Author Topic: Sun Tracker - Part 2 (Read 12125 times)

David HK · « **on:** March 21, 2007, 09:50:07 AM »

Further to my previous posting concerning the electronic control arrangements for a Sun Tracker, I have spent the last few days redesigning the Light Dependent Resistor (LDR) module.

Originally, the first module consisted of one weather proof plastic box with a clear plastic top. Black plastic sheet was glued on top of this to provide a shadow maker between the LDR's. (Above).

The LDR's themselves were mounted on one circuit board so that each one appeared to sit either side of the shadow maker. The LDR's were also covered with copper pipe which in turn was covered by some copper sheet soldered to the pipe. This created a pot with a peephole which was originally about 4 mm (⅛") in diameter.

Weather conditions for testing have not been the best, and generally Hong Kong has recently experienced heavy cloud with intermittent periods of bright sunshine, which is normal for this time of year. Despite this, the system did track, but on occasions the LDR module tracked in advance of the sun. I always had the feeling something was not right and after some thought I decided that the peep holes were too large. It was also possible for bright light to be transmitted through the clear plastic cover of the box, thus affecting the way the LDR's reacted to different light levels.

A complete redesign has been made and the module now consists of two weatherproof boxes each housing one LDR. Sitting between them is the black plastic shadow maker. The LDR's are again covered with copper pipe which is topped with a flat piece of copper. The difference this time is that the peephole is very small - about 1 mm in diameter.

I have been testing this new configuration over the past few days and I am pleased to report that the module is now working as intended and the solar panel array is almost directly facing the sun at any given time. The difference in power output (Watts) and the time spent facing the sun is a wonder to behold. The tracking is spot on until the system stops at the West limit, or, stops because clouds appear and it ceases to track. In this condition it just waits until darkness when the system automatically drives itself back to the East to wait for the next day.

Needless to say I am most pleased with the outcome. The increase in performance and available power now makes it clear that anyone who wants maximum power from solar panel arrays should set them up on a motorized mount. At the height of summer the sustained power output from 8 am until late in the afternoon will be worthy of a study. It merits note that this arrangement has provided the highest number of Watts output I have ever seen from my two solar panels. The mathematics are:- 14 Volts x 0.8 Amps = 11.2 Watts thereabouts. This is a small value, but for an experimental system it now clears the way for me to purchase larger panels and set up a permanent mounting.

Overall, I am very pleased with the electronics and I am confident that this system will last for the rest of my life and beyond. Time and money well spent.

If anyone has any questions, or can offer any hints and tips for improvement, please write a note.

This system is the best thing since sliced bread was invented!

David HK

dinges · « **Reply #1 on:** March 21, 2007, 03:36:12 AM »

Interesting write-up on the sensors. Conduction of light by the housing makes sense. There's a reason in photography most equipment is black (if only on the inside). Might use black encasings myself (though it would heat up in the sun much more).

Your 'shadow plate' seems very large to me; does it have to be that big?

Can you give an estimate by what percentage it increased the output of your system?

Both in gross & nett terms, if possible (i.e. with and without the extra energy used by the electronics and tracking motor).

I doubt that for such small systems tracking is very useful/economical (but hey, my 10W panel will be tracked too, if only for the fun of it ), but it allows one to weed out any errors before starting to design/build a larger tracking array. Plus it's fun to see the panels move all by themself as they track the sun...

David HK · « **Reply #2 on:** March 21, 2007, 03:55:09 AM »

Hello Peter,

Yes, the shadow plate is large - deliberately, so that I can cut it down later. This will have to wait.

I am like you, I have built my control system first to prove it all. My next step is to acquire larger photo voltaic panels to suit my power needs. For this reason i cannot give any worthwhile performance estimates on the output of the system because it does not yet exist - other than my two play around educational panels.

Your last paragraph reflects my own sentiments.

Thanks for your short note.

David HK

wdyasq · « **Reply #3 on:** March 21, 2007, 04:32:55 AM »

I too will be interested in performance boost - even if just a good guess. I plan on putting my massive 88W system on tracker. One of the things I am planning is a concentrator similar to this:

http://www.solar-trackers.com/pictures.htm

I do believe trackers and MPPT charge controllers have uses in PV system. Getting real numbers to see where the greatest actual return of power on one's costs can only be verified by independent testing. I do believe manufacturers and their sales forces at least take the best possible results and on the other end just plain flat lie about what their gear will do.

With a bit of honest information one should be able to accurately guess the parameters of a system before spending a large amount of cash.

Ron

scottsAI · « **Reply #4 on:** March 21, 2007, 10:18:02 AM »

Hello wdyasq,

I like the tracker designs at your link for several reasons.

Snow, rotate the panels to face down during snow. No cleaning, no waiting to melt.

Hail, not too common but when it does hit, has been known to damage panels. Have not designed a hail detector.

Summer, winter positions are easy to add, this axis moves once daily or weekly. (up/down)

Active tracking is not needed. The sun follows a set course. Move to follow. Clouds are not a problem.

Melexis makes a very interesting angle measurement sensor: MLX90316 digikey ~$4.

Problem with trackers are their cost.

The performance benefit of a tracker can be matched by spending equal money on more panels.

Hard to beat stationary panels for long term survivability.

Now if you're willing to build the tracker then the benefits to cost may be in the favor of the tracker. Same for MPPT.

Have fun,

Scott.

TomW · « **Reply #5 on:** March 21, 2007, 11:40:26 AM »

Scott;

The performance benefit of a tracker can be matched by spending equal money on more panels.

I have to whole heartedly disagree here.

In real world use I can nearly double [certainly add 2/3] my incoming power by tracking. My entire tracking system probably would cost $500 if bought new [mines used] To add that much solar would be at least 1500 bux.

Nothin personal but i probably could disprove your theory right here on my setup hands down.

I just wonder if you have experience with both types or are just regurgitating something you read somewhere? I do have experience with both. My tracker happens to be in for repair and I sorely miss it. Curious where you get some of your information, really.

Thats my opinion and I am sticking with it.

Cheers.

TomW

SamoaPower · « **Reply #6 on:** March 21, 2007, 01:33:37 PM »

Scott,

I also have to disagree with much of what you say. This is based on actually having such a system (based on Poulek design) for the past year.

"Summer, winter positions are easy to add, this axis moves once daily or weekly. (up/down)"

Not so. This is a polar axis tracker where the rotational axis is aligned parallel to the earth's rotational axis. To accomplish compensation for the changes in solar declination, the angle of the panels RELATIVE to this axis needs to be changed, not the axis itself. Mechanically, this isn't easy since each panel would have to be changed independantly, not the array as a whole. Fortunately, the loss from not doing this is pretty reasonable - about -5% in total annual energy harvest.

"Active tracking is not needed. The sun follows a set course. Move to follow. Clouds are not a problem."

I strongly disagree that clouds are not an issue. I've found that, for my location, the direction of maximum energy (brightness) is not aligned with the solar position about 15-20% of the time. Differences of up to about 60 degrees have been noted. This is due to variation in cloud layer thickness in different directions. The tracking sensor needs to have sufficient sensitivity to respond to the brightest sky position for best harvest. Active tracking IS needed.

"Problem with trackers are their cost.

The performance benefit of a tracker can be matched by spending equal money on more panels."

The cost of this tracker was about US$120 (mechanical and electronic, sans reflector/concentrator). Over a six month period I measured an improvement of 22% in energy harvest over the same array in a fixed position. This was the tracker only before the soft concentrator was added. The calculated improvement factor for my latitude is 26% for single axis tracking. Not a bad return considering that I would have to spend about $400 on additional panels to accomplish the same thing in a fixed array.

I definately consider a tracker worthwhile, particularly for older panels where output has decreased with age. Mine turned 20 years old last August.

scottsAI · « **Reply #7 on:** March 21, 2007, 02:48:46 PM »

Hello SamoaPower,

The Trackers on the link are single axis trackers that keep the panels perpendicular to the sun in the X and Y planes all day long, resulting in the most power harvested. The seasonal adjustment is raising or lowering the north or south mounting, ONLY. Which can be done with a simple up/down motor.

Other single axis trackers keep the sun perpendicular only for one axis, the other axis is a compromise. Like these: http://en.wikipedia.org/wiki/Image:SolarTrackerRoofView300W200H.jpg

Two axis trackers can match the benefit of the linked trackers with added complexity and cost, as in the drive motors, electronics and mounting.

http://en.wikipedia.org/wiki/Solar_tracker

Polar trackers have one axis aligned close to the axis of rotation of the earth, hence the name polar. So-called "polar" trackers are actually aligned to a perpendicular to the ecliptic, an imaginary disc containing the apparent path of the sun. Simple solar trackers are manually adjusted to compensate for the shift of the ecliptic through the seasons, usually at least twice a year at the equinoxes to establish a position for Autumn and Winter, and another for Spring and Summer. Such trackers are also referred to as "single axis" as only one drive mechanism is needed for daily operation, and so reducing cost or allowing the use of passive tracking methods (described below).

Your picture shown looks like the sun is behind or directly above the panels, the panels should be pointing at the sun. The shadow on the ground should be the shape of the mounted panels?

How many watts are the panels?

Not sure where you got your tracker, most I have looked at are much more expensive. Did you build it?

I did say if you build it... it was good.

The further north you are the more benefit you get from a tracker.

Thanks.

Have fun,

Scott.

ghurd · « **Reply #8 on:** March 21, 2007, 03:48:18 PM »

It's all from one's perspective.

I understand where Scott is coming from.

The biggest non-commercial array I have ever seen is around 300W (this isn't the solar capitol of the world with <0.5H/day, worst case month), `big' local systems are 120W, average is ~50W.

Using Samoa's 22% means 26W on the `big' local systems, or about $200 total for equal that in PVs. Only $100 for the average local 50W system. And nothing extra to maintain or break.

I'm sure the guys I deal with would rather spend more money on more PVs than worry about a tracker they don't understand. It is a little different around here.

G-

scottsAI · « **Reply #9 on:** March 21, 2007, 04:15:09 PM »

Hello TomW,

I always have so much fun talking to you.

Please attach a link, I'm looking for a cheap tracking system.

Everything I have looked at for a 2kw system cost about $2k, not installed. Installation more than doubles cost, (concrete, labor etc) assuming it's in an easy location, then lets remember the area around it that must be available etc.

So the math is easy, 2kw panels cost $12k, Tracker is $4k installed = 30% cost, on par with cost of more panels increasing output 30%.

DIY is definitely worth doing it, if you can handle it. The Why we are here.

In real world use I can nearly double [certainly add 2/3] my incoming power by tracking.

I have never seen a claim above 30% for tracking, most as SamoaPower states just below your post in the mid 20's. Now the further north you go the more benefit tracking gives, you must be at the north pole?

I just wonder if you have experience with both types or are just regurgitating something you read somewhere?

Yes! I read a LOT. I can design hardware, electronics and software for both systems.

Send me your requirements and I will show you, it will be fun! (I do ask a lot of questions:-)

There is no magic in tracking, the math is simple, understanding a bit harder. Now coming up with the best solution is magic. The single axis shown at the link is for me, one of the better solutions.

Thanks.

Have fun,

Scott.

SamoaPower · « **Reply #10 on:** March 21, 2007, 04:46:56 PM »

Wow, I hardly know where to begin.

"The Trackers on the link are single axis trackers that keep the panels perpendicular to the sun in the X and Y planes all day long, resulting in the most power harvested."

This is only true at the equinoxes (like today) when the plane of the panels are parallel to the rotation axis. It's the autumnal equinox for me - southern hemisphere. As the solar declination changes from today's zero degrees, the tracking error increases in both X and Y. We seem to agree here.

"The seasonal adjustment is raising or lowering the north or south mounting, ONLY."

WRONG! A polar axis tracker has to be maintained parallel to the earths rotational axis at all times. In tracking, we are trying to compensate for the earth's rotation to point to a fixed point on the celestial sphere. If you change the axis parallelism additional errors occur. Do the geometry to see it.

To correct for changes in solar declination throughout the year, the plane of the panels must be changed relative to the rotation axis to match the solar declination, but the rotational axis must remain constant.

"So-called "polar" trackers are actually aligned to a perpendicular to the ecliptic ..."

As many have discovered, wikipedia is often wrong and this is another case. By aligning the rotational axis parallel to the earth's axis you are actually aligning it perpendicular to the CELESTIAL EQUATOR, not the ecliptic. It is tilted at 23.5 deg. to the plane of the ecliptic as is the earth.

"...to the ecliptic, an imaginary disc containing the apparent path of the sun. Simple solar trackers are manually adjusted to compensate for the shift of the ecliptic through the seasons ..."

Another definition of ecliptic is: the plane of the earth's orbit extended infinitely (Webster's).

The ecliptic does not shift.

You have a sharp eye. The array was taken out of track mode and manually positioned for photographic purposes only.

The panels are 40 watts each.

Yes, the tracker is homebrew with some parts coming from eBay.

Yes, the gain from a tracker is dependent on latitude and time of year. Long term energy harvest is the thing to look at. Six months is adequate to see the seasonal variations since the other half of the year is a mirror.

David HK · « **Reply #11 on:** March 21, 2007, 05:06:11 PM »

Chaps,

This is all wonderful comment.

However, could I just say that Hong Kong's latitude is around 22.6 degrees North and during the summer we have a huge amount of sunshine. All contributors seem to be missing the point that we all live at different points of latitude, in different countries and experience different weather conditions.

Surely a more scientific way of looking at this would be for a group of us to liase directly with each other, document our tracking systems, and on certain nominated days (or day) in the year, all take readings on power generation, weather on the day, and so on, and publish the data on this site.

The merits of whether a system is a single axis tracker, or an azimuth and elevation tracker, often depend on personal preference. Costs that are expensive for one man may be of no consequence to another. So everything is different for everybody.

David HK

TomW · « **Reply #12 on:** March 21, 2007, 06:04:29 PM »

All;

I seem to have made some outrageous miscalculations of the boost I get from tracking. An honest mistake, I assure you.

Math is my worst subject. Ask the IRC guys and they will tell you the same.

In full sun my 300 watts [rated] of panels at nominal 24 volts puts out somewhere above 9 and just below 10 amps on my meter. I think of it as 9.75 but the hash marks are too close to tell the fractional values accurately.

Scenario:

Early AM, clear day with panels facing due South. I would get a couple amps this way and gradually increases till solar noon where it reverses and slowly losses amps. I see that even a few degrees of angle off of straight on that the amps drops significantly.

Same day or as close as possible anyway. With tracking the panels are dead on when Sol gets above the horizon. As I track I maintain that same full ampere reading thru the entire day until sundown. That means I get a good bit more power all day and, while it may not be as much as I stated earlier, there are times when I get several times the amps I would fixed south. I live at or about 43 North latitude. My Sat dish polar mount is capable of tracking in both planes if I added another jack but I just adjust elevation every few weeks or when I think it looks like it would help.

My panels can be adjusted 100% dead on to the sun any day of the year at any time so i guess the mount is capable of exact tracking in both planes.

Costs were a $75 4" X15 foot steel pipe, 3 bags of sacrete [<$10], digging a hole, setting pole and capping with my old unused C band satellite dish mount. The actual rack that the panels mount to is 3" angle iron and designed to hold 4 SP 75 Seimens panels with some room between. I could easily add some steel and double the rack space or more. The polar mount claims it is capable of supporting 1500# so it will handle anything I can afford. The rack itself was $75 I think but not sure. I don't weld so I hire welding done.

Tracking electronics are fairly user unfriendly and computer based and I can't cite an actual cost being one off prototypes.

Regardless, in my situation, I find tracking to be very cost effective compared to more panels.

Out of pocket expense was under $200 and that wouldn't buy much of a solar panel. These options are not for everyone but work for me in my situation.

If I was motivated I could do data logging of 2 panels in each configuration for a year and get actual values. I am confident it would exceed the oft stated 30%. Without proof my numbers are guesses.

I apologize if I misled or offended anyone. I always prefer real world observation to trusting someones numbers derived mathematically. You see, nobody has told the bees that mathematically speaking they should be unable to fly.

Just 'fessing up.

Cheers.

TomW

scottsAI · « **Reply #13 on:** March 21, 2007, 08:49:08 PM »

Hello SamoaPower,

I always learn something from you or because of you:-)

This link is showing the apparent motion of the sun, Never seen anything different:

http://www.redrok.com/images/sunpath.gif

Movie or applet showing the same: (Cool)

http://www.math.nus.edu.sg/aslaksen/applets/earth_fixed_lat/earth_fixed_lat.html

Your description says picture is not correct as shown?

Single axis tracker and how it works: (not that great, anybody have anything better?)

http://www.solar-trackers.com/p1-2.htm

With a single active axis (rotation east to west) the panel is kept perpendicular in X and Y planes all day long. Seasonal tilt is fixed by raising or lowering the north support for example.

Unfortunately I do not understand your point. Can you supply a link showing what your saying (with the math and or a picture)?

Wikipedia has some of the best explanations for complicated things around. But I did not use it this time!

Looking over your 8 panels of 40 watts each, assuming $6/watt Cost:

$6 * 40 * 8 = $1920 new. With 25% gain with tracker, for break even the tracker must cost less than $480. Since a tracker has moving parts and something else to break, then I would buy the two extra panel(s) for the $480. Or DIY for less then go for the tracker.

Have fun,

Scott.

wpowokal · « **Reply #14 on:** March 21, 2007, 10:15:57 PM »

Never mind Tom I believe you, I know I over double my incomming power when tracking, and I did publish the figures herin.

But even then there were doubters, working on my second tracking aray now, which will see all panels tracking. I'm at 35 deg south the sun rises and sets south of me in high summer and well north in mid winter, and I need the most power in summer, so for me it works.

I suspect people can not comprehend the increase in the sholder periods, despite you explaining it.

allan down under

RP · « **Reply #15 on:** March 22, 2007, 05:11:58 AM »

Scott,

If I may: I think I can explain the confusion. Picture a line of panels mounted on a pipe-axle as shown in SamoaPower's picture above.

I believe he is agreeing that the panels must be retilted throughout the year to stay in alignement with the sun but the actual alignment of the pipe-axle does not. Notice that in the gif image link you gave above, the "axis" of the sun's rotation above the sky is the same all year even though the vertical angle changes.

It may help to picture your solar panel as being at the center of a "celestial sphere" with the sun moving in a great-circle across its surface.

dinges · « **Reply #16 on:** March 22, 2007, 05:57:54 AM »

Samoa,

I'd be interested in a few close-up shots of that tracker, if possible. Checked your diary and stories but I can't find a story of that project of yours.

Also, the mirrors that you use, are they stainless steel ?

Regards,

Bruce S · « **Reply #17 on:** March 22, 2007, 07:43:16 AM »

David et-al;

I too would like to take part in this suggestion.

My HF panels are not mounted at all. I have them pointed according to DOE's website, and have been comtemplating a tracker as well.

Been right in front of the meter when clouds have gone over head and I can say there is almost an instantaneous reaction.

I can easily begin documenting what the panels are doing on a day by day production now and then do the same after I pick a tracker type and mount it.

I am intreagued by Soma's system and have been wondering about this system and have several questions about it, but I am also interested in David HK's DIY system.

I have been wanting to put together a logging system with one of our old laptops, just always find other "stuff" to do. Maybe this'll get me going on it.

If there are others willing to this I'm game.

My system is currently a modest 45 watter from HF.

Thoughts?

Bruce S

hydrosun · « **Reply #18 on:** March 22, 2007, 08:39:57 AM »

One point of tracking is often overlooked. In the real world we want the system that gives the most usefull energy. In most systems the solar panels fill up the batteries on a sunny day and then the charge controller disconnects the panels. Even though the panels are producing power none is being used. On a fixed array that peak in output in the middle of the day the voltage is artificially pushed up above the charge controll level and the power is rejected. With a smaller array the output is more even all day long and more power may be put into the battery. And if you are using power during the day the input is spread over a longer period and less is stored and retrieved from the battery, increasing efficiency. A lot of times I'm not home until after 5 pm and with the tracker I have more direct solar power coming in to cook dinner with. I have 1200 watts on a home built tracker so the difference between a south mount and tracked mount is significant at that time of day. I used to tell people that it wasn't cost effictive to buy a tracker but now a home built one tilts the advantage over adding more panels.

Chris

SamoaPower · « **Reply #19 on:** March 22, 2007, 09:02:21 AM »

Hi Peter,

As requested, with apologies to David HK:

Yes, the ridge soft concentrator is 0.7mm stainless.

A post is in the works with gist being new life for old panels. Need to finish the buck converter - doing PC board now.

alancorey · « **Reply #20 on:** March 22, 2007, 09:50:55 AM »

I like the idea of tracking overall. Back in the late 70's I built a tracker with an antenna rotor, a couple of relays, and some op amps. The collector was a reflector that was parabolic in one direction heating a piece of copper tubing painted black that I was running water through. It was all just a small experiment and I was mostly unemployed at the time. I had the same kind of detector arrangement with 2 photoresistors and a shadow board between.

Now that I'm getting ready to build for retirement and considering 1 - 1.5 KW of PV cells because it's cheaper than getting on the grid, I think differently about tracking. Too risky. Not only is there the extra expense to begin with and whatever lubrication/maintainence may be needed, but am I really going to trust something I built to hold $5000 worth of solar panels through high winds? Probably not. I'm at about 42.5 degrees north latitude, and I think I'm going to just build my roof at 45 degrees and hitch the panels down well. I plan on using wind and hydro too, so the loss of 25% or so on solar isn't that big a deal. I've seen too many pictures of smashed panels (and windmills) on here. I don't like to gamble. Maybe I'll track my hot water panels but PV is too expensive to take chances with.

Alan

rossw · « **Reply #21 on:** March 22, 2007, 02:49:00 PM »

Well, most of what I was going to add has already been said, however some points I'd like to make:

I have a home-brew setup of currently only 320 watts, single-axis tracking with second axis manually adjusted. (I'll fix that one day).

[img width= height=]http://house.albury.net/17sep2006/thumb.MVC-488X.JPG[/img]

[img width= height=]http://house.albury.net/17sep2006/thumb.MVC-492X.JPG[/img]

It uses a conventional linear actuator for east/west tracking and a large bolt for inclination (I adjust it every couple of months).

The tracker has doubled my daily watt-hours of collection compared to when it was a fixed panel. Getting 1650 watt-hours/day peak as a fixed panel to getting more than 3500 watt-hours/day from the same panel. (Granted, this has been over summer, the gain may be less in winter). (I'm about 36 degrees south).

As for cost - the tracker cost me about $500 (Australian) in parts and materials, a couple of hours to dig a decent hole and concrete the post etc all up probably about a day to assemble, erect, test, align etc.

The cost of adding enough panels to double output, excluding the cost of mounting frames, additional charge-controller and so on which would be required) would be $3200, probably close to $4400 by the time one includes the extra hardware.

My tracker uses two LEDs as photosensors - mounted with a 90 degree angle between them. No shade device, the LEDs are connected back-to-back, and run to an amplifier and comparator to provide east/centre/west detection. It's quite sensitive enough to position to within 1 degree, which is less than 0.02% reduction in output due to cosine error.

Bruce S · « **Reply #22 on:** March 23, 2007, 08:23:33 AM »

rossw;

Only 320 watts? I thought I was doing good with my little 45 watter.

Would you happen to have a write up of your tracker? nice looking land too.:--)

Bruce S

rossw · « **Reply #23 on:** March 24, 2007, 04:54:40 PM »

Yeah, "only". I have a 1000W turbine too - which is great on windy days, but sadly I live in a lousy wind area.

I am planning to add another 660W of PV in the next few months, that should really help me out. (I run too many computers and other equipment, I need the watt-hours!)

I don't really have a write-up of the tracker, although I'm building a new and improved controller. Perhaps when thats done I can post a write-up of that, there has been more than passing interest in it.

The mechanics are fairly self-evident from the pics

News:

Author Topic: Sun Tracker - Part 2 (Read 12125 times)