Tracking - why so complicated ?

[Dave B]

I'm new to solar and leanring a bunch but one thing I don't understand ... maybe I'm missing it some where but why doesn't there seem to be more "clock type trackers" rather than the complexities of actually tracking the sun ? I have never seen the sun do anything different than what is already known about where it will be when in the sky for any given date any where in the world. Seems to me you could manually tilt the array twice a year and have it always rotate horizontal to the ground just based on time. It wouldn't even have to be continuos to simplify things even more. How about rotating (not tilting) about it's pole axis maybe a certain number of degrees every half hour or something ? I would bet the sun will be there, why have the tracker track the sun ? I mean, don't we know it's not going to do loop the loops ? If that happens I think we are all in trouble. This could be as simple as you want, maybe even just a motor with an on off timer jogging through the day then returning home at a given time with limit switches. Any of this would be better than a fixed North South array.  Dave B.

[TomW]

I did this with a computer based relay driver board. Every X minutes move Y ticks of the actuator sensor and  reset to east at midnite.

So it has been done and I posted here about it.

Very simple and effective but needed a computer to do it.

Tom

[Dave B]

Tom,

 That's great to hear, I did not see that any where. I'm thinking simpler yet, just a bit more sophisticated than rotating the array by hand every hour or so. Slip rings are readily available too, make it non-reversing and go all the way around to home (or by time) during the night. Just a BA clock. Tracking the "beam" though is cool and makes for a neat demo, I guess that sells more.  Dave B.

I did this with a computer based relay driver board. Every X minutes move Y ticks of the actuator sensor and  reset to east at midnite.

So it has been done and I posted here about it.

Very simple and effective but needed a computer to do it.

Tom

[rossw]

Tom,

 That's great to hear, I did not see that any where. I'm thinking simpler yet, just a bit more sophisticated than rotating the array by hand every hour or so. Slip rings are readily available too, make it non-reversing and go all the way around to home (or by time) during the night. Just a BA clock.

As Tom says, tracking by time isn't terribly difficult, but the process to actually work out where the sun will be with any precision is a computationally demanding task. I'm doing it on one of my systems for other purposes, and it does take a reasonable number of CPU cycles.

That said, I'm doing similarly to Tom - I have 4 arrays "tracking" (I use the term very loosely), by a very simple system of a gated clock. From sunup until sundown, the system drives the panels west 2-4 seconds every 20-30 minutes (exact times vary depending on time of year). A while after sundown it drives the panels to an overnight "park" position, and just before sunrise it drives them fully home.

With flat panels, an angle offset of 10 degrees from "ideal" represents only 1.5% loss of power, and for the sun to move from 10 degrees "east", past the panel to 10 degrees "west" takes roughly 90 mins to 2 hours (depending on time of day). So as long as you can move the panels to "close enough" to the right place, you're not really wasting a lot of potential.

Tracking the "beam" though is cool and makes for a neat demo, I guess that sells more.

Tracking the "beam" as you put it, is actually a fairly trivial task for a very basic electronic device to do. Certainly much easier than a clock of sufficient accuracy and features, and way easier than something able to compute the suns position at any given time. The mechanics may be more challenging!

As for ideal tracking, you should look up "polar mounts" or "polar tracking". Ideally, tracking in the polar plane, but with an additional offset for time of year should make the tracking process straightforward. The mechanics of such mounts often makes it easier to simply do 2-axis tracking electronically. 2-axis tracking also gives you the advantage of being able to (potentially) stow the panels in a reasonably safe position during storms.

[Dave B]

Hi Ross,

 Interesting to hear of your set up and comments. The kicker to my thinking about time tracking is that the array will always be at the ready when the sun peaks out. No seeking, computating, moving, settling, hysterisis etc. etc. for when the sun finally appears an hour or so later from behind the clouds. Maybe I see tracking in a different way because I am from the Northeast. I want to be ideally ready where ever and when ever the sun does show it's rays, no sense in looking for (and then reacting to) the sun when I already know where it is even if I can't see it. Just like wind, opinions and comments vary widely on solar and are often influenced by location.  Dave B.

Tom,

 That's great to hear, I did not see that any where. I'm thinking simpler yet, just a bit more sophisticated than rotating the array by hand every hour or so. Slip rings are readily available too, make it non-reversing and go all the way around to home (or by time) during the night. Just a BA clock.

As Tom says, tracking by time isn't terribly difficult, but the process to actually work out where the sun will be with any precision is a computationally demanding task. I'm doing it on one of my systems for other purposes, and it does take a reasonable number of CPU cycles.

That said, I'm doing similarly to Tom - I have 4 arrays "tracking" (I use the term very loosely), by a very simple system of a gated clock. From sunup until sundown, the system drives the panels west 2-4 seconds every 20-30 minutes (exact times vary depending on time of year). A while after sundown it drives the panels to an overnight "park" position, and just before sunrise it drives them fully home.

With flat panels, an angle offset of 10 degrees from "ideal" represents only 1.5% loss of power, and for the sun to move from 10 degrees "east", past the panel to 10 degrees "west" takes roughly 90 mins to 2 hours (depending on time of day). So as long as you can move the panels to "close enough" to the right place, you're not really wasting a lot of potential.

Tracking the "beam" though is cool and makes for a neat demo, I guess that sells more.

Tracking the "beam" as you put it, is actually a fairly trivial task for a very basic electronic device to do. Certainly much easier than a clock of sufficient accuracy and features, and way easier than something able to compute the suns position at any given time. The mechanics may be more challenging!

As for ideal tracking, you should look up "polar mounts" or "polar tracking". Ideally, tracking in the polar plane, but with an additional offset for time of year should make the tracking process straightforward. The mechanics of such mounts often makes it easier to simply do 2-axis tracking electronically. 2-axis tracking also gives you the advantage of being able to (potentially) stow the panels in a reasonably safe position during storms.

[rossw]

Interesting to hear of your set up and comments. The kicker to my thinking about time tracking is that the array will always be at the ready when the sun peaks out. No seeking, computating, moving, settling, hysterisis etc. etc. for when the sun finally appears an hour or so later from behind the clouds. Maybe I see tracking in a different way because I am from the Northeast. I want to be ideally ready where ever and when ever the sun does show it's rays, no sense in looking for (and then reacting to) the sun when I already know where it is even if I can't see it. Just like wind, opinions and comments vary widely on solar and are often influenced by location.

Well, just playing devils advocate for a moment....

The highest output from the panels is not ALWAYS pointing where the sun is.
Sure, on averages it is.... undeniably. But SOMETIMES, like when there's a dirty great black cloud between you and the sun, you can get a LOT more power by pointing your panels 20, 30, 40 degrees or more away - where a big white fluffy reflective cloud is.

A suitably designed tracker can take advantage of these scenarios and point the panels "where the power is" under these conditions, whereas a "lets point at the sun and hope" system cannot.

[DamonHD]

Many claim (eg in patents and experiments), and I've seen evidence to support, that on a cloudy day generally the best orientation is directly upwards, not tracking at all.

For that reason I have a small amount of near-horizontal off-grid stuff to help in winter.

Rgds

Damon

[defed]

i have some old c-band parts, including a the mounts, actuator and receiver w/ timer functions to move the dish.  i was thinking about messing around w/ this and using the timers (4 or 6) to move the array.

ross, do you think tracking w/ a polar mount is better than just putting the array on the pole, titled to the correct angle, and then just spinning it on the pole?  i was trying to picture if one way was better than the other.  i do recall from my c-band days, that on certain days of the year (equinoxes?) that the sun tracked the same as the satellites...so that when the satellite was in between the sun and dish, the solar radiation would knock out the satellite signal.

[Tritium]

I have a telescope clock motor and gear drive I am planning to use.

Thurmond

[Dave B]

"With flat panels, an angle offset of 10 degrees from "ideal" represents only 1.5% loss of power, and for the sun to move from 10 degrees "east", past the panel to 10 degrees "west" takes roughly 90 mins to 2 hours (depending on time of day). So as long as you can move the panels to "close enough" to the right place, you're not really wasting a lot of potential"

  This explains my thoughts exactly about trackers that follow the beam as opposed to always being there (or close enough) with just rotating either stepped, continuos clock or timed. There would be no searching and delay to find the sun. Manually tilt the array maybe twice a year and be done with it. Again, tracking the beam is cool though but I'm not buying, I'm building.  Dave B.

[hydrosun]

I can buy a tracking circuit for less than $50 from several sources. Done.......   Or I can reinvent the wheel and use time to track the position of the sun. There was an article in Homepower magazine a few years ago about someone used a timer to open a water valve to fill a bucket to provide weight against a spring to move the solar panels. When the valve closed the bucket slowly leaked out the water to return to the east.  So there are a lot of ways to get the solar panels to face the sun. I chose to use an established avenue.
Chris

[rossw]

ross, do you think tracking w/ a polar mount is better than just putting the array on the pole, titled to the correct angle, and then just spinning it on the pole?

Well, splitting hairs a bit, but yes I think a properly aligned polar tracker would be better.

Do I think it's sufficiently "better" to justify the additional complexity and cost is a different question with a different answer.

[birdhouse]

i would love to see someone on the board build a tracker with the copper tubes connected to a hydraulic ram and filled with oiled freon-12.  much like 

http://www.motherearthnews.com/Do-It-Yourself/1977-11-01/Mothers-Super-Simple-Solar-Tracker.aspx

then they could work out all the kinks, so i could follow in their footsteps! 

i've always wanted to make one like that.  takes zero electricity, auto returns to east in the am, and should run flawlessly for a very long time. 

build it, and i will follow!   

adam

[rossw]

i would love to see someone on the board build a tracker with the copper tubes connected to a hydraulic ram and filled with oiled freon-12.  much like 

I've always been concerned about how these finely-balanced systems work in the real world - with storms, buffeting wind etc.

I had three of my trackers litterally screwed out of the ground during a storm. They weigh around 130Kg (290 lbs) each - and thats without half as much again in concrete in the 3' deep hole.

[birdhouse]

ross-
i hear ya.  in the first portion of the article, they talk or just using the pressure of the freon to tranfer from side to side, then later hooking the tubes to a hydraulic cylinder.  maybe double cylinders for a larger mount?  i was planning on 600w or so max!

i'd bet the hydraulic cylinder would hold if everything else was built strong enough.  maybe a damper for the evenings, when sun isn't keeping he pressure up??

bummer about your mishap.  did you post a story/pics of your wind mishap?  i'd love to see/read about it.

may the wind be on your turbine and not your panels!!

adam

[Dave B]

I guess the near perfection of generating electricity with no moving parts is just too boring. I like neat stuff too but to me the simplicity of photovoltaic in itself is already about as neat as it gets. Wind is fun and photovoltaic is perfect, great combination.  Dave B.

[TrackerJack]

With me, tracking the sun is half of the infatuation I have with inventing. It could be the mechanism that Harrison Ford use to make the ceiling fan go round in the movie, Mosquito Coast.
Of course there was the professor in Gilligan's Island, but with me it all started with an old cartoon of two little poor kids waking up on Christmas morning found no presants under the tree.
All of a sudden this old bearded inventor dude shows up and starts making presents out of things around the house. I specially remember the toy steam train made from cups and stuff.
For a locomotive's engine he use a coffee peculator.  Man, at 6 years old, I got to work.

[ghurd]

I've always been concerned about how these finely-balanced systems work in the real world - with storms, buffeting wind etc.

How do they work?
Seems they don't.  Or don't work very well in many places.
At least 1 major retailer stopped selling them because they only work when they feel like it.
Here, not uncommon to get a 20F drop at noon, or a 20F rise at midnight.

Then again, back when I was important, some guy from a German company showed up HERE, in late winter(?), with his fancy 2-axis tracking doohickey to demo.
It looked like a break dancer on crack having a seizure.
We brought it into the garage so it could only see "the South sky".
It looked like a break dancer on crack having a seizure, while mostly facing South.

It would have been a fine unit in New Mexico, but he showed up in winter in the snow belt.
Middle of summer here now.  Been waiting ~3 weeks for 4 hours of semi-clear sun around solar noon for a silly experiment.
G-

[birdhouse]

yea, i've got some doubts about a freon tracker as well.  from reading the article, freon 12 boils at 32deg F.  that seems like it would create massive pressure at 110deg F.  yet might be sluggish at -10deg F. 

it also mentions only filling the side tubes to half way full to eliminate blow outs in the copper pipes or sweating job.  it also mentions using oiled freon 12 to keep the o-rings and such in the hydraulic cylinder from drying out with just the freon. 

i've put quite a bit of thought into this, and was really hoping someone else would do it first, but maybe i'll have to tackle it, though i do move slow with my re projects. 

my thoughts were to pour a bit of mobile one, or hydro fluid into the tubes before pulling the vacume and adding the freon.  hoping that the oil/fluid would stay in the bottom area near the hydraulic cylinder and it's flexible transfer tubes, and the freon would sit on top.  maybe this is a misconception, as i don't know a ton about how different oils/freon mix/don't mix. 

i also thought it would be a good idea to fill the tubes all the way full, yet use a blow off valve on each tube set to the max the builder thought the copper/sweats were good for.  that way you'd maintain the most amount of fluid, yet not have a copper blow out ruin your panels.  idea being it may work over a wide variety of temperatures. 

for any of you that have read the whole link i posted above, do you think this is a feasible system?  i really like the idea of it, and am willing to go above and beyond it, yet don't want to waste money and time on something that will never work. 

thoughts welcome, and let me know if this is hyjacking, and i'll gladly start a new thread.

adam

[rossw]

for any of you that have read the whole link i posted above, do you think this is a feasible system?  i really like the idea of it, and am willing to go above and beyond it, yet don't want to waste money and time on something that will never work. 

Call me a killjoy, but *I* sure as hell won't be spending any of *my* hard earned cash on one.

[TrackerJack]

I'm new to solar and leanring a bunch but one thing I don't understand ... maybe I'm missing it some where but why doesn't there seem to be more "clock type trackers" rather than the complexities of actually tracking the sun ? I have never seen the sun do anything different than what is already known about where it will be when in the sky for any given date any where in the world. Seems to me you could manually tilt the array twice a year and have it always rotate horizontal to the ground just based on time. It wouldn't even have to be continuous to simplify things even more. How about rotating (not tilting) about it's pole axis maybe a certain number of degrees every half hour or something ? I would bet the sun will be there, why have the tracker track the sun ? I mean, don't we know it's not going to do loop the loops ? If that happens I think we are all in trouble. This could be as simple as you want, maybe even just a motor with an on off timer jogging through the day then returning home at a given time with limit switches. Any of this would be better than a fixed North South array.  Dave B.

 This will be a rotating tracker controlled by a Flexcharge Programmable Digital 12 volt d.c. timer and  load controller. Simple yet effective.
It will hold 16- 130 watt Sharp panels. The Flexcharge will allow 8 -1 min..power on to any motor up to 10 amp.
I will be using a relay to turn on a 3/4 hp geared motor 40:1 ratio.  Belt driven  
      

[ghurd]

This will be a rotating tracker controlled by a Flexcharge Programmable Digital 12 volt d.c. timer and  load controller. Simple yet effective.
It will hold 16- 130 watt Sharp panels. The Flexcharge will allow 8 -1 min..power on to any motor up to 10 amp.
I will be using a relay to turn on a 3/4 hp geared motor 40:1 ratio.  Belt driven

A DC motor will turn on for a set amount of time?
Not sure that will work out quite as exact as expected.

7 cycles moving west?  1 moving east to park?  About 25 degrees rotation per west cycle?

No good sun for a few days.  Battery at 12.2V.
And a cold day when the controller decides to equalize at 15.5V.
27% difference in voltage makes a 27% greater rotation per minute.  25 degrees goes to 32 degrees.
Not a big deal I suppose.  Even if it is off 20 degrees at 3PM, it is better than not tracking at all!

Looking forward to seeing how it works,
G-

[DanG]

8 -1 min..power on to any motor up to 10 amp

No reason its forced to run for the entire time - cam following switch triggers relay reset etc...

Five peak solar hours tracking is 75°, six hours 90°, seven hours 105° - then dusk and dawn is catch as catch can.

I almost posted here a method I'd been thinking about but would never post thinking* out loud...

A nice dual-cable-spool winch center drawing a cable bow for a carousel to pivot 120°, curved pipe rails and concave rollers with relay logic for 8 or 10 adjustments after start signal... Ya, I used to work for a Railroad.

[TrackerJack]

I'm on the road right now, but when I get home I will show with diagram and explain how I will be able to place each movement into the same place 8 times in a day with the return making 9.
Also the mechanics of tilting array while rotating. All with low tech switches and such. The array will be capable of being locked into an upright position during freak winds, and left to swing freely to follow the wind, somewhat like a weather vain.

[adobejoe]

Here is another way. Mount PV panels on diagonal (these are 210 W, 16 of them) with inexpensive fence post and rotation sleeve. Use linear actuator on the high corner and operate all 16 panels with one PIC program, like the Parrallax STAMP, on a separate battery, that also powers the linear actuators. All based on time-of-day, and rotate every hour 8 times a day. Returns at midnight. This is the most stable system in high winds, and is very strong and low cost.

Adobe Joe in Wyoming

[BrianSmith]

I like your setup Adobe Joe, do you have a pic of what it looks like underneath a panel?  It looks like its only supported on the two corners.  I'm having trouble figuring out how it works.  Thanks.

[XeonPony]

it also mentions only filling the side tubes to half way full to eliminate blow outs in the copper pipes or sweating job.  it also mentions using oiled freon 12 to keep the o-rings and such in the hydraulic cylinder from drying out with just the freon. 

My thoughts were to pour a bit of mobile one, or hydro fluid into the tubes before pulling the vacuum and adding the freon.  hoping that the oil/fluid would stay in the bottom area near the hydraulic cylinder and it's flexible transfer tubes, and the freon would sit on top.  maybe this is a misconception, as i don't know a ton about how different oils/freon mix/don't mix. 

adam

For this idea you would use 300sus mineral oil as it must be soluble by the R-12 (Freon tm) How ever I would not recommend R-12 due to it's ODP value, R-152a or R-134a is readily available from air duster cans of all things with a bit of propane mixed in to assist in oil transfer so you can use mineral oil.

Freon is the Dupont trade name for R-12, so you can use one or the other, using both gets redundant and can confuse the issue

Sincerely the HVAC pony

[Tritium]

R-12 has almost the same boiling point as propane.

Thurmond

[birdhouse]

thanks for the info pony and tritium! 
soo, you folks with a far better understanding of these unusual fluids...  do you think a collector of the type i linked to above is workable through a fairly wide range of temps?  like -10F to 110F?? 

in my head it would work perfectly, but what goes on in my head, and reality can be quite different at times!   

adam

[oztules]

Weigh up the cost.... is it cheaper to use 30% more panel to fix the array?...usually is.
My experience is limited. On this island, very very fancy gps tracking arrays FAIL.... timed arrays FAIL... fixed arrays do not.


..............oztules.

[WindriderNM]

Zomeworks makes a tracker with no moving parts. There is a plan for one in an old Mother Earth News There are pics of this on a previous post

[WindriderNM]

Universal Track Racksâ„¢ F-Series passive solar tracker with integral early morning rapid return system, partial pre-assembly and easy installation.
How Trackers Work

1 Sunrise

The Track Rack begins the day facing west. As the sun rises in the east, it heats the unshaded west-side canister, forcing liquid into the shaded east-side canister. As liquid moves through a copper tube to the east-side canister, the tracker rotates so that it faces east.

2 Morning
The heating of the liquid is controlled by the aluminum shadow plates. When one canister is exposed to the sun more than the other, its vapor pressure increases, forcing liquid to the cooler, shaded side. The shifting weight of the liquid causes the rack to rotate until the canisters are equally shaded.

3 Day
As the sun moves, the rack follows (at approximately 15 degrees per hour), continually seeking equilibrium as liquid moves from one side of the tracker to the other.

4 Evening



The rack completes its daily cycle facing west. It remains in this position overnight until it is "awakened" by the rising sun the following morning.

This is copied from the zomeworks web site the site has diagrams that make this clearer this looks fairly easy to make. I have seen a few of these and they seem to work well.

[Nil]

I have three of the zomeworks trackers holding 12 Arco panels each. I never would have bought them for 'baked' Arco's but $1k for the trackers and panels I could not pass them up.

I have never seen another tracker work for long periods of time but these ones seem to work as expected. During clear days they follow the sun as they should. During cloudy periods they become erratic, at times I have seen all three pointing different directions. The temperature does affect them, if the night was above 50F they 'wake up' and track within ~30 minutes. If the nigh was below 50F they take increasingly longer to 'wake up' and track the sun; sometimes over an hour. On overcast breezy days they hunt much more than if the wind is calm. Once again the colder the temperature the more sluggish and unstable they become.

I have heard that they are finicky about the balance, I have not seen that. I believe they have more than enough working fluid to balance out any normal amount of poor weight distribution. The way I see it the fluid is working off of the shading on the tanks, and enough fluid will be pushed to the appropriate side to get the shading balanced. For an example I have a 5' piece of angle iron that I use to force them horizontal during windy conditions. I have it hinged on one side so that it can be stored flat on the underside of the moving portion of the tracker when it is not needed. The swinging side of the brace is held on the tracker using a simple bailing wire hook(I do live in New Mexico...) The brace has come loose a few times so that it is swinging entirely on one side of the tracker(~3 lbs) and the tracker happily follows the sun.

Dave B, I don't intend to steal your thread, only to give some details about this style of tracker. Just let me know and I'll try to clear up this post and start a different thread.