh Dual Port Li-polymer Power Bank 5
8000
694
59.42
Folding plastic.
10.5W USB Output Foldable Portable Solar Panel Charger
10.5
-
507
51.10
Folding plastic.
Miniisw SUNWALK-070 Portable Foldable 7W Solar Panel Charger
7
-
469
40.38
Folding fabric.
WN-02 10W 2A Fold-up Solar Panel Power Battery Charger
10
-
445
56.01
Folding plastic.
Miniisw 6W Folding 1100mA 4-Solar Panel Charging Power Bank 6
-
350
27.57
Folding canvas.
Miniisw SW070B Portable 7W 1050mAh Solar Panel Battery Pack for Cellphone 7
-
345
34.45
Folding canvas.
Miniisw SW-050 Portable Foldable 5W Solar Panel Charger
5
-
338
29.03
Folding canvas.
W-01 6W Portable Monocrystalline Silicon Solar Power Battery / Cell Charger for Cellphone + More
6
-
335
32.73
Folding.
7W Folding Solar Panel Charger for Mobile Phone + Camera + More
7
-
332
32.93
Folding plastic.
Miniisw SP6W Portable Outdoor Folding 6W Solar Panel 6
-
312
33.31
Folding canvas.
ZNOODA SW-050 4W 4-Section Folding Solar Panel Power Battery Charger
4
-
275
34.06
Folding plastic.
WN-07 4W 800mA Fold-up Solar Panel Power Battery Charger
4
-
255
28.50
Folding plastic.
WN-09 5W 1A Fold-up Solar Panel Power Battery Charger
5
-
230
32.51
Folding plastic.
CS-450 Portable Folding Solar Power Charger w/ USB Cable for Cellphones / Camera + More
2.2
-
167
39.25
Folding plastic.

A note about the power ratings: take them with a grain of salt! Putting it bluntly, the vendors tell lies about panel capacities. For example, a few years ago I tested a "120W" panel, Chinese import, but couldn't get any more than about 60W out of it.
Apart from getting sufficient charging current, my main concern is weight, and all of those panels, 5W and above, are quite heavy.

Introducing pass-through charging

There is something else to think about also, and that is, just how to charge my smartphone. Can I plug the solar panel directly into it? If the panel is slung over my backpack while walking, charging is going to be fading in and out -- how will the smartphone handle that? Some panels claim to have smart electronics for this situation -- unfortunately, some cut-out and charging has to be manually restarted.

The idea of having an intermediate battery, what is referred to as an "external battery", may solve the "fade in and out" problem mentioned above.  Traditionally, a panel is plugged into an external battery, then after the external battery is charged, it is then plugged into the smartphone.
That, however, is a very slow and inefficient method. Slow, because of the two charging steps required, inefficient because of energy losses in each step.

However, in this page I have investigated what is known as "pass-through charging", in which the panel and the smartphone can be plugged into the external battery at the same time.
However, the external battery must be especially designed to support this, and most don't. Those that do, do not necessarily do it in the most efficient way possible (in fact, jumping ahead, none of those tested do pass-through efficiently).

The pass-through charging setup is very good, as it buffers the charging to the smartphone. That is, when the sun fades, the intermediate battery will ensure that there is still charging current to the smartphone. This is also likely to avoid stressing the smartphone.

Read on, the principle of pass-through charging will be further explained...

Rollable solar panels

I turned my attention to integrated solutions, with inbuilt battery, and more specifically the thin-film rollable panels as they are very light.

 I eventually found one product that ticks all the boxes, the Companion Epak Solaroll 4W, with 2600mAh battery. I bought it for AU$129, from Outdoor Shack. Here is the manufacturers page:
http://outdoor.companionbrands.com.au/index.php?p=line&sid=1382090590  
epak solaroll

Weight of the Solarol is 400g total, including light globe, USB cable and carry case.

The panel is rated at 4W, which is reasonable.
Opened, the battery hangs at the bottom, and measures 70cm x 19cm. A bit longer than I would prefer, when hanging off my backpack.

Here is a youtube video:
https://www.youtube.com/watch?v=MHOlmOv28Ew

Note, there are other rollable solar panels out there. Solarfilm, a US product, has a range, but all are 12V output (which can be regulated to 5V but that needs to be done with an efficient switching voltage-converter). They are also expensive.
Another US company, Bushnell, sell a range of rollable panels called Solarwrap. The manufacturer hides the true specifications, most importantly the panel power ratings -- because they are not so good. The Solarwrap Mini, specifically for charging via USB, is very light, with 2200mAh battery, only 110g, however, after much digging I found the solar panel is only about 1 - 1.5W.
Some guys have provided feedback at Amazon.com, one guy tested solar charging of the Solarwrap Mini, and after a week in full-sun, his Nexus 4 battery only got to 70%.

Inspecting the Epak Solaroll

First impression is very high quality construction...

epak unpack
Extremely well made padded nylon carry bag. Inside is the Solaroll itself, a USB-to-microUSB adapter cable, LED with 3m cable, user manual and plastic hanger for the LED lamp. Weight breakdown:

4W Solaroll (with 2600mAh battery)
175g
USB-microUSB cable
14g
LED 1.4W lamp
85g
Case (incl manual and hanger)
131g

The all-up weight is 405g. But, compare the 175g Solaroll with all the folding panels at the top of this page -- the Solaroll is by far the winner. Most of those folding panels are just that, panels, no battery, which makes the Epak even more impressive.

The case is just too lovely, I have to carry the whole package in my backpack!

The downside is that there are mounting grommets at the top of the panel only, unlike the (smaller) folding panels that can be strapped firmly on all sides to a backpack.
This is the big problem. Yeah, I could restrict charging to only when I am camped, but I intend to do a lot of walking and really do want to mount the Epak on the backpack somehow.

The Solaroll can be strapped at the top ok, but the bottom is going to sway around as I walk. This may result in damage to the panel. I may document a solution in a later web page.

Monitoring charging

I purchased a couple of great little items from dx.com:
battery torch

This is a 2600mAh lithium-ion battery, with plug-on 1W LED lamp. It cost US$10.56, postage-free:
http://aud.dx.com/product/cylinder-shaped-external-2600mah-power-battery-charger-w-1w-led-flashlight-for-cell-phone-black-961217813

I bought this for three reasons. First, one of my solar panels (without builtin battery) can be plugged into this. Second, at night I can use it at night for illumination in my tent or room, or thirdly, carry it around as a regular torch. Weight, including lamp: 144g.

The lamp has on/off switch, battery has charging LED.

Note, the micro-USB socket shown above is the input for charging the battery, the normal USB socket is the 5V output.

So, you plug a charger to the battery, but is it actually charging? Well, the status LED in the battery will show red if charging, but what we really want to know is how much current we are getting out of the charger...
charger doctor

This is just what the doctor ordered: the Charger Doctor. Rated up to 7.0V and 3.0A, plug this between panel and battery to see exactly what your panel is putting out!

This lovely little device cost just US$3.98, including postage:
http://aud.dx.com/product/usb-av-usb-power-current-voltage-tester-translucent-blue-silver-961235090

Weight: 16g.
charge testing
When these batteries are new, it is recommended to give them a good charge. Here I have it plugged into a USB port on my laptop, and using the Charger Doctor to see what is going on.

The display flips automatically between showing volts and amps, and it is showing the charging current is about 0.25A.
Well, it has been dropping, now at 0.21A.

The battery wasn't flat when new, but if it was then I would be looking at 11-12 hours to fully charge (from laptop USB port).


Of course, I am assuming that the battery is actually 2600mAh! Quite likely it is a bit less than specified. Refer to the mini-tutorial at the top of this page. It is most likely that the manufacturer is specifying that 2600mAh as for the internal battery, which typically is 3.7V. So, I will calculate the claimed energy storage as 2.6*3.7 = 9.6Wh.

Mini-tutorial:
I have used the nominal voltage of 3.7V in energy calculations on this page, however in actuality the battery voltage is "all over the place".
For "3.7V" Lithium-polymer batteries, the voltage can be between 3.4V and 4.1V, as this tutorial explains:
https://learn.adafruit.com/li-ion-and-lipoly-batteries/voltages
I am hoping that using 3.7V is a reasonable average value for the purpose of energy analysis on this page.

Note also, I can plug the battery into my smartphone to charge it. So, the battery serves two purposes, as a torch and a phone charger.

I also tested the LED lamp. After charging the battery, I plugged in the lamp, turned it on, and left it on. That was 8.07pm, later on I went to bed, woke up at 6.30am and the lamp was still shining, except was flickering on and off. I turned it off. Total running time, 10 hours and 23 minutes.
However, it did seem that the "1W" rating of the lamp is bogus. I tested with the Charger Doctor (before the battery had discharged), and got a current of about 0.05 amperes, which means the lamp is only drawing about 1/4 watt.

Note: that flickering on and off of the lamp is bad. This could be a bad situation if the battery is used to charge a smartphone. A good external battery should cut off its output completely when output voltage drops to a usable minimum.

Hmmm, something fishy here. 0.25*10.38 = 2.6Wh
That is, according to the Charger Doctor, it only extracted 2.6Wh of energy from the battery, only 27% of the claimed capacity.
...see more bad news below...

take2
kcx-017
I need a second opinion, so I bought another charge monitor. This one has a LCD display, so will likely have less power drain than the Charger Doctor (which has LED display), plus it records the accumulated energy, in mAh.

This unit is the KCX-017, cost US$8.55 from here:
http://aud.dx.com/product/lcd-display-usb-power-charger-data-transmit-current-voltage-tester-capacity-tester-white-961315314

Nup, still get 0.06A with the "1W" lamp, about the same as above.

kcx-017
The accumulated mAh is very useful. Right now I am charging the "2600mAh" battery, will find out exactly energy needed to fully charge it:

The bad news:
Using the KCX-017 charge monitor, the Epak lamp and an AC power charger, I have put the "2600mAh" battery through a few charge-discharge cycles.
it takes about 1060mAh to charge the battery, which is 5.1*1.06 = 5.4Wh
I only get about 630mAh out, which is 5*0.63 = 3.15Wh.
That is really awful. The actual capacity is about (3.15/9.6)*100 = 33% of that claimed.

The energy loss is also bad news: (3.15/5.4)*100 = 58% efficiency
Losses would be incurred at three places: switching converter voltage step-down at input, storage loss in battery, switching converter step-up to 5V at output.
I hope other USB external batteries are not this inefficient!

I did expect the Chinese suppliers to over-state the specs, although not by that much. A battery that stores only 1/3 of its rating, a LED lamp that draws 1/4 of its claimed wattage.
Testing continues, more over-stating revealed...

Testing WN-09 5W folding panel

I purchased one of the smaller folding panels from the table at top of this page, the WN-09, rated at 5W, for US$32.36 including postage:
http://aud.dx.com/product/wn-09-5w-1a-fold-up-solar-panel-power-battery-charger-camouflage-961335924
folding 5w

I chose this for testing as its power rating is comparable to the Solaroll.

Right off, the small size and solid construction creates a good impression. Folded, it is only 18 x 17cm, opened it is 18 x 40cm.
The small size does immediately make me suspect that the output will not be anywhere near 5 watts, unless they are spectacularly efficient solar cells -- which is highly unlikely.

I eagerly weighed it, and it is 230g, same as officially claimed. Heavier than the Solaroll, but let's see if its other features tip the balance in its favour...

Charger Doctor showed only about 0.1A output. What!!! So, I applied various resistor loads to the panel, and yes, peak power output, in full sun, is 4.99V at 0.153A, which is only 0.76W.
Note, I used a digital multimeter, not the Charger Doctor, to obtain accurate readings. There has to be a fault somewhere, either in a panel or the electronics -- yes, the little black rectangle that you see in the photo contains an electronic voltage regulator.

Right, I have removed the voltage regulator, testing direct output of the WN-09 panels. Oh so weird, no matter what resistive load I put on the panel, current output stays about 0.14A - 0.23A. My multimeter is OK, this panel is broken.

Testing GP Solar 4.5W quasi-foldable panel

I own a panel that I bought awhile back (from an Australian supplier), rated at 4.5W, made by GP Solar, that cost AU$39 (plus postage), weighs 252g (website claim, on my kitchen scale it is 269g):
http://www.lowenergydevelopments.com.au/5W-5V-Solar-Panel?manufacturer_id=24  
gpsolar 5w

It is solid plastic, and doesn't seem to have any "smart electronics", just straight from panel to smartphone.

it "quasi-folds", but with a big hole in the middle, awkward to pack in my backpack -- though perhaps I could wrap it around the towel.
I refer to this as quasi-foldable, as it is made of stiff plastic and the fold is actually a large curve. I had earlier "retired" this panel, as the way it folds up is not convenient to place in a backpack, plus I felt that it is a bit too heavy. However, after the dismal result of testing the WN-09 "5W" folding panel, I decided to take a fresh look at the GP Solar panel.

A European manufacturer, so more likely to be truthful about the output power, and judging from the surface area of the panel, I would say that the 4.5W rating is reasonably correct. Testing confirms this...

gpsolar plot
By attaching various resistor loads, I obtained enough points to plot a voltage-versus-current graph.

Notice the linear voltage drop as more current is drawn, reaching 4.62V and 0.825A, the maximum power output of 3.81W.

After that, it falls off fast, and see how it curves back -- that would be due to electronics in the panel. I haven't cut open the junction-box on the panel, but there is likely some limiter device in there.

Testing was done in Perth, Western Australia, Saturday 23rd August 2014, about midday, a sunny almost-spring day, with the sun reasonably high in the sky and irradiance (intensity of the sunlight) likely to be close to 1,000W per square metre. Hey, I'm remembering stuff! (I am a retired electronic engineer, worked in the solar field in the late 1970s). Note, I used a digital multimeter to take the readings.

Ha ha, they cheated a little bit! The panel is advertised as "5W", in detailed specs it shows 4.5W, my test shows actual output 3.81W. Looking at the curve, perhaps a tiny bit more power could be coaxed out of the panel, plus a little bit more mid-summer.

The end-user will want to know what this means in terms of charging up a battery. My "2600mAh" battery needs 1050mA (into the USB socket) for 1 hour to charge (see above). The maths is 1050/825 = 1.27 hours approximately. But, that is in full sun, which you are not going to get on your backpack. Unfortunately, the reality is very different, as the battery will only charge at a certain rate...

Firstly, I plugged the GP Solar panel into my cylinder "2600mAh" battery. Then I unplugged it and plugged it into my Sony Xperia SP smartphone. I used the Charger Doctor to measure the voltage and current that the panel was generating...

The bad news:
Tuesday 26 August, 3pm, full-sun. It was cloudy earlier, bright sun now, but of course a bit down in the sky. For each of these loads, this is what I got out of the panel:

Charging "2600mAh" battery:
4.89V  x 0.29A = 1.42W
Charging Xperia smartphone:
4.85V x 0.41A = 1.99W

In other words, I am only getting about half, or less, of what the GP Solar panel is capable of.
The battery and smartphone will only charge at a certain rate, and will not take full advantage of what the panel offers.

Well, it is bad news, but only in the sense that I have not chosen the correct loads for the panel. The panel is capable of 3.81W, so I need a load that wants that much power (or use a lower-power panel).

This brings us to the concept of pass-through charging. The cylindrical "2600mAh" battery is not capable of pass-through charging, but if it was, we could conceivably charge and discharge the battery at the same time. Assuming that the "2600mAh" battery did support pass-through charging, then it could be charging from the panel at the same time that the smartphone is charging from the battery output socket, which would (in theory) draw 0.29+0.41 = 0.70A from the panel.

...ah, this is starting to look good! We need to investigate pass-through charging...

Testing Epak Solaroll, 4W rollable

I can't test the solar panel properly, as the battery is permanently wired to it, however, I can obtain some overall usage impressions.

When I first obtained the Solaroll, I plugged it into a AC mains charger, so as to fully charge the battery.

Now, the battery capacity is a nominal 2600mAh, but I am planning to purchase a smartphone that will have an approximate 3500mAh battery. Therefore, the smartphone cannot be fully charged in one go.
However, I discovered that the Solaroll is capable of charging the smartphone at the same time that it is itself charging from the sun. This is known as "pass-through charging", and I came to learn that very few USB battery charging systems support this.

The good news:
Support for pass-through charging is very good news indeed. It means that while walking with solar panel over the backpack, charging its battery, it can simultaneously be charging my smartphone. What is particularly good about this is that if the sun goes behind a cloud or tree and solar charging stops, my smartphone will continue to charge from the battery in the Solaroll.

This is important! it means that the Solaroll's battery is acting like a buffer, so that charging of the smartphone will not be chopping on and off as the sun disappears and reappears -- such chopping may stress the smartphone circuitry.

Pass-through charging also means that the Solaroll battery does not have to be greater capacity than my smartphone.

epak pass through1
This is demonstrating pass-through charging.

The AC adapter is supplying 0.73A, the lamp is drawing 0.21A. Thus, 0.52A difference is partly losses and part going into the battery.

if I disconnect the AC adapter, the lamp continues to work.

I replaced the lamp with my Xperia SP smartphone, and it charged at 0.32A.
The AC adapter continued to supply about 0.70A.
Thus, about 0.4A is losses and going into the battery.

Curiously, the Solaroll user manual does not actually state that it supports pass-through charging!
It should not be taken as "assumed", as most external batteries do not support it.

OK, so I can't measure the power output of the Solaroll panel directly, at least not without pulling it apart (which I am itching to do). However, I can make a pretty good estimate based on the surface area of the panel:

More good news:
GP Solar: 16.5x11.8x4 = 779cm2
Epak Solaroll: 14.4x59.7 = 860cm2
Oh wow, the Solaroll has a bigger surface area!

Even if the efficiency of the cells in the Solaroll is lower, it seems pretty definite that this panel is going to deliver the claimed 4 watts. It means that Companion, the manufacturers of the Epak Solaroll, have been scrupulously honest.

The Epak Solaroll is looking good. But, apart from the mounting problem that I mentioned above, is there anything else that I don't like about this panel?
Yes, having the battery attached to the panel means that it is in the sun. In Australian summer weather, that could be very bad news for the battery. it could get too hot -- I don't know what that will mean -- reduced capacity, catch-fire, melt? Only extended field testing will expose any problem here.

Oh yes, another thing. The Epak also has a "2600mAh" battery. Again, it is probably 3.7V, so claimed capacity is 3.7*2.6 = 9.6Wh.
I fully charged the battery, then discharged it using the Epak lamp. I got 1376mAh out of it before the lamp died, which is 5.06*1.376 = 6.96Wh. Incidentally, the lamp ran for a bit under six and a half hours.
So, the battery actually has a storage capacity that is (6.96/9.6)*100 = 73% of claimed capacity.

One more thing: the LED lamp draws 5*0.22 = 1.1W, a bit short of the claimed 1.4W. Quite bright, good enough to read under.
The efficiency of LEDs, that is the luminance (light output) per watt, is considerably higher than traditional incandescent globes, and even higher than fluorescent lamps. There is a great deal of variability though, in the efficiency of LEDs -- if I assume a fairly good quality LED, then 1.1W would be approximately equivalent to a 10W incandescent globe. But, the Epak lamp could be better than that -- it seems to have a phosphorescent housing, which may improve the light output considerably. However, a luminance meter would be needed to find out for sure.

Testing efficiency of pass-through:
I have shown above that the Epak Solaroll supports pass-through charging, however there are a number of caveats. One caveat: it is possible that the pass-through mechanism may not work in an efficient manner.

Consider what happens inside the external battery: if being charged from a 5V panel or AC adapter, the voltage has to be stepped down to the battery voltage (3.7-4.2V). For the 5V USB output, a separate step-up is required. This is normally achieved with switching converters.

If you want to see what these voltage-converters look like, Tim's Blog has pulled apart a cheap external battery:
http://cpldcpu.wordpress.com/2014/03/31/external-usb-battery/

So, going through step-down-converter, store-energy-in-battery, then out through step-up-converter, each step has a loss, and I have got only 58% efficiency in the external batteries that I have tested.

The big question is, if I have a load connected at the same time the external battery is being charged, is the energy being drawn by the load going through those three stages. Or, is the circuitry able to divert current direct from input to output, without the 3-stage step-down-battery-step-up?

To illustrate what I mean. Say that the battery is fully charged. The AC adapter is plugged into the micro-USB input of the battery. Let's say the Epak lamp is plugged into the USB output of the external battery:
epak pass
                              efficiency

The AC adapter is plugged into the white KCX-017 monitor, the Epak lamp is plugged into the Charger Doctor on the left.

Input current exceeds output current!

...I started with a fully-charged battery, and ran the above setup for a couple of hours to see any variation. It settled down at 5.15V * 0.36A = 1.85W going in, and 4.42V * 0.24A = 1.06W coming out of the battery.

The bad news:
The Epak Solaroll does support pass-through charging, but not efficiently. Electric current still goes through the step-down then step-up converters (even though the battery does not require charging), with considerable energy loss.
Efficiency is (1.06/1.85)*100 = 57%

In summary, the Solaroll, when charging from the solar panel, with my smartphone simultaneously attached, does have the feature of acting as a buffer, continuing to charge my phone even when the sun fades away -- that part of it is good.
However, it does so in an inefficient way -- regardless of whether the external battery needs charging or not, current flows through the converters and you will loose 43% of the solar panel's energy -- compared with plugging the solar panel directly into the smartphone, you lose 0%.

Further investigation of pass-through efficiency:
What I am really looking for is a pass-through external battery that behaves like a UPS (Uninterruptable Power Supply). This is how the battery in a laptop works -- when the mains is plugged in, and battery fully-charged, current comes in from the AC adapter, and goes through directly to power the laptop -- the battery is bypassed. Current only goes into the battery when it needs charging. The laptop runs from the battery when the AC adapter is unplugged.
These guys are describing the same problem:
http://www.stackprinter.com/export?service=raspberrypi.stackexchange&question=1360   

Back to checking out a cheap panel from dx.com...

Testing M-6261 5W rigid solar panel

I bought this as I wanted to test a panel that is a good match for my "2600mAh" cylindrical battery.
rigid 5w panel

This is also a change from the foldable and rollable panels that I have considered so far.
I bought it for US$25.38 from:
http://aud.dx.com/product/m-6261-5w-usb-solar-powered-moible-charging-panel-power-bank-black-961334441

I like the look of it, seems like it should be easy enough to store in a backpack. Weight is 222g.

The active area of the panel is 19.5x15.5cm, an area of 302cm2. That means I shall expect a power output considerably less than the advertised 5W!
However, the M-6261 is a rigid panel, not a flexible thin film, and it may be that the cells are more efficient...

I am p****d off. This panel is also broken. With no load, I do get 5.0V out, but as soon as I try to draw even the tiniest current, the voltage drops off rapidly. I can't even get a tiny fraction of a watt out of it.

I wonder if customs x-ray machines are damaging these panels? Micro-cracks incurred in transit? Factory rejects? Whatever, from now on I will only buy solar panels from reputable manufacturers/suppliers.

Note, I am not bothering to return these panels to DX, as I would have to pay return-postage, which from Australia is just not worth it. I have written them off.

"Credit card" battery

if the external battery supports "pass-through charging", it does not need to have very large capacity. It just needs to be enough to keep the charging going into my phone when the sun goes away and the panel is no longer giving adequate output.

I want the external battery to be as light as possible, yet still have adequate storage to serve as the proposed "buffer" for charging my smartphone. Adequate capacity is something that I have yet to determine: field testing is required.

Anyway, I found an external battery that is certainly light enough. It was just too cute, the size of a credit card, I had to buy it...
battery mini card

I bought this from DX, for AU$12.45:
http://www.dx.com/p/top-flight-ultra-thin-mini-card-type-1500mah-polymer-mobile-power-source-bank-white-black-314516

It weighs only 45g, incredibly small and light. But, as I have come to expect, that "1500mAh" specification is highly suspect.
Now, a great thing about this battery is that it does support pass-through charging, though it doesn't state it in the documentation.

I have got to give this little guy a name. It is made by Top Flight, and described as "Super Slim Card Power Bank", model TOP-D-001. How about if I just refer to it as my "credit card battery"? -- good enough.

Testing credit-card-battery

Upon receipt of the new battery, I charged it from my AC adapter, it took 594mAh. I then removed the charger and attached the Epak lamp (which draws 0.22A), and it sucked out 760mAh before the light died. Charged it again, and it took 1301mAh. Discharged again with the lamp, and got 770mAh out of the battery.

In other words, 1.3*5.1 = 6.6Wh of energy was needed to charge the battery, and
0.77*5 = 3.85Wh energy was got out of it.
(rough calculations, as charging and discharging voltages varied through the cycle)

The official specs state 1500mAh, well they are probably referring to the internal battery, which I will assume is 3.7V. Thus claimed storage capacity is 1.5*3.7 =5.55Wh

Thus, the actual storage capacity of the battery is (3.85/5.55)*100 = 69% of claimed capacity.
Ah, I see that Top Flight have covered themselves -- there is a "range" spec that states the capacity could be as low as 1000mAh.

This manufacturer is being much more truthful than the manufacturer of my "2600mAh" cylindrical battery, that only stores about 33% of claimed capacity.

The conversion efficiency is (3.85/6.6)*100 = 58%
Same as my "2600mAh" battery. This is quite low. Interesting because I recently read the specs on a Xiaomi 10400mAh external battery, claiming 93% efficiency.

Is 770mAh enough buffering? I don't know.

Just for the record, when charging, the AC charger delivered 0.64A initially, dropping down to 0.51A after 7 minutes, continuing to drop slowly, down to 0.42A after 68 minutes, 0.32A after 131 minutes, 0.19A after 3 hours and 24 minutes. I was asleep when it reached full charge, so don't know exactly when, but the total charging time was probably about 3 hours and 40 minutes.

Also for the record: discharging to my Epak lamp, I got 3 hours and 31 minutes before the light died.
Notice:
I tested pass-through charging, but found anomalies, so need to re-do it.
Will update this page with pass-through results for credit-card battery soon.

Moving on...

Voltaic V15 6W solar charger kit

In hindsight, I can see now that it is going to be difficult to match an external battery that has USB-charging input, to a solar panel. With a solar panel, we want to extract the maximum power that the panel is capable of.

I have read about the Voltaic products in various places, and they are described as being of good quality, with claimed specifications that are reasonably close to reality. This is a USA-based company.

Their V15 4,000mAh (15Wh) battery, coupled with 6W 6V solar panel, is described as working well together. These items are available separately or as a kit. Here is the manufacturers page for the kit:
http://www.voltaicsystems.com/6-watt-kit
voltaic kit

I purchased this in Australia, for $AU119 plus AU$9.95 postage (receipted delivery):
http://www.multipoweredproducts.com.au/products/Voltaic-6W-Solar-Charger-Kit-with-V15-Battery-%252d-V6WK.html

The solar panel weighs 241g (8.5oz) and the V15 battery weighs 126g (4.4oz).

Here is the manufacturer's promotion video for the kit:
https://www.youtube.com/watch?v=2aTBpg6yH24

There is also a video for the V15 battery:
https://www.youtube.com/watch?v=kYONqP0-2Hk

Thanks to the support for pass-through charging, this kind of setup is possible (LED lamp permanently connected to battery):
https://www.youtube.com/watch?v=Nvz50EEN35I

voltaic 6w front
First impression: this is one strong panel!
I saw a video where they hit it with a mallet and it survived -- I can believe it. The panel seems to have two layers of aluminium, between what looks like epoxy. The cells are coated with a UV-treated protective layer.
Weight (on my kitchen scales): 249g (8.79oz)

voltaic 6w back
I want to use carabiners to attach the panel to the top of my backpack, so need to figure out an adapter for the screw lugs on each corner.
There is probably something readily available that would do the job -- a simple plastic thing with two holes -- any ideas?
The panel is very small, no way can it give out 6 watts! Beautifully made though. The V15 battery bank looks well-made also, but, it is the internals that matter. Testing is required...

Testing Voltaic V15 6W kit

Having a solar panel separate from battery is good, as I was able to test the panel independently, by applying resistor loads and measuring voltage and current with my digital multimeter.

Here in Perth, Western Australia, we typically have total blue skies, not a wisp of cloud (or smog) in sight -- this was one of those days. A very high level of irradiance. The test was on September 18, 2014, starting 9.10am, and this panel performed superbly...

voltaic 6w vi plot
The peak power point is about 5.1V at 0.80A, which is 4.08W.

OK, it is not the claimed 6W, but still very impressive for the size of the panel, certainly much more than I was expecting, given that the panel is only 320cm2.



The good news:
High power output
For its size, this panel is impressive, giving over 4 watts. Compare the area of the Voltaic "6W" panel, 320cm2, with the GP Solar and Epak Solaroll panels -- about half the area!
The explanation, basically, is that the thin-film rollable panels have much less efficient cells. Even so, I am still surprised at how much power I got out of this panel.

Mini-tutorial:
Solar cell efficiency
I did some reading-up on the efficiencies of solar cells, and found an explanation for the relatively high power output of the Voltaic panel.
The Voltaic "6W" panel is made with monocrystalline cells rated at 19% efficiency. The Epak Solaroll is made with a-Si amorphous silicon cells, that may have efficiency somewhere between 6% to 12.5%, as explained here:
http://energy.gov/eere/sunshot/amorphous-silicon

On the otherhand, monocrystalline cells have efficiencies from 15% to 21.5%, and also last the longest (25 years):
http://energyinformative.org/best-solar-panel-monocrystalline-polycrystalline-thin-film/

In fact, I can roughly calculate the efficiency of the Epak cells, taking power output of the Epak and Voltaic panels to be about the same: (320/860)*19 = 7.1%

Wow, that really is something to consider when choosing a solar panel to fit on the limited real estate of your backpack!

v15 discharging
Now for the V15 battery bank. I gave it an initial charge, and it took 760mAh to fully charge (from an AC adapter).
I noticed the sophisticated charging algorithm, as the charging current started at 0.73A, dropping to 0.22A, then 0.13A, then 0.07A, then 0.05A -- this tapering off of the charging current is to top off the charge very nicely to the maximum that the battery can take.

Note, that 0.73A is the limit of the AC adapter, possibly an external battery would draw more if it could.

After charging, I discharged with the Epak lamp, as shown.

Note also, new batteries are usually supplied partially-charged, which is why it took only 760mAh to bring it up to fully charged.

After the initial charge, I discharged the battery using the Epak lamp, as shown in the above photo. The battery output 5.30V at 0.20A, and 1980mAh was logged before the lamp died.
So, the battery capacity is 5.30*1.98 = 10.5Wh.

I then took the battery through another complete charge and discharge cycle. The second charge logged 3570mAh and took almost 7 hours, charging from my AC adaptor. The voltage from the AC adaptor was about 4.7V through most of charge, so energy put into the battery was 4.7*3.57 = 16.8Wh.

Now, the second discharge should show what this battery is capable of!
Actually, I did the pass-through test next, as documented below, then, with a full-charged battery, plugged in my Epak lamp, and waited...

Done, the lamp died after drawing 2000mAh. It was running consistently at 5.30V and 0.21A output. So, the energy extracted is 5.30*2.0 = 10.6Wh.
So, efficiency of the battery is (10.6/16.8)*100 = 64% This is ratio of energy got out of battery to that required to charge it. I would say that is accurate to plus and minus 2.

The claimed storage capacity of the V15 battery is 15Wh. 10.6Wh is much less than that, in fact, (10.6/15)*100 = 70.6%.
The claimed storage is also claimed as 4,000mAh. Well, let's state this as being for the internal battery, so claimed energy storage is 3.7*4.0 = 14.8Wh. So, actual storage is (10.6/14.8)*100 = 71.6% of that claimed.

Testing efficiency of pass-through:
The Epak external battery supports pass-through, but with an efficiency of only 57%. In theory, higher-quality voltage-converters could raise this considerably, as could extra electronics to bypass the battery under certain conditions. I am eager to see how the Voltaic V15 performs...

v15 pass thru

The battery started off fully-charged, the AC adaptor plugged into its micro-USB charging socket, the Epak LED lamp plugged into the USB output port of the battery.

There are some errors here, due to the Charger Doctor and KCX-017 not being exactly accurate on their least-significant-digit, and some switching action on the charging input to the battery causing fluctuation in readings.

After one hour, I unplugged the lamp, and then I realised a potential gotcha, as the battery started charging (after unplugging the lamp), and 123mAh went in before charging stopped.
Question: was the battery draining in the above pass-through test? Or, did the charging algorithm decide that a bit more could be pumped into the battery?

To answer this question, I ran the pass-through test overnight. Starting with a fully-charged battery (as per above). Eight hours and twenty minutes, to be exact.
Now, the logic of my calculations starts to get a bit convoluted, as I then had to recharge the battery, and log how much had been drained from it in that 8 hours and 20 minutes, and include that in the calculation (for the record, a top-up of 182mAh was required after completion of the 8-hour 20-minute test). I will just post the end result here...

The
not-so-good
news:

Efficiency of pass-through is 62%, plus or minus 3

Not much better than the other external batteries. The error of about 3, that is, efficiency could be 59% - 65%, is due to my very limited test instrumentation.

Having tested the Voltaic v15 battery, what do I think of it overall? How does it compare with the other batteries? A summary...

Summary of Voltaic V15 power bank:
I had high hopes that this battery would have a capacity near that claimed, and use high-efficiency voltage-converters. However, neither is the case, this battery being a bit better than others but not by much.

The battery stores 10.6Wh, about 71% of that claimed. I got out 2,000mAh at 5.30V, which I estimate would be enough to charge a 2300mAh battery in a smartphone -- but then, claimed storage of smartphone batteries may also be exaggerated also, so maybe a "so called" 3,000mAh battery could be charged -- but that is a wild surmise.

The storage efficiency, that is, how much energy you get out compared with how much goes in, is 64%

Then there is pass-through charging. There I was getting 62% efficiency.

So far, it is "par for the course" that manufacturers, even the reputable ones, overstate battery capacity by about 30%.
Also, fairly low-efficiency voltage converters are used, which means that you have to put in about 55% more energy than you get out.

Another opinion:
I found someone who has posted a review of this kit:
http://waldenlabs.com/voltaic-systems-6-watt-solar-charger-kit/ 

I expect more tests will be added to this page.

Camera

A couple of the photos in this page were taken with my Sony Xperia SP C5303 smartphone. I have a lot of trouble with this camera, as the focus "hunts" in and out when shooting indoors. Usually I take several shots of the same thing, and choose the one with best focus. It is weird, I can be holding the phone still in front of what I want to photograph, and see focus fading in and out repeatedly as it hunts.

For a long time, I thought it was just me doing the wrong thing. Note, I have the same problem with a Canon Coolpix L310 camera, to a smaller extent -- in fact, with that camera, if I drew a picture of a face on a piece of paper and placed it beside the object that I was photographing, the face-recognition kicked in, and I achieved perfect focus every time!

I haven't tried that trick with the Sony. Anyway, I googled and found that I am not alone. It seems to be a "fault" of the Sony S series smartphones. For example:
http://talk.sonymobile.com/t5/Xperia-S-SL-acro-S/Camera-xperia-s-focusing-problems-and-poor-images/td-p/149424

The focusing problem is in low light levels. Well, I take a lot of shots indoors, with fairly bright room lighting, and using the phones inbuilt flash. For a AU$297 phone (purchased early 2013), not very satisfactory.

Anyway, I will be upgrading my phone soon. I am planning to purchase a Samsung Galaxy Note 4 -- I will update this page after acquiring and testing it.

Conclusions

I absolutely love the Voltaic v15 solar panel. This is my choice to take traveling, either in my shoulder bag or backpack. I just have to figure out a means of strapping it on.

Regarding a battery, well, I have not yet found one that handles pass-through efficiently. I know it is possible to get up to 90%, as I can conceive in my mind how it could be done.
Until I find this mythical battery, the Voltaic V15 is fine.

If you don't want to use the pass-through feature to charge your phone (or other device), the storage capacity of the V15 may not be enough for modern smartphones, not to fully-charge them anyway.

Future projects

These guys have really got their act together:
https://learn.adafruit.com/usb-dc-and-solar-lipoly-charger

I gave a cursory examination of the chip that they use in their charger-controller, and it seems it may actually achieve my dream of efficient pass-through.

Adafruit use the Voltaic panels, which is really good, as their charger is designed to work with my "6W" panel and I won't have to buy some other panel.

I have to give the purse a rest for awhile though. Soon, I will probably buy the Adafruit charger and test it. It is DIY, so I will have to make a case for the charger, and therein lies a problem -- taking a homemade kit like that to an airport -- picture it, a little circuit board with a rectangular block plugged into it -- ha ha, the customs officers will freak out!

if indeed the Adafruit charger does look like the "answer", a professional-looking case will have to be created. Anyway, I will cross that bridge when I come to it.

If you want to provide any feedback, send me an email, at STARTbkaulerATgmailDOTcomEND (obfuscated to deter automated spammers, but you can figure out my real email address).
It would help if you put the text "BK" somewhere in the email title, so I can easily recognise it is feedback and not spam.


(c) Copyright Barry Kauler 2014, all rights reserved.
Please do not copy this page anywhere, instead link to it. I will probably be editing it every now and again, so it is wise to link to this original page.