Solar tracking without an optical tracker

This is the place for any magazine-related discussions that don't fit in any of the column discussion boards below.
User avatar
kheston
Posts: 348
Joined: Wed Dec 03, 2003 1:01 am
Location: CA
Contact:

Solar tracking without an optical tracker

Post by kheston » Thu Dec 25, 2008 4:02 pm

Perhaps I'm just looking in the wrong places, but I have yet to find a solar tracker that calculates the sun's position with math alone and aims a dish/panel accordingly. Popular designs rely predominately upon optical tracking.

It seems that Galileo (et al) gave us all of the math required to track the sun some time ago. It's used regularly in control apparatus designed for star gazing. Anyone know why the same technology employed to aim a Meade consumer-grade telescope isn't used with PV? Is there some constraint inherent in solar energy harvesting which makes pure mathematical position calculation impractical that I'm missing?

I've got a parabolic dish and a couple of linear actuator that I'm looking to build a 2-axis control system for. Suggestions?
Kurt - SF Bay

User avatar
MrAl
Posts: 3862
Joined: Fri Jan 11, 2002 1:01 am
Location: NewJersey
Contact:

Post by MrAl » Thu Dec 25, 2008 4:43 pm

Hi there,


With todays microcontrollers anything is possible.

I would think you could take measurements and store them,
then calculate the next angle using say the last 20 measurements.
Might have to measure input power and output power to/from
the converter, but it should be possible.

If you already have equations you are thinking of using,
please present them here so we can all take a look at what
you've got so far.
LEDs vs Bulbs, LEDs are winning.

User avatar
dacflyer
Posts: 4547
Joined: Fri Feb 08, 2002 1:01 am
Location: USA / North Carolina / Fayetteville
Contact:

Post by dacflyer » Thu Dec 25, 2008 6:00 pm

if you search some of the solar sites, i have seen a few that use a thermal means of tracking..not sure how it works tho..
i use a optical one on mine, its a led based tracker..the sensors are actually green leds..go figgure. its from a site called redrock energy
perhaps they might have info on other types of tracking ideas..
good luck.

User avatar
kheston
Posts: 348
Joined: Wed Dec 03, 2003 1:01 am
Location: CA
Contact:

Post by kheston » Thu Dec 25, 2008 6:08 pm

Check out:

http://rredc.nrel.gov/solar/codes_algs/solpos/

I'm thinking Mr. Rymes code will give me the values I'll need (and then some). If I do my job right, I won't ever have to worry whether my dish or array is pointed correctly when the clouds part...
Kurt - SF Bay

User avatar
dacflyer
Posts: 4547
Joined: Fri Feb 08, 2002 1:01 am
Location: USA / North Carolina / Fayetteville
Contact:

Post by dacflyer » Thu Dec 25, 2008 6:21 pm

redrock energy has the units that can work to modify a old satalite dish
i did this to mine, its been great.. you can also make a 2 axis tracker also.
and its not expencive. i paid 35.00 for a single axis..and its simpler to use than anything else i have seen on the market.

User avatar
MrAl
Posts: 3862
Joined: Fri Jan 11, 2002 1:01 am
Location: NewJersey
Contact:

Post by MrAl » Fri Dec 26, 2008 7:06 am

Hi again,


I like this idea very much! I wouldnt mind experimenting with
this a bit myself too at some point.

What i would like to ask now are these:

What kind of load are you driving with the panel?

Have you considered any means of maximum power tracking,
or maybe you dont need that?
LEDs vs Bulbs, LEDs are winning.

User avatar
kheston
Posts: 348
Joined: Wed Dec 03, 2003 1:01 am
Location: CA
Contact:

Post by kheston » Sun Dec 28, 2008 8:08 am

Al,

I think what we're talking about here may function better than an optical tracking scenario when it comes to MPPT. I hadn't thought about it yet, but I'd certainly be interested in any optimization that would get me more throughput in a given year.

I'm going to start with a 100w panel that will simply float-charge my battery bank while I'm away. At present, the 4-alternator charger you (and many others) have helped me with and 1000ah bank make the 240 mile round-trip to my vacation place twice a month. The tracker is over-kill for such a small system, but I don't plan for it to stay small for very long.

Where this sort of tracker will come in really handy is my parabolic dish project. I have an 8ft spun-aluminum dish and two linear actuators. The plan is to try driving a steam turbine or stirling with it, but the math doesn't seem to work until one's solar collector gets much bigger. In the end if all I end up with is a perfectly-aimed death ray to make YouTube vids with it's still a worthwhile project.

--K
Kurt - SF Bay

User avatar
Forrest Mims III
Posts: 12
Joined: Thu Dec 18, 2008 6:59 am
Location: Texas
Contact:

Solar Tracking

Post by Forrest Mims III » Sun Dec 28, 2008 6:51 pm

There are various ways to calculate the position of the sun based on your coordinates, date and time of day. Long ago I developed a set of algorithms to do this for calculating the angle of the sun for a long seriers (since 1990) of daily sun measurements.

But there's much to be said for optical tracking. For example:

1. Optical tracking will account for slight changes in the solar panels.

2. Clouds can sometimes cause the peak solar power to be some distance away from the sun.

In the final analysis, solar tracking for solar panels does not need the precision of a telesope or one of my sunlight monitoring instruments.

Forrest M. Mims III
www.forrestmims.org
www.twitter.com/fmims

User avatar
haklesup
Posts: 3040
Joined: Thu Aug 01, 2002 1:01 am
Location: San Jose CA
Contact:

Post by haklesup » Mon Dec 29, 2008 9:46 am

In theory, a GPS reciever should be able to output this information with great precision. I don't know of a complete project or product but I am aware of numerous applications that run on my GPS device (car navigation like tom tom). The device has windows CE OS and a USB connection you can access. Other devices have similar features.

Once you know your percise time, position and bearing, the location of the sun can be calculated or found in a lookup table.

Anyway the price of new GPS devices is way down now and older models (without built in maps) are starting to populate shelves collecting dust. OEM modules are easy to sample and are intended to be easy to interface to a uController of many types and come with development software libraries. I think projects using these nearly obsolete recievers would be great. I personally own two I use and two I don't.

User avatar
kheston
Posts: 348
Joined: Wed Dec 03, 2003 1:01 am
Location: CA
Contact:

Post by kheston » Mon Dec 29, 2008 10:36 am

Haklesup,

The code I reference earlier in solpos.c does exactly what you allude to, though it calculates the sun's position on the fly rather than looking it up in a table.

I'll be able to pinpoint my lat/lon with one of my eTrex handhelds and it will never change, so no problem there.

Now what I need to do is figure out how to use some combination of encoders and/or accelerometers to point my collector in the direction I calculate.
Kurt - SF Bay

User avatar
haklesup
Posts: 3040
Joined: Thu Aug 01, 2002 1:01 am
Location: San Jose CA
Contact:

Post by haklesup » Mon Dec 29, 2008 11:11 am

Now what I need to do is figure out how to use some combination of encoders and/or accelerometers to point my collector in the direction I calculate.
With a fixed location, you could certainly control it with a PC or stand alone box (w/uController inside) I don't think you need an accelerometer unless you are tracking a fast moving object or wanted feedback so you don't slew the dish so fast you overload the gearbox trying to stop it. Encoders to report the azimuth and inclination should be enough.

I was just thinking that a GPS would also supply you with a very accurate clock that the gadget could use in its calculations. Precision is not all that important to sun tracking though. Take a look at how its done for telescopes, just point it at the nearest star (the sun). Celestron has a few nice integrated products. I wonder if there are user groups who hack those systems.

http://www.eetimes.com/news/latest/show ... =198700125

User avatar
MrAl
Posts: 3862
Joined: Fri Jan 11, 2002 1:01 am
Location: NewJersey
Contact:

Post by MrAl » Mon Dec 29, 2008 11:46 am

Hi again,

This is interesting because i just built a similar panel tilt rotator,
but it was for an LCD panel, not a solar panel :smile:
The idea is the same but it only has to work over a small range of
angle: about +/- 15 degrees from the vertical.

Anyway, on the chance that Forest might return:

Mr. Mims:
Do you think then that his original idea is too much of an overkill
even though microcontroller chips are readily available these days?
The reason i ask is because i was thinking that if this accuracy is
available and it doesnt cost too much to implement why not use it?

BTW, what problems come up with optical sensors with changing
cloud cover?

Back when i was dealing with million dollar solar panels we were
most concerned about max power point tracking, and we did this
based on the delta of power input available. The companies
eventually started to prefer the optical sensor over the purely
electrical algorithm, but the sensor had to be kept clean because
the algorithm for that was depending on an absolute measurement
of the suns intensity. I never quite understood why they would
prefer this method over a purely electrical one where no secondary
measurement is needed and the result works even when the panel
gets a little dirty.
LEDs vs Bulbs, LEDs are winning.

User avatar
haklesup
Posts: 3040
Joined: Thu Aug 01, 2002 1:01 am
Location: San Jose CA
Contact:

Post by haklesup » Mon Dec 29, 2008 12:46 pm

The way I see it, The goal of a solar tracker for a solar panel is to maximize power not necessarily to point exactly at the sun.

On a clear day in an open field that may be the same thing but on a day with multiple layers of clouds, close to dawn or dusk or due to tall landscape or even a body of water, the position of the sun and the angle that produces the most power might be different. An analog optical solution is simple, a mature design with multiple sources, low cost and most of all, inherently maximizes power not position.

I'm no expert but it seems to me that a solar tracker would adjust the inclination in one degree of freedom. Change of the suns path due to seasonal tilt would either be a manual adjustment or set to a compromise. A digital system controlling X, Y and Z position could avoid the second problem but maximizing power when it is not the same as sun position would require auxiliary optical sensors to detect the condition. I see the solar market as growing substantially, there should be room for such control systems so long as they can be manufactured cost competitively. For a sucessful product I would think it will need to do more than just position a solar array but it would need to interact in a greater system to provide input or control by a program.

User avatar
kheston
Posts: 348
Joined: Wed Dec 03, 2003 1:01 am
Location: CA
Contact:

Post by kheston » Mon Dec 29, 2008 5:57 pm

I can certainly see how the optical method would work well for cloudy conditions in the case of PV. I'm just not sure the mechanical device (whatever it turns out to be) I plan to drive with heat generated at the focal point of my parabolic collector will work with anything but direct sunlight.

Since I'll be using both PV and parabolic collectors in my experiments, a system that leverages both would be worth building. Any ideas about how the mechanical/electronic positioning part might look?
Kurt - SF Bay

User avatar
MrAl
Posts: 3862
Joined: Fri Jan 11, 2002 1:01 am
Location: NewJersey
Contact:

Post by MrAl » Tue Dec 30, 2008 1:17 am

Hi again,


Are you asking about how to make a mechanism for moving the
array panel?
If so, how big is the panel?
LEDs vs Bulbs, LEDs are winning.

Post Reply

Who is online

Users browsing this forum: dacflyer, Google [Bot] and 32 guests