My apologies to any of the 1400 or so people who've read this Instructable in the day since it was published; I described an idea I was sure would work, and I was so excited to get it out there where people could try it, that I didn't wait until I could get the parts I needed to fully test it myself, and I posted based on a partial test and an assumption about the performance of small motors that turned out not to be accurate.
Now normally I would take a faulty Instructable down (and run away and hide), but in this one instance I'm going to leave this online after editing it to correct my mistakes, because I thought that there was still a good idea here that could be developed even if it had to be by someone with more electronics experience than me, and I wanted to put this germ of an idea out there so that maybe it could inspire someone to come up with a solution that works. I offered the 1Yr pro membership upgrade I received when this went front-page as a prize, and as you can see from the comments a lively discussion ensued. (I gave the prize to jtlowe for his suggestion of using a clockwork rotating platform, but although there's no more prizes to give, I would still welcome any more suggestions you all can offer).
The problem is sun tracking: pointing a solar panel directly at the sun so that it can harvest significantly more light - and the difficult aspect of the problem is that the cost of adding a sun tracker to a solar panel in order to gain X% extra output has to be less than X% extra cost, otherwise it's more cost effective to simply add more solar panels.
The solution that I had was to take two small PV cells from a couple of solar garden lights, and connect them not in series or even in parallel, but head to head, connecting the ground lead from one to the ground lead from the other, and determining which of the two panels was receiving more light by looking at which one was able to generate more voltage than the other. For instance if one generated 2V and the other generated 3V, then the voltage between the two positive outputs would be 1 volt, and that volt would be used to drive a DC motor to turn the platform in the direction of the cell that was reading the stronger light signal. (Although in practice the voltage was actually less because driving current through a PV cell against its natural direction (since it acts a little like a diode) causes a voltage drop in excess of just the voltage that would be cancelled out, though that's not relevant to the problem)
Although that is indeed what happens which you can confirm by looking at the voltage on a voltmeter, what I didn't realize was that these panels don't generate enough amperage to drive even the smallest motor, as far as I can tell. I'm talking about 3V motors that need so little power that they'll spin from a single half-dead AA battery!
So what I'm going to show you here is half a solution, and I'm asking the smart readers of Instructables to help come up with the other half.
Now normally I would take a faulty Instructable down (and run away and hide), but in this one instance I'm going to leave this online after editing it to correct my mistakes, because I thought that there was still a good idea here that could be developed even if it had to be by someone with more electronics experience than me, and I wanted to put this germ of an idea out there so that maybe it could inspire someone to come up with a solution that works. I offered the 1Yr pro membership upgrade I received when this went front-page as a prize, and as you can see from the comments a lively discussion ensued. (I gave the prize to jtlowe for his suggestion of using a clockwork rotating platform, but although there's no more prizes to give, I would still welcome any more suggestions you all can offer).
The problem is sun tracking: pointing a solar panel directly at the sun so that it can harvest significantly more light - and the difficult aspect of the problem is that the cost of adding a sun tracker to a solar panel in order to gain X% extra output has to be less than X% extra cost, otherwise it's more cost effective to simply add more solar panels.
The solution that I had was to take two small PV cells from a couple of solar garden lights, and connect them not in series or even in parallel, but head to head, connecting the ground lead from one to the ground lead from the other, and determining which of the two panels was receiving more light by looking at which one was able to generate more voltage than the other. For instance if one generated 2V and the other generated 3V, then the voltage between the two positive outputs would be 1 volt, and that volt would be used to drive a DC motor to turn the platform in the direction of the cell that was reading the stronger light signal. (Although in practice the voltage was actually less because driving current through a PV cell against its natural direction (since it acts a little like a diode) causes a voltage drop in excess of just the voltage that would be cancelled out, though that's not relevant to the problem)
Although that is indeed what happens which you can confirm by looking at the voltage on a voltmeter, what I didn't realize was that these panels don't generate enough amperage to drive even the smallest motor, as far as I can tell. I'm talking about 3V motors that need so little power that they'll spin from a single half-dead AA battery!
So what I'm going to show you here is half a solution, and I'm asking the smart readers of Instructables to help come up with the other half.
Remove these ads by
Signing UpStep 1: What you'll need
Grab a couple of those cheap garden path solar lights, the kind they sell at the Dollar Store for... well, a dollar. (If you miss the occasional Dollar Store deal on those, damn, you'll probably be out of pocket by two dollars from a more expensive store! :-) )
That's it - to test the circuit you won't need any more components. Just your trusty volt meter, and a soldering iron... (I finally treated myself to a programmable temperature controlled iron from Radio Shack and I have to say I'm really enjoying using it compared to what I had before. The volt meter is that cheap one that's on sale for a couple of dollars every weekend at Harbor Freight)
When I was trying to drive the motor and realised I didn't have enough power I added a second pair of PV cells in parallel, so some of the images below have 4 cells and some have 2. As long as we're just using these as light sensors, 2 will do fine - I didn't feel it was necessary to reshoot the photos...
That's it - to test the circuit you won't need any more components. Just your trusty volt meter, and a soldering iron... (I finally treated myself to a programmable temperature controlled iron from Radio Shack and I have to say I'm really enjoying using it compared to what I had before. The volt meter is that cheap one that's on sale for a couple of dollars every weekend at Harbor Freight)
When I was trying to drive the motor and realised I didn't have enough power I added a second pair of PV cells in parallel, so some of the images below have 4 cells and some have 2. As long as we're just using these as light sensors, 2 will do fine - I didn't feel it was necessary to reshoot the photos...
Visit Our Store »
Go Pro Today »
This appears to be a fairly well documented technique (even here in Instructables) but it was new to me.
I've also got a transistor version, not sure if I posted it on my website though.
It's a nice simple circuit, but are you powering it from the panels that you're shining the light on or from that DC power supply I can see off to one side? :-)
I may build this for the fun of it, but the biggest thing I still want to implement is a cheap way to power the rotation (or control it if unpowered) using no more than the small PV cells. i think your design may need more power than I have spare.
Thanks for the link, your tracker works nicely.
I have a one-year free pro membership available which I'll give to the person who comes up with the most elegant and cheapest way to use this circuit in an actual tracker!
I was hoping someone would actually build something and try out their ideas, and if they had, that would have made the selection of a winner much easier!
However many good ideas were suggested and chosing a winner was tough - for example I thought Wroger-Wroger had a good grasp of the big picture; alarrrd and jtlowe made low-tech analog suggestions in the sprit of the instructable; perfo, eecharlie, and others contributed valuable detailed information on electronics that would definitely help in a scaled up version of this project. Every commenter had something helpful to say - there really were no bad ideas here.
In the end I decided to give the prize to jtlowe for his suggestion of using a wind-up clockwork mechanism, which was at the level of complexity that I was looking for and is what I'm going to work on next to improve the design.
Congratulations JT! I'll email you the code.
However, if you are using solar lights you have a readily available power source, the batteries.
And another thing is that all the solar lights I've seen have another sensor to turn off the lights during the day so that the batteries will charge more. I don't know if these are phototransistors or LDSs but I would think this would be a better source of your 'light' differential signal. So what I would try would be to run the differential signal into a transistor for gain. The transistor is powered by the battery or batteries and then you would have enough power to drive the motor. Since the motor isn't going to be moving that much, you should theoretically have your efficiency. Now you may have to connect two in series to get enough voltage for the motor.
Now, I'm not an analog guy or a motor guy but I think this would have a better chance of working.
LOG
I'm wondering if the pulse of power from the coil in the Joule Thief circuit is enough to turn the motor a little - it would be more like a stepper motor if it worked, as long as it didn't drive it too far past the optimum position.
Here's my suggestion. Take out the output LEDs from the solar lights, Leave the rechargeable batteries in. They will charge up to maybe 1.4V. Get a low voltage comparator. I know some can be powered by less than 1V. Power the comparator with one of the batteries. Now take the PV output from the two solar lights. This voltage is going to be higher than the battery so divide it down with a voltage dividers. Tie the two voltages to the comparator. Add hysteresis so the motor will stop when you get close. Now you know that the direction of the sun goes in only one direction so the motor needs to move in only one direction. Drive the output of the comparator to the motor. You might need a transistor to give it enough current but the battery should have any capacity to turn the motor especially since the duty cycle for the motor will be so low.
Now if this isn't enough power, you could tie two solar lights in series so the battery voltages would be about 2.8V.
By the way a cheaper solar light is the $1 solar light keychains sold on ebay. They actually put out 3 volts.
LOG
maybe what I said didn't make sense out of context so let me explain what I was thinking: the motors need a higher current to get started turning than they do to keep turning. I know that the power output from the PV cells isn't enough to start the motors turning, but I was hypothesizing that if you collected the output and sent it all at once in a short pulse, it *might* be enough to overcome that initial inertia and turn the motor a little. I was also guessing it wouldn't turn far and wouldn't have enough duration to keep it turning with the risk of overshooting the target. Again - all assumptions, but ones I plan to test on Saturday morning in the heat of the day :-)
Now ideally I would suggest using a capacitor to build up charge, but since we have the JT circuit already - came with the $1 unit that the PV cells came in - we might as well reuse that if it would work. From what I read in another instructable, it looks like JT uses the collapsing field of a coil to generate a large pulse - at a voltage sufficient to charge a 1.2v battery, so it is plausible it would be enough to kick over a gearmotor since a AAA cell alone can do it even when the AAA cell is almost depleted.
I'll give it a try and report back. Chances are it won't work, but it's definitely worth the effort to try.
But instead of adding a capacitor, you could probably add the battery as it's already part of the $1 cost anyway. The battery is going to act just like a capacitor.
By the way, we only have a Dollar Tree out here but I've never seen a solar light in there. Maybe I'm looking in the wrong place.
Good Luck.
LOG
While this is an interesting theoretical project, I question the practical application. As you said it's not very practical outdoors.
The problems I see indoors are first, you probably need an unobstructed south facing window. Even with that I don't think there would be much tracking involved.
I have a south facing window but when the sun is out, I always have my blinds down and there ain't no room for a moving panel.
I would think a slightly larger panel would cover any charging panel.
Other questions, how much is solar charging a cellphone going to save over plugging it in.
By the way, I charge my solar keychains under a lamp. Well, these solar lights charge under room lighting? Guess I've never tried it.
I guess if you're going 'Green' I would say a bigger solar panel is more efficient as you are not wasting energy moving the panels.
A question, however. Since a solar panel is taking energy from the sun, will it be cooler? In theory, I would think that since some of the energy is being converted to electricity, there would be less heat.
LOG
Here are a couple of the tricks.
Your PV cells should be polycrystalline with a resistor load, so that they work as solar power sensors not solar energy sensors. A mono crystalline cells needs full light cover to generate a poly- doesn't. Measure the voltage across the PV cell to measure exposure to the sun.
This will work for adjusting alignment but you need to turn back east when it gets dark (night time).
To save power only turn east-west.
Use a polar mount to get your panels roughly parallel to the earth's axis. Dont bother adjusting the north- south alignment, there isn't enough to gain.
I hope this helps
It's pretty simple to prove, Get one easy to handle solar panel, and hook up to a digital multimeter, and then log the voltages with pencil to paper and accurate estimates of angular alignment to sun in say 5* Degree increments.
I have found that as long as solar panels are within about 15* of true alignment to the sun there is very little difference between being pointed directly at the sun AND being 15* out of alignment with the sun.
It was after that, that the power gained started to drop off..
In regards to accurate tracking and the North South Alignment - this I would say is NOT an issue at or near the equator, as setting the tracking to the North South alignment, at the equinox or between the summer and winter day light lenghts, would keep the sun within the 15* of alignment.
But when ones location is in the lower or higher latitudes, for instance, the sun is at 29* above the horizon at the winter solstice, and it's 75* above the horizon at the summer soltace.
That is a difference of 36*, so in saying if the panels are pointed at the sun, within 15* of true perependicular alignment to the sun, and more especially so in winter, where the say light is weaker, the days are far shorter etc...
At the worst, having the panels set for the summer solstace alignment, would cause a dramatic loss of harvesting capacity in winter, and vis vera for alignment with the winter solstace.
If one were to set the panels at the suns noon equinox angle of 56* (where I live), there would be an appreciable loss at the solstace periods, because the panels will be more than 15* off set to the sun in winter, which is the most important time to have reasonably accurate alignment and also with the summer sun.
So my take is, the lower or higher your location latitude wise, is to get up and clean the panels and service the equipment, on at least a monthly basis, and to manually set the angles to point the panels at the more or less noon sun, for that part of the year - especially in winter.
Cleaning the panels, especially in dusty locations, makes a HUGE difference to the amount of gain as well.
Get out in the early morning or late afternoon sun, and also the winter sun and then align a panel with the sun and measure the voltage drop, the more the panel is out of perpendicular alignment with the sun.
I do know that absolutely perfect tracking systems, that align with the sun on a minute by minute basis, are a waste of time....
I do believe that a simple stepped tracking of turning the panels, to lead the sun by 5 degrees, and then when the panels lag the sun by 5 degrees, to make another turn of the panels, is about optimum.
But keeping the panels adjusted for north south alignment, this issue is basically irrelevant at or near the equator, but it becomes progressively more and more important, especially seasonally, and the further you are away from the equator and are closer you are to the poles.
Actually, if you put your mind to the problem and get away from the solution, you will come to conclude that the traverse we make around the sun is quite regular.
All you need to "track" the sun for a fixed solar array is a timer and some small adjustment as the seasons change.
Now, if you want some sort of "robotic" style sun tracker, then you'll probably need more than $2 worth of parts.
Keep trying.
Even this author clearly states that he posted a project that he hadn't really even built ("...I described an idea I was sure would work, and I was so excited to get it out there where people could try it, that I didn't wait until I could get the parts I needed to fully test it myself, and I posted based on a partial test..."), but he still won the prize. Projects here are a lot about winning prizes. You just have to sort out the junk.
I applaud this author for admitting his mistake and posting the retraction.
I've been vilified on here at times, because I can't always find a way to "be nice" with some of these projects. Criticism is not being un-nice, but can be quite constructive. I got flamed for pointing out that one project was based on a stolen shopping cart. Many argued that the store that owned it was OK with it being converted to something else for the "author." Why present projects that teach the wrong thing?
You take criticism well and must be a good instructor to those around you. Thanks for that!
The basic rule of thumb I'm exploring here is whether it's cheaper to add tracking or to add more solar panels. For huge panels it's a no-brainer. For medium panels it's borderline. For small ones like cell phone chargers, I'll be interested if you can show me *any* existing tracker designs that give you a bigger win than just adding another panel for the cost. Honestly the only cheap efficiency improvements I can think of that are ways of concentrating the sunlight by mirrors or lenses but those are bulky and fine for outdoors but not so practical for say behind an office window.
.
But in winter the sun is low in the sky and (here it is cloudy) so it doesn't much matter where the panel points. It will perform badly no matter what!
There is a helpful site called sollumis that shows where the light comes from all day. If you track well, you must beat a stationary panel by a long ways.
I heard an interview (I thiink with the guy from red rock). He claims that the statistics for cloudy days and sunny days are just averaged over wide areas and you local climate is far more important than you would think. If you are in a sunny area, work at tracking. If you are cloudy, it might not be worth it.
Brian
One thing that's cool about an active follower rather than a fixed path tracker is that it works fairly well indoors for pointing at the indirect illumination through your window. And it's cool for demos where you can make it follow a flashlight :-)
Winding may be driven by a wind operated device.
If it is turning a gear against a pring force, then a reset switch can be activted onlce it reaches a set position and return the mechanism to its start postion waiting for another switch to engage the start of the days tracking activity.
Perhaps this too is too simple an idea as clock mechanisms do not move that large of a mass.
This is the entirely wrong approach.
In life you get what you pay for.
In terms of the TIME and RESOURCES used in buying the parts, transporting the parts, doing the design and assembling and then making it all work, then all the adjusting and repairs and problems etc..
Then you get all the time spent in getting up and down ladders, with the risks of falls, injuries and disabilities or death.
I think it's better to do it properly in the first place, and to make the unit so it works without flaw for the next 50 years, or pay someone who has done all of the research and development, and has amortised the costs of this, across thousands of units.
I want strategic long term investments, of my time and finances, and limiting the consideration to "coming in to under $5" leaves out all of the other really important issues.
Sure I don't want to pay $900 for a controller, that has $25 worth of parts in it and I can build it in 3 - 5 hours myself.... but I want really good outcomes for whatever I do.
As for your adage "in life you get what you pay for" if only that where true. You have to work like a (insert something that is renowned for working a lot in here) just to get something anywhere near what you pay for as most manufactures sell for what they can get not what it's worth which can be the same thing but can also be very different.
Truth is the best strategy is do nothing and look at what's available in a few years time when efficiency is way up and cost is way down. But meanwhile let us have our fun, after all that's what instructables is all about.
respectfully,,
Graham
- Ed
I'm interested in the subset problem of small systems, around the size of a cell phone charger that could be put on a desk by a window (or maybe hanging from the ceiling as one poster suggested :-) )
The basic challenge is can it be done for less than the cost of using a larger panel (or adding a second panel)? While remembering that time costs money and even free parts salvaged from trash are not free if someone else building this can't find those parts...
I reckon seven or eight dollars and 20 minutes of effort is a fair goal to acheive, versus say and extra $20 for a larger panel.
Some good thinking here, however…
I don’t believe that this can be accomplished for $20.00, or anywhere near there .In all probability the electronics alone would exceed that amount. ”Two” Photo Voltaic cells (PVc) could start and/or stop two motors but not reverse direction of either motor, which is necessary for proper tracking. Also you must consider what happens when the sun sets for the evening. If the PVcs are set to “go towards the light” will your tracker be hunting a sun that it won’t see until morning?
For proper tracking of a Solar Array (SA) the array should not only track east to west (for following the sun’s position during the day) but also north to south (for following the sun’s position from season to season). For this scenario four PVcs are needed. One PVc to track west, one PVc to track east, one PVc to track north and one PVc to track south. In addition to the four PVcs one addition PVc would be required to 1) automatically turn on/off the entire array and 2) to return the array to a “home” (Easterly facing) position after the sunset.
There are many, many more things that must be considered for a successful SA tracker. Perhaps you could start a group here and turn this into a true open source project.
Some really good thinking here, keep up the good thoughts,
williamj
1) let it die facing east and rely on the morning sun to turn it back west
2) as the sun sets, use all remaining power in the battery to return to west - this can either be by reversing the engine or by continuing to rotate through to 260deg
3) mechanical reset (eg rubber band, which makes the initial cost of rotating higher) or manual reset
if this is an indoor desktop toy, which is the way I'm inclined to go now, a manual reset and manual powering (eg a clockwork table such as a microwave turntable) tweaked to the right speed is probably my favorite bet at the moment
I would assume that because you are using this indoors, the ambient temperatures that this would be exposed to is at least be kept constant? Ideally it absorbs and releases heat purely radiatively... Being thermally conductive to random temperature fluctuations in the air would render a thermal expansion piston design quite useless!
I can't seem to embed an image in these comments, but I was going to show you the temperature graph for yesterday - almost perfect sine curve except for a small notch when we had a short rainfall. Usually it is a perfect sine wave for the whole day, with only a few days where there is a precipitous drop as a cold front passes over. Few and far between.
I used to have a 'nodding duck' toy that bobbed up and down all day which relied on an expanding liquid as well as evaporation. Of course it did need a cup of water to drive it but I could live with that.
I also see that you are also contemplating mechanical means of orienting the solar panel and to return it to the same position every morning... I was thinking that a passive mechanical system was the best way too, since sunlight and mechanical items are in great abundance, while solar panels and the electricity they produce are relatively scarce
Either that, or balance the weight of the panels with a counter-weight to minimize torque and use the smoothest bearings you can find to minimize friction. At that point, the weight of the whole thing will move it west until noon and then your motors move it until sunset and reset it. The circuit really would be as simple as two poly arrays, an opamp and possibly an amplifier circuit.
At the end of the day, the weight should swing the whole thing a little east of noon position, and a hinged weight could bring it all the way back east. the weight is just latched back in when daylight comes, and the whole process starts over.
"Comments about RadioShack 1.5-3VDC Metal Gear Motor:
I bought this to make a prototype solar-powered boat. It works great. However, it requires at least 0.25amps to spin it with no-load. It goes up to about 1.5amps full load. A single solar panel was incapable of spinning it no-load (I bought the PV from the Shack as well). Just be warned, you'll need about 5 6v/50ma PV's just to spin this thing at no load! Batteries, though, work great!"
Maybe I just need more cells or a different motor? ... :-)
Your set up is similar.
I think you need better shading between the detectors. There is too much incident light.
Maybe isolating each detector in a deepish can, like Pingils or tennis balls are packaged, and painting the insides flat black would give a bigger difference between the readings. A circuit could rotate the assembly to maintain equal readings.
One thing I have given a little thought to but don't know if it is even feasible is using the wind and a ratchet to always be trying to move the platform one more increment. The ratchet ensures it turns in one direction only. It could be machined out of wood, doesn't have to be high tech.
We have a fairly low level of wind here but it's seldom completely absent for any length of time.
Good Luck.
If it's for something you want to run all day every day, such as say a solar-powered wifi hub or maybe a wireless surveillance camera, then manually wound clockwork isn't an option but some way of scavenging occasional wind bursts or rain power might be.
Another clockwork-related techmology worth thinking about is the good old-fashioned pendulum...
For something that you are going to run all day every day I'm sure you could hook up a wind turbine of some sort, get it to wind a spring to store energy and have the spring wind/power the clockwork. The only problem I see is if you need to get the power of the turntable, a trailing cable would just get wound up, but if the power was used at the source ie on the turntabl then that not a problem.
1) let it die facing east and rely on the morning sun to turn it back west
2) as the sun sets, use all remaining power in the battery to return to west - this can either be by reversing the engine (could be a mechanical switch) or by continuing to rotate through to 360deg
3) mechanical reset (eg rubber band, which makes the initial cost of rotating higher) or manual reset (rewind the clockwork)
If this ends up as basically an indoor desktop toy, which is the way I think I see it heading, a manual reset and manual powering (eg a clockwork table such as a microwave turntable) tweaked to the right speed is probably my favorite bet at the moment
But a cheap and simple way of implementing your suggestion is just to flip the PV cell over inside the plastic cover, so it faces inwards instead. Then point the covers with the base out instead of the flat side.
Now if only there was a way to do sun tracking without moving parts...
Hmmmm.....
There could be another way extremely similar to yours but using light dependent resistors (LDR) something like an ORP12 (less than £1 in the uk) . You connect two LDRs in a Wheatstone bridge arrangement with a couple of resistors as the other two legs. The little motors you have will sit across the mid points of the bridge and will drive forward or backward depending on which LDR is conducting more. The LDRs will be sinking current but are capable of handling enough for a little motor and can be connected in parallel to get a couple of watts.
If the LDRs donât give you enough power then a few tricks:-
As the movements will be (hopefully) fairly infrequent then a simple capacitor charge store (charge pump) would give you the extra power youâd need i.e. store three minutes of power to give you a couple of seconds blast.
The second trick is the system I recommend and would probably opt for if I was doing it. The LDR or even PV set up above is a used as the master in a follower system. Thus the imbalance moves a very small motor or solenoid which in turn presses against a micro switch that powers your big motors (as big as you want). As the big motor assembly moves it takes the micro switches with it and thus the switch will be free until the small motor (solenoid) presses it again.
The return to start position can also be achieved fairly easily.
The basics are you have the set up as above but you then have a third LDR (or PV if that's where you are going), the third ORP drives the motor toward the start position but hereâs the fiddle. The third LDR is hidden behind a shade, as the motors pan during the day the third LDR is connected to a bit of string that pulls it out in to the open as the assembly reach near its full end of day travel. The LDR will now be in the perfect position to capture that morning sun. Now with this LDR in the sun the assembly will motor all the way back to the start as it gets there it physically bumps in to the start LDR and pushes it back behind itâs shade and the system then controls as per normal.
I am an electronics guy and happy with programming Arduino etc etc and have done so for years but I will always look at a simple mechanical system first and only head for components and micro controllers if I canât get a very simple mechanical system to do what I want.
So Iâve tried to make suggestion here which will preclude the use of a load of electronic gismos and should be very reliable.
Other have already mentioned shielding the sensors to get selectivity and you may have a problem with dithering when clouds come across etc but maybe not as there would have to be a fair difference across the bridge to start the motors rolling.
If you use your PV's in a bridge set up and put a voice coil (possibly out of a speaker) in the middle of the bridge with a bit of iron in it (not a magnet) then you could put a couple of reed switches on it with one reed at either end of the iron core, so when the coil magnetises one way one of the reed will close and when the other way the other reed will close these reeds would then connect to your bigger motors making the control system very cheap and easy and extremely reliable. Possibly still within your $2 budget not including the bigger motors for actually driving the assembly.
Thanks.
One solution that comes to mind is to minimize the friction of the rotation by suspending the solar chargers on a wire. You would only need really small forces to move it. Of course this depends on the size of the chargers and how you are planning to use them.
How about trying to see if you can get to move the suspended chargers by creating a small electromagnetic force? If you send the surplus current through a long coil [fixed to the base] and you stick a magnet inside, attached to the chargers, it might just provide enough force to move it.
You already pointed out in a comment that your system only needs to run a full circle in one direction. A low-tech solution to get the tracker back to start in the morning, would be to have an extra little motor connected to a battery [or other external power source] with sliding contacts in a semi-circle that are only connected when it reaches a certain angle in the evening, so the motor spins the panels clockwise to the morning position. Then the contact is broken and it waits for light again.
Problem with this solution [if you understand my clumsy description...] is that it will never reach the contact when there is nog enough light in the evening...
I know these are just ideas, but they might give you some new direction. I hope you will manage to find a low-tech way to fix it. Don't give up.
I'll give this idea some thought. Thank you.
PS Does 'muscle wire' count as still low-tech? :-) Just something that came to mind as I was thinking about wires...
1) Physical arrangement of your sensor cells: rather than put them at the bottom of tubes or otherwise use occluding objects to create a large discontinuity in the amount of light they receive as a function of their angle to the sun, why not do away with all that and just have the sensor cells oriented 90 degrees from one another? In this case the equilibrium position would have each cell 45 degrees off of head-on towards the sun.
2) Short-circuit each PV panel with a ~10 ohm resistor so that it behaves like a pyranometer, with a roughly linear response to light intensity. Example and explanation.
3) Use blocking diodes such as a BAT54 to also connect your solar panels to a single Joule Thief circuit, with a common ground, so they are *not* fighting each other.
3) Use an op-amp circuit to measure and compare the short-circuit current of your solar panels, as measured across the 10 ohm resistor. You'll want to use a differencing circuit with adjustable weights to each input so you can compensate for differences in each panel and tune the equilibrium position of the tracker. Example circuit on Wikipedia. You'll want to use a rail-to-rail single-supply op-amp such as the LM324 which can be powered by the Joule Thief output.
4) Got #3 to work? Great, now do it again, except swap the inputs to the differencer. The LM324 is a quad op-amp package so you already have all the parts. When this circuit is working as well, you should always have 0V and Vcc as the outputs of each of your op-amp circuits; which is which will depend on which panel produces more current.
5) Now you have two signals that will be 0V or Vcc depending on your sensor panels - assuming your motor is tiny and geared down a whole lot, you can get away with driving it right from the two op-amp outputs (read the datasheet!) If not, throw in a couple of MOSFETs or power BJTs.
6) Done! One thing you need to think about is how your tracker is going to get from its sunset position to the sunrise position. You will need the sunrise to shine enough on one of your panels to drive the motor. It may be a good idea to separate the orientation of them by even more than 90 degrees; as close as you can get to the limit of 180 degrees while still driving the motor would be best.
7) Note that this design does not deal with hysteresis. It lets the slow motion and limited power available to the positioning motor limit back-and-forth motion of the assembly.
Well, I'm inviting folks to brainstorm ideas in the forlorn hope that we just might come up with something that's both cheap and simple to make. If the analog circuitry gets too complicated, you're right - an Arduino-style controller is an easy fallback, though rather than dedicate a whole board I would expect to just use the underlying chip, and of course power it from the same PV cells :-) If not the ATtiny then there are other PIC chips such as from TI that could do the trick and cost around $1.
Meanwhile yours is the best anwer so far if we stay with the analog circuitry route! (Good links, too). thanks!
You would have to pump the water from the lower can to the upper can every day but this could move TONS! of hardware depending on the size of the cans. If the floats are fairly snug fitting in the cans it is VERY stable. I thought wind gusts would blow it around but the energy of the wind gets absorbed in the water.
http://www.instructables.com/id/Solar-Oven-With-Tracking/
Step 7 describes how he made the tracking system.
I realize he was working with bigger panels and parts, but you may be able to scale the results .. if not perhaps adding a few more small PV cells to increase your electrical output will solve your power problems.
I hope this helps you.
The solar cells could be placed at the bottom of a black tube of about 5 inches in length, this would prevent ambient light for filtering down to the cells when not in direct sunlight. You then can arrange them in a 4 leaf clover pattern with one in the center and 4 around them at approximately 10 degree angel. Using an arduino or other off the shelf micro computer, you simply measure the voltage from each cell, the circuit has got to have control of the motor so that it can always move the center tube into position as the highest output voltage. From there, you simply attach your charging cell in parallel to the center detecting cell.
Overall cost will be more than $2, but since the parts are relatively low cost, it will certainly be cheaper than the pre-assembled sun trackers. The optimal use of a system such as this would be to act as a direction finder for a larger panel array with it's own motor synced to the smaller array, thereby directing larger panels into correct alignment and providing power to a large system capable of running the needs of a small home or camp site while still coming in at lest than $50 for the tracker and if you assemble the PV panels on your own, significant cost savings in the overall system.
This is of course just my first blush on your concerns of ineffective implementation, but you certainly have chosen an excellent low cost source for a fairly complex simple mechanism.
You could alternately use any number of cells in tubes as indicated above mounted on a simple arc of aluminum or light steel and use your control circuit to direct the plane of the solar panel into parallel alignment with the highest voltage sensor cell. This would simplify the plumbing because your suntracker would be stationary and only your collecting panel would move, reducing the weight of the moving parts and reducing the needed voltage to move the motor.
Hope these suggestions help you out. I spent 18 years as an electronics repair tech in the U.S. Army, but my experience with solar cells is purely research related with no hands on work to date. Please keep us posted on your progress.
I understand the joy of using digital microprocessors if you already have one available (I programmed Mindstorms for years before Arduino) but people are rapidly losing the value of simple analog circuits for things like low voltage switching.
The problem gets harder as the panels get smaller - I don't know of cost-effective solutions for say a 5 to 15 watt panel mounted on an RV or a boat. And as you get down to the size of panel used to charge cell phones etc, they get so cheap that it gets really hard to do anything at less than the cost of a second panel.
The arc you mention is what I meant by 'polar mount'. I'll see if I can draw something to add to the instructable over the weekend. (My drawing skills aren't great so I'll probably have to install a CAD package first)
Graham
His design, which was much more compact, consisted of a set of dividers in a "plus" shape with a photoresistor tucked into each of the four corners of the "plus". When the sun shifted enough such that a bit of shadow from the "plus" fell on one half a pair of photoresistors (left-right, up-down), it is time to reorient the panels.
Simply put, the voltage across each photoresistor in a pair is run through a voltage comparator. With a shadow on one half of a pair but not the other, the voltage difference is used to determine which motor needs to turn and in what direction.
I think the mechanical concept and the use of dirt cheap photoresistors is sound perhaps there is a more elegant electronics solution than described. In the 70s, all we had was near single-purpose op amps, voltage comparators, 555 timer chips, and other chips available in DIP packages. :-)
If it ends up that a circuit board has to be designed and populated, I guess using an op-amp would be an option at that point, but I'm hoping we can do this with minimal components. The cost (in manpower if not parts) of making a PCB swamps the cost of just adding a second solar charger. It's all about the cost efficiency.
Using four cells for vertical adjustment is a reasonable idea but needs a second motor and more power to drive it, and basically doubles the electronics. Using a polar mount ought to work just as well and needs no electronics.
You could probably use some good size transistors and some resistors to step down that voltage.
I will try to prototype the circuit when I get a chance but I am sure this will be the cheapest solution without using an external battery or power supply. It would also shut down the drive circuit when the output of the solar charger drops due to fading sunlight.
Terrific idea, thanks for making me think about it.
Cheers,
Chris
I have a feeling that it will oscillate a bit when it looks right at the sun, but there has to be a way to smooth that out.
This adds a bit of complexity, but one solution would be to use a Power Smart Head connected to the PV cells, and the large solar charger mounted to it somehow.
.
Even in full sun, working in parallel, the PV cells don't seem to generate enough power to turn a motor. They *might* be able to turn a motor if we leave the Joule Thief and battery in place from the lights, and only attempt to adjust the platform angle at say 30 minute intervals after the battery has had time to recharge.
If you mean drive the motor from the solar charger instead, then I would prefer not to, simply as a design goal. For example, with an off-the-shelf charger (like the one in my last picture) it's likely to be plugged into some device via the USB power output and there's no other tap for taking power off it for the motors, so we'd need to add a usb T-cable (at the cost of an extra plug and socket) and a diode at least, not to mention one or more likely two solenoids. (Remember, the current has to flow in either direction to turn the motor in either direction - unless we do something crazy like a 340 degree turn in order to turn by -20 degrees!)
Actually that's not a bad point - it does really only need to turn in one direction and then have some sort of reset at the end of the day, and no-one says the reset has to be by driving the motor in reverse... the forward and backward tracking may be cool when waving a flashlight around but the sun only every goes across the sky in one direction...
When I was in post-school we had to design a suntracker. I thought mine was cool until I heard of a guy who just put a slit in a rotating cylinder as the sensor. A photo-transistor near the radial axis sensed when to stop the drive motor. Simple! But needed to be able to go the whole 360degrees fro the next attempt. Limit switches w/ reversing could be easily added..
My attempt was very similar to this instructable, but with 2 simultaneous axes . It was a controls class so I went overboard and tried to get them working together with feedback, all using analog circuitry. Poor thing looked like a lollypop wobbling, spiraling, until it started trying to beat its head on the floor. I paralyzed the one axis and got the lab signed off 100% due to time restraints.