Solar Tracker Project

[Xringer]

Works almost the same as before! The good news, it didn't try to track the moon last night!
I think it likes to track the outdoor IR security lights a little.

Today, it got 15-20 degrees ahead, again! (Before the sun clears the easterly trees, it tracks to the brightest part of the sky, and waits).

At 10:30, when the sun hit the array, it came back east, but not all the way! It was tracking about 20 degrees ahead of the sun..

I know that reflections don't help, so I painted over the white center plate (used RedRock paint).



That cut the glare some, but didn't help with 20 degree problem.

I moved it dead-on, and it locked up just fine. With no visible lead or lag.

[Dave]

Thanks for the update. I know what you mean about the moon. I think it's been the brightest I've seen all summer. Today is a very bright sunny day down here in N.J. I'm about to order the redrock this weekend, just so I can start tinkering with it. Keep us posted.

[Xringer]

Originally Posted by Dave

Thanks for the update. I know what you mean about the moon. I think it's been the brightest I've seen all summer. Today is a very bright sunny day down here in N.J. I'm about to order the redrock this weekend, just so I can start tinkering with it. Keep us posted.

I was going to get the kit, but reconsidered.. The extra $10 was worth it. It's a great little board.
I manually tweaked the array onto the sun today and it tracked perfectly the rest of the day.
I never saw any shadow at all on the shadow stick. It was dead-on perfect.
The other board was never this accurate.

If I didn't have those giant trees to the east, this thing would perfect.
Since I don't want to wait for fall, I'm going to have to figure out
some way of starting up on a good track in the morning. Get a good lock, right off the bat.

Since the sun doesn't clear the trees until about 10:30, I've been thinking
about using a timer to disable the RedRock's power supply from 21:00 until 10:30 the next morning..

Another idea is to give the RedRock some sun-glasses. . In the form of a tinted sun-shield,
that will cut down the amount of voltage produced by the LEDs.
That might keep it from seeking out the brightest part of the sky, at the crack of dawn..

One other thing on my list.. Get rid of the aluminum shield around the glass.
I suddenly realized today, that thing is a snow & ice catcher. Don't need it.

Re: Motor too fast.
I'm using a 24v actuator with 12v and it's still not slow enough,
or there is some slop in my mount that allows some over-shoot.
When it's tracking well, many times, the motor over-shoots the sun
and the RedRock slams the motor into the reverse for a quick backwards movement.
(Since the board is firing short pluses into the motor, the reverse is actually pretty smooth).

RedRock sells an Anti-reverse feature, which you might want to consider,
if you think your actuator & linkage arm might be a little fast..
The reason I'm worried about those little quick reverses?
They might not be too good for the actuator motor and gears..

I had a similar problem once with my Ham radio satellite tracking rig,
and fixed the too-fast motors by installing large series resistors.
Slowed the motors right down to 1/4 speed.. Might check this idea out tomorrow!

8-29-2010 update:
Yesterday, I put 5 ohms in series with the motor. This slow it down a lot, and improved
the reverse problem quite a bit. It's reversing a lot less now and not going as far backwards.

It's still tracking slightly ahead of the sun. Once manually 'aimed' it stays locked..

[Dave]

Hi,
What would increasing the voltage to 24 do to the performance?
I think I will go for the anti-reverse unit. Thanks for the tip. I did buy the same actuator but the 36v model. I'm still a long way off so I'm following your progress regularly. Thanks for your updates so far.

[Xringer]

It would be way too fast with 24v.. Be too jerky..

I think your 36v actuator might be perfect for a 12v supply. Give you a nice slow array movement.
It's not like the sun is zipping across the sky like a low orbit satellite.. No speed required.

Have you got your mount started yet?

[dakicha]

For me the best two axis solar tracking system to be applied to any mechanical system we have is the SolarElectronica Company.
Works form 10 to 36v with any dc motor. You can see details solarelectronica.com
and best of all, very affordable.
That works great. I bought the unit, some time ago and works with incredible accuracy regardless of my engines are the worst!
I use CPV (concentration modules)

[Xringer]

I like the wireless remote, with LCD! I could use one today.
It's too hot (for me anyways) to go out and manually tweak the array dead-on..



It looks very professionally made, like something from an R&D lab pro-shop..
If it was priced like the RedRock units, there would be a lot of interest.

Anyways, I don't think super precision tracking is going to gain much,
over regular old tracking, when you are using PV panels.
Looking at my amp meter, it seems like 15-20 degrees off-sun,
doesn't make much difference in power.
Of course, my measurements are too coarse to show +/- 2 or 3 percent.

[Xringer]

I don't feel like the Solar is heating as well. Most likely because the heat is being applied to the colder lower part of the tank..
But, if a lot of the heat is being radiated off the load into the air, some of it is escaping up the flue.
Maybe a little too much.?.

So, to enhance contact heat transfer from the hot plate (10 ohm load),


I placed a cloth under the hot-plate, soaked in a liquid heat-transfer solution.(10w30 motor oil).

I've also added some more insulation to the boiler. It still needs more. On my list for next week.

The Hurricane is over, so maybe today we can see if there is any improvement..

Cheers,
Rich

[Xringer]

Yesterday began as usual, with Miss RedRock moving the array towards the brightest part of the sky,
while the morning sun was still behind the westward trees!

After 10:30, the sun had cleared the trees, and sure enough,
the array was tracking about 20-30 degrees ahead of the sun!!

Since it would stay ahead all day, I've been going out and manually
aiming it right at the sun. It tracks fine after that.


When the sun finally hits it, are the sensors overloading? Are the sensors (LEDs) out of balance?
It seems like the 'Night Return East' offset bias would help keep
the tracker pointed east as the sun was coming up behind the trees.
But no! It sees that bight sky and thinks it's the sun behind clouds..

~~~

That's when I tried to imagine what the sun looked like from inside a covered wagon..
Looking up from inside a covered wagon, the light coming in would be defuse,
seemingly at equal magnitude from the whole sky.. A nearby tree line might not be noticeable.?.
BUT, if you looked in the direction of a bight sun, it would be pretty bright..
Maybe just bright enough for Miss RedRock to track.?.

So, I wrapped a white cotton cloth around the RedRock module.

Within seconds, it was moving towards the sun (in the eastern sky)!
All day long, it stayed right on the sun within a few degrees!!

This morning(9:30AM) before the sun is over the trees, it's not tracking towards the bright sky!

I'll try to be watching at 10:30 (at First Light)!

[redrok]

Hi Xringer and others;

First:
It seems that you are not understanding what is going on when the LED3X tracker is not aimed directly at the sun.

The LED sensors are integrating the light from a wide area of the sky. This can cause the tracker to aim quite a way away from the sun. This is usually a good thing.

Remember, only 80% of the light is coming directly from the Sun. The 20% is from the sky. OK, the closer to the sun the brighter the sky radiation.

To evaluate this you must measure the power coming from the panel at different orientations. You will probably find the power is close to greatest where the tracker is aimed and not directly at the sun.

When setting the orientation of the tracker choose a clear sky day when the sun is clear of any obstructions. At this time both the direct and the sky radiation should be centered on the sun.

Your example of the tracker leading the sun just as the sun clears the trees to the east sounds correct to me. Even if the sun is 10deg off axis the cosine loss is only about 1.5%. But the added power coming from the sky radiation is probably greater.

The only way to tell is by direct power measurements.

Second:
If the motor is doing a bit of hunting you might want to reduce the motor on time setting. It's set to about 2 seconds now. The tiny square black surface mount pot near the Eastern LED sensor controls this. It looks like a Phillips but it is really a small flat blade screwdriver.

I don't recommend adding a resistor in series with the motor. True, the speed is reduced but the torque is severely limited which is required for high load conditions such as wind, snow, or when the mount is near the movement extremes.

This is why I can't put linear actuators in the "Good" category, just "OK". Sure, they are well built and have built in limit switches but they are linear actuators essentially connected to a lever. The speed of movement changes throughout the day.

"Good" motor drives are rotary with high gear ratios of 100,000/1 or more. The motor can be quite small yet will produce quite high torques. The drive can be V-belts, cables, or chains. Note, V-belt pulleys or chain sprockets are not required when the belt or chain is bolted to the opposite side of the sheave.

Duane
redrok(dot)redrok.com