"a controller just adds inefficiency."
"You have to disconnect it at night."

How do you monitor for overvoltage disconnect and undervoltage disconnect? How do you do equalization? You don't need some of these features if you have no battery in the middle.

Charge controllers are very handy.

My point was to study which of the features might be handy for little or no penalty, and roll them in.

--B.G.--

Edited by --B.G.-- on 09/19/2012 02:11:31 MDT.

WOZ: That graph is the key to understanding pv solar panels. Note that as the load attempts to draw more current, the voltage goes down. 7.2V is the point of maximum efficiency for conversion of light to electricity. So, if you hook a 7.2V panel directly to a 6V battery, the panel will simply supply as much current as it can at 6V, (the voltage on a charging battery does not change much). You need a reverse blocking diode, but otherwise this is a very electrically efficient system. A linear regulator can also be used to prevent overcharging/overvoltage but adds nothing to the overall efficiency.

Now, an active regulator is a different beast - it will draw as much power as it needs to supply its load. So, if you want to supply a load of say 5V @ 1A, the regulator will draw 5/0.9 = ~5.5W from the panel. IF you have sufficient area of panel with sufficient sunlight intensity to supply this power, then ok. But in less optimal conditions, as you attempt to draw more current from the panel, the voltage goes down = LESS power. The regulator will drop out, but then the panel voltage will come back and you get in a cycle of oscillations. A (normal) active regulator will not work unless you have oversized panels that can always supply sufficient power.

"That graph is the key to understanding pv solar panels."

As Stuart points out, the graph is the key.

Most single crystal silicon panels have similar curves, although some are slightly shifted. So, you really need to work with the curve for your specific panel product, and you don't want to just blindly assume a curve that somebody else has for a different one.

Personally, I am a big believer in reverse blocking diodes, typically with very low voltage across it.

Linear regulators were good about twenty years ago, but active regulators are in now.

Once you get the charger thing cooking just right, you can play around with setting it to the sun angle above the horizon, tracking the sun from east to west, and all sorts of things to improve total charging performance. A dinky shadow can really foul it up.

--B.G.--

Bob, that's a really good idea that never crossed my mind. I was thinking a single 3mA LED for low drain, but with a 'test' switch, I could put a descent brightness LED on there, or maybe a few SMD indicators to see how much the panel is producing and I no longer have to worry about the loss. That would definitely increase my options without sacrificing any performance (other than a few grams of weight). Plus brighter LEDs could be easier to see in bright light, which by definition is the condition that I'd prefer to use the panel! Thanks! BTW, I am using a 12V spec'd panel, the 7.2V was just an example. It was actually cheaper to buy this Powerfilm 5W unit rather than 5W of 6 or 7 Volt Panels. I'm not sure why that is. All the wiring in the panel unit is inside the stitching so I can't really see how they are wired to guess what spec panels they are and series vs parallel design. BTW, the quality of the build on the Powerfilm is fantastic! High quality wire, Brass rivets (non-corrosive), electrical parts are sealed with silicon and routing and wiring is taped in place to minimize stress. I was very impressed when I took it apart! The little circuit board has two diodes on it and only one is being used, so it must be a generic PCB that they used for their 10W version as well, which is basically 2 of mine in parallel.

Jerry, battery charging is freakishly complex! Luckily the iPhone has alot of built in protection so as long as I stay in the USB 4.75-5.25V spec, I should be ok. You can actually charge batteries at a fairly wide range but it will affect their lifespan and cause excess heat. One of the huge advantages of Lithium Ion batteries, besides their low weight, is the ridiculous # of charge cycles that they can go through. Technically each charge takes a little bit of their capacity away, but poorly designed/spec'd chargers make it much worse. Lead Acid batteries are much more forgiving except that they can heat up and be damaged if overcharged, especially at high (quick) charging voltages. But obviously, the assumption is that if you're charging your car with a solar panel, your battery is dead, and you just want to up the juice, not replace your alternator with a solar panel :)

Jack, think I'm into this one for $100 bucks already :) I'm not sure how well they would sell at that price! But maybe this project will give me some insight into building another one much cheaper. NOW TAKING ORDERS!!! (MUST PAY UP FRONT!!!)

Two problems with solar panel into lead acid batteries:

You want to keep the battery fully charged. If you don't have enough panel then the battery will be undercharged and over time it can kill the battery, like if the battery goes below 50% charged or so.

If you over-charge the battery, the water in the electrolyte get's conveted to hydrogen and oxygen. If it goes low enough and exposes the plates inside the battery, that will kill the battery.

If you have solar panels on your house charging batteries, you need good controller to prevent these problems. You won't remember to do it manually reliably enough.

If you're vacationing and hooking up the panel each day, you can remember to check and prevent. And disconnect at night to prevent reverse current. The controller and blocking diodes consume some of the power.

Stuart, that's a really good point about the oscillation. I didn't think of that. So I'm not sure how best to reduce that on my boost regulator. e.g. So let's say that it's lower than ideal light and I'm charging my device with my panel, and it's pulling .5A so it's folding back the voltage on the panel. I can put a voltage divider on the enable pin of the regulator, to disable it at any given level, but when the reg shuts off, it'll obviously stop drawing current, and the panel voltage will bounce back up. repeat ad infinitum. I don't think I need the diode since the regulator will shut down. I guess I may simply not care if my phone doesn't complain. The output cap on the regulator may be enough to keep the oscillating output voltage of the reg ever getting high enough to even trigger charge mode, or the iPhone may be smart enough to wait a small amount of time seeing a 'good' charge voltage being charging. Either way, it's def a concern...test needed...come on UPS, I need my parts!

I saw this alternative that I thought was promising for lightweight 'general use'. ($20, you assemble)

http://www.ladyada.net/make/mintyboost/index.html


It uses a buck regulator (step-up) to charge USB devices with 2 AA batteries. A rechargable AA battery can have up to about 2500mAH of charge! But they say this won't quite bring an iPhone up to full power, and will only last two partial charges on two new AA batts.

But then again each AA battery is about 1oz, so I'm guessing this is a 3oz charger, and you only get two partial charges out of it. For another oz. I'll take my infinite number of charges under clear skies :)

"Jack, think I'm into this one for $100 bucks already :) I'm not sure how well they would sell at that price!"

People will pay Big Bucks to save grams. There might be one or two folk eyeing a summer-long hike and planning to take an iPhone.

"But maybe this project will give me some insight into building another one much cheaper."

You'd need to buy the bits wholesale and in quantity to drop prices. Hie thee to Shenzhen city!

"NOW TAKING ORDERS!!! (MUST PAY UP FRONT!!!)"

That's the spirit.

Keep us posted on development.

>David, "can the iPhone survive 12 or more volts -- which is what a solar panel can put out -- on its 5V input line?"

I'm not proposing using an 80-watt, 12-volt panel. But rather, an array of cells that puts out somewhat more than 5 volts (open circuit) at something less than 1 amp at 5 volts. The array will have a performance curve in full sun. The batteries will have a very differently shaped charging curve (almost nothing until near 5 volts, then steep as 5 volts is approached. Where those curves (amps versus volts) intersect is where the combined array and battery will charge. Until the battery nears capacity. THEN, yes, bad things could happen if the phone has no protective circuits and is left connected.

It would be akin to filling a gasoline tank with no auto-shut-off on the hose. How much empty tank? Filling rate? Don't leave it unattended for longer than that. Filling the hot tub, refilling propane cylinders, or drinking beer all take a certainly mindfullness so as not to overdo it. The OP seemed to understand circuits well enough to plot those graphs and select an appropriately small array of cells.

Going with cells only, no circuitry; reduces weight AND increases efficiency. It does place the burden of oversight on the user. You definitely couldn't "set it and forget it".

Touché on the beers David. It's easy to know when you've had enough...you run out of beer! Seriously good points!!

How were you planning on waterproofing your electronics? Not having any other practical solution in my head, I've built mine into a heavy Otterbox:
http://www.backpackinglight.com/cgi-bin/backpackinglight/forums/thread_display.html?forum_thread_id=64564
If you've thought of a lighter solution, I'd love to see it...

Also, I don't know if you considered building it with just the WeatherPro panels without the canvas backing; you might save some money and total weight.... I haven't posted my update to the thread above, but the bare panels held up fine on an overnighter a few weeks ago.

As a continuation to the comments about battery chargers, I strongly recommend these:

New Trent iTorch 5200 mAh $33
http://www.amazon.com/New-Trent-IMP52D-Thunderbolt-Blackberry/dp/B0013G8PTS/ref=sr_1_4?ie=UTF8&qid=1348246805&sr=8-4&keywords=smartphone+battery+charger

Anker Astro2 8400 mAh $50
http://www.amazon.com/Upgraded-Version-External-Flashlight-Smartphones/dp/B0067UPRQ4/ref=sr_1_6?ie=UTF8&qid=1348246805&sr=8-6&keywords=smartphone+battery+charger

These are both brands that have been in the business for at least a few years, and seem very high quality to me. Plus they have LED flashlights (the iTorch even has a laser) built-in!

My wife keeps the iTorch, but the Astro2 that I keep on me is 187 g, 6.6 oz... a bit heavy but enough to charge a smartphone several times over... I use it for travel

Matt,

The Panels actually are the weather proof already, so I'm good there! Which is why I went with the canvas backed type instead of the individual bare panels. This was MUCH cheaper that buying them individually.

So I have a couple of issues with weatherproofing my controller electronics. First, obviously the open USB port can't be waterproofed. It has to be open to the world to accept the USB cable/Ipod cord. Second, there is a small amount generated by the inefficiencies inherent in the design and I don't want to insulate it too much so that is overheats. I'm requiring one amp from this guy, so I will be generating a bit of junction heat! The circuit does have a thermal shutoff, but that kind off defeats the purpose! Especially because I'll be setting this in the hot sun. So I'm looking at two options. The easiest, is plasti-dip spray. It's easy to apply, won't damage anything and can be scraped of easily enough to make changes. Of course that offers no mechanical protection. My other option is a small cover which I'd sew directly to the fabric on back of the 'last' panel. I'd use something akin to half of a square dental floss case. This would provide the mechanical protection and could be combined with the plastidip spray or even some liquidtape brush on if the heat buildup turns out not to be an issue! That being said, it's only a one sided board, so I'd only need to seal one side, which helps considerably for heat.

So here's a pic of the actual board and a sketch of the modifications I plan to make to it. The LED is switched, and therefore can be a brighter 8mA light. The USB D- contact can also be switched from .5mA to 1A charging mode. Note that this charger will be mainly for Apple Devices since Apple requires the strange data voltages, so I may wire straight into the iPhone cable to further minimize weight and separable parts. I may also put a mini connector on the solar panel wires to the board, so I can swap this out for a 12V or other charge controller on the fly. I intend to use surface mount parts glued to the very small 'spare areas' of the board accessing the traces as needed. And I have no idea where to put the switches, which may have to go on the inside of the cover!

> The voltage out of a solar panel (before the regulator) will decrease as more current is pulled.

Correct. But I don't think many people really understand what happens with a solar panel. All the discussion has been about volts and amps, but that is the wrong way to look at the whole system. Let me explain.

What really matters with a solar panel is the incoming POWER from the sun. The solar panel can only deliver as much power as it gets (allowing for inefficiencies of conversion). What the output voltage is does not really matter, nor is it fixed. The panel voltage will adapt to the situation.

This means that you can stick a solar panel rated at 12 V onto a 7 V battery and it will happily deliver power to the battery at around 7 V and charge it up. Of course, if you try to use a 3 V solar cell you won't get very far. If you try to use a 100 V solar cell, bad things might happen as well.

The next thing to consider is how the rechargable Lithium battery behaves. Most rechargable Lithium batteries have internal circuitry to limit the input voltage. Very often that circuit also limits output current in the event of a short circuit. That is a safety feature they are required to have, to prevent fires. When the battery is fully charged the internal cell voltage rises of course; at some point the little safety circuit inside says 'enough' and shuts off. It won't accept any more charge. At that point the apparent external terminal voltage can rise several volts.

So my solar charging system sticks the output of the 12 V solar cell straight onto the 7.2 V recharable lithium battery, and charges it. When the battery is fully charged the safety circuit switches off and the battery stops taking charge. Ah, but that means the solar cell is no longer delivering power, so its terminal voltage rises up towards 13 V. Yes, the safety circuit in Lithium battery can handle that. The high voltage is then enough to push current through a series combination of white LED (3.6 V) plus 10 V zener diode. So when I see the white LED shining I know the battery is fully charged, and I can take it out.

Yes, this does mean that I recharge the 7.2 V lithium battery OUT of whatever it goes in. It may well be that the phone/whatever cannot take 13 V on the input. That's OK, I don't stress it. I carry two small lithium batteries and recharge one externally while the other one drives the phone/whatever.

Of course, this also works on a PAIR of 3 V rechargables for a UV Steripen. Been recharging them that way for years.

Cheers

Of course, here we are talking about charging an iPhone which does not have removable batteries and we don't know how much voltage it can take at its USB input charge line without breaking. The designers don't have much reason to design for voltages higher than the USB standard of 5.25V. Alas, Apple doesn't give out any useful specs for things like maximum input voltage.

Wow, that's a pretty clean looking board, you have some serious skills. I don't envy one who needs to glue surface mount components to the spare areas and "access traces as needed"!

If it was me, I'd just access the pin outs on the side (assuming that is what's going on). Do you have a bench to validate functionality, or will you just field test it cold turkey? Fun project indeed!

> an iPhone which does not have removable batteries
Oh well, defective design.
And now you have Apple maps...

Cheers

"[...] an iPhone which does not have removable batteries [...]"

"Oh well, defective design.
And now you have Apple maps..."

Oh, it gets better: the CDMA (non-GSM) version used by Verizon, a major carrier in the US, does not permit shutting off the phone's radio without also shutting off the GPS receiver. So if you're in the backcountry, away from any cell towers, the phone goes into high power "search" mode, rapidly draining the battery. Unless you put it into Airplane Mode, which then kills the GPS receiver.

Why? It is a mystery.

(Not trying to create thread drift here, just sayin')

Why is it that, despite being highly opinionated and blunt, Roger and I don't get into arguments? I suspect we're "reading from the same book(s)" and they are mostly engineering references.

Roger: I agree with your logic and practice of using a 12-volt array to charge a 7.2 volt lithium battery. BUT, I think you could save weight with little loss of charging rate if you had a 9-10 volt array. Alas, that's not an off-the-shelf item, but if you wire individual solar cells in series, than you can choose a lower max voltage, save weight, and gain a bit of efficiency.

Question: Do even AA lithium batteries have the internal circuits you describe? Disposables and/or rechargeables?

Anecdote: I had a ChemEng friend from high school Math Club days who went to work at a lithium battery company back in the early days. The fastest way to empty a room was to yell "hot cell" and then everyone would belly-crawl out to the fire escapes. These were auto-battery-sized lithiums and held A LOT of energy. Their joke, since they were going into cruise missles, was, if it doesn't work, it blows up. If it works, it blows up.