Author Topic: Heliostat Array Project  (Read 12085 times)

Gabriel

  • Administrator
  • Hero Member
  • *****
  • Posts: 648
    • View Profile
Re: Heliostat Array Project
« Reply #15 on: August 17, 2009, 06:36:14 AM »
Hey Brendan,

I abandoned the gimbal design because, after two months of working with it,  I couldn't get that infernal machine to work. ???

I still don't know what was wrong either. I must have checked everything 100 times, and it all seemed fine. It was the strangest thing. It should have worked, but it didn't.

At first, I thought that it was the heliostat program I wrote that was messed up, but I tried out a little test heliostat with it and it did fine.

I decided to just start over with a new machine and see if the problem would disappear, which it did. I guess that the gimbal machine I built might have been faulty, but I still don't know exactly what the problem was.

In all practicality, a heliostat's alignment doesn't have to be super precise, but, for the gimbal machine that I was using at least, setting it up was a bit of a pain because there were two posts (one on each side) involved. If you're only aligning one machine, plumbing each post isn't a big deal, but I wanted to set up three that were controlled with just two stepper motors. I never actually tried it, but it seemed as though trying to get six posts set up correctly would be tedious.

I have found the heliostat design that I'm using now to be much quicker to put together. I have also found that the weight of the mirror can easily be overcome with a counter balance, so this design does actually have all of the benefits of a gimbal design.

I hope that clarifies why I made the switch.

Have Fun
Gabriel


Gabriel

  • Administrator
  • Hero Member
  • *****
  • Posts: 648
    • View Profile
Re: Heliostat Array Project
« Reply #16 on: August 28, 2009, 06:57:59 AM »
I finally had a chance to upload a video of my heliostat array.

Time Lapse Video of a Heliostat Array Prototype Test


The tracking definitely isn't perfect, but it's good enough to give me the chance to do some experiments with it.

You can't see it in the video, but the reflection from the mirror on the left moves up and eventually misses the target almost completely. At the end, it is actually hovering just above the wood target.

The reflection from the right mirror does the opposite and moves down throughout the day.

In some cases, this might not really matter. For example, if you had a sliding glass door there wouldn't be any issue because it is a big target and easy to hit.

I know that I could probably make some improvements to the tracking, but it takes a long time to do it because I have to run it for the entire day to see whether or not any adjustments I make have an affect. The fact that it has been overcast lately hasn't helped either.

It's good enough for now though, so I'm just going to live with it.


Davetech

  • New Member
  • *
  • Posts: 47
    • View Profile
    • Davetech's Projects
Re: Heliostat Array Project
« Reply #17 on: August 28, 2009, 09:29:30 PM »
Great job, Gabriel!

I was thinking... about testing your unit out... if you mounted a sight tube on each heliostat, you could rapidly step the mirrors through a day's movement and check the alignment through the tubes?

Have you tested a piece of vinyl siding with it yet?

Getting rain down here lately too.

I put my ramp idea on hold. I'll tinker with it later. Right now I just want to get the thing working, so I'm going to go ahead and add a tilt motor and the electronics for it. Ran into a snag though. Turns out that the two wireless doorbells I got DO work on the same frequency and will interfere with each other. Thought I had checked that out. So I've wasted two days playing with other ways to get the signal wirelessly delivered. Fooled around with a couple of old wireless computer mice but I couldn't get more than about 10 feet of range out of them. Now I'm bidding on a toy helicopter that is controlled by a two channel IR system supposed to be good up to 50 feet. Also got a cheap Bluetooth dongle and a headset, but I've got no experience with Bluetooth and don't know if I can get them to communicate without software/computer.
« Last Edit: August 28, 2009, 09:36:47 PM by Davetech »

Gabriel

  • Administrator
  • Hero Member
  • *****
  • Posts: 648
    • View Profile
Re: Heliostat Array Project
« Reply #18 on: August 29, 2009, 05:20:10 AM »

I'm already pretty certain that the heliostats are aligned correctly. I can check the altitude of each mirror pretty easily, and they are all right where they are supposed to be. I think most of the trouble is caused by the fact that the left and right mirrors are turned inward to focus the light and are therefore off target, although I've been wrong before.

Right now, I'm just going to experiment with it for awhile and make adjustments as I go.

I haven't tested it with a piece of vinyl siding yet. I did measure the temperature of the target though, and it was at around 150F.

I've looked into bluetooth myself for an idea unrelated to heliostats. There doesn't seem to be a whole lot of information on the subject.

I've also looked into those XBee wireless modem modules because I've seen a lot of people using them successfully. They look interesting, and I've heard that you can get them for pretty cheap off of EBay. One day I want to try one out.



travis77

  • Jr. Member
  • **
  • Posts: 75
    • View Profile
Re: Heliostat Array Project
« Reply #19 on: September 08, 2009, 03:26:18 PM »
Someone's been busy updating the site! That write up looks really nice, i bet it took some time. Almost makes me want to build one haha, but I dont know what I'd do with it, nor do i have extra steppers laying around.


Gabriel

  • Administrator
  • Hero Member
  • *****
  • Posts: 648
    • View Profile
Re: Heliostat Array Project
« Reply #20 on: September 09, 2009, 01:38:22 PM »
Yes it definitely took some time. I guess after having worked on the design and just heliostats in general for so long I had a lot to say.  ;D

Now, I have some new ideas, and I want to see if its possible to control multiple sets of stepper motors with just one driver board. So, a two axis driver board could be used to control as many stepper motors as you would need just so long as you don't mind them taking turns when moving. I don't know a whole lot about electronics, but I think that it is possible.

This would bring the cost of the electronics down a lot.

Sometime soon, I'm going to start a new thread explaining my idea in detail in hopes that someone might be able to double check the circuit design (which I haven't made yet) that I would need to pull this off. Mostly I just want to be sure that I'm not going to fry anything.


travis77

  • Jr. Member
  • **
  • Posts: 75
    • View Profile
Re: Heliostat Array Project
« Reply #21 on: September 09, 2009, 09:40:53 PM »
Okay, talking about running more than one motor on a single axis on a driver board, I think you can just wire the motors in parallel.... I thought I saw some people talking about this over on CNC zone where someone was running two motors for his x axis on his moving gantry. I'll try to find a post, I think I have a picture.

Edit: I found a good post: Multiple Stepper Motor Wiring

Read that. Well he gets pretty heavy into the math behind parallel wiring, which looked like he spent way to much time explaining it, maybe even over explaining making it all pretty looking. Anyway, what is important is to stay within your driver board's current and voltage limits. They were initially talking about wiring them in parallel, but then at the very end they were talking about wiring them in series because it would draw less current than if wired in parallel (resistance is less in parallel, thus draws more current). Anyway check that post out... So it is possible to do it. Just have to figure out what your max current and voltage your driver board will handle. Oh ya, and its easier if you use all of the same steppers so you dont have to worry about one drawing more current than the other... Hope that helps out.

By the way the guy on the third post has a pretty sweet rep rap... hey Gabriel, you should build one so i dont have to! :) Maybe one day I will... I love how RepRap's dont have any mess or cut away material like with CNC's. They only use the material they need to make the part, hardly any stock material gets wasted.
« Last Edit: September 09, 2009, 10:14:04 PM by travis77 »

Davetech

  • New Member
  • *
  • Posts: 47
    • View Profile
    • Davetech's Projects
Re: Heliostat Array Project
« Reply #22 on: September 10, 2009, 05:54:27 AM »
Wow Gabriel, that's good outside the box thinking!  The place a micro shines is its versatility and the nice thing about heliostats is that the timing of the adjustments is not critical. Doesn't matter if the mirrors are adjusted in a sequence and a few relays (or electronic equivalent) could be used by the micro to address each each mirror in the array individually. That would cut out the problems of linking the mirrors mechanically and let software do the work.

I think that idea deserves much consideration. Meanwhile, I'm still stuck just getting one mirror working correctly. Although my cnc will plot the control pc board perfectly with a pencil, when I try to mill a pcb, even with a brand new bit, there is enough sideways pull on the z-stage to slightly move it and skew the cut. And it does it even at 2 inches per minute travel. I guess I'm going to have to rebuild the stage stronger and it is ticking me off.

travis77

  • Jr. Member
  • **
  • Posts: 75
    • View Profile
Re: Heliostat Array Project
« Reply #23 on: September 10, 2009, 08:22:07 AM »
Davetech, have you tried making multiple passes to reach your target depth instead of just making one pass?

redrok

  • New Member
  • *
  • Posts: 3
    • View Profile
    • Red Rock Energy
Re: Heliostat Array Project
« Reply #24 on: September 15, 2009, 06:17:45 AM »


Hi All;

You asked about using heliostats in arrays.

HELIOSTAT ARRAY MOVEMENT:
The subject of "single point tracking" of heliostat
arrays is a bit difficult to explain properly but I'll
give it a try.

On a hard technical basis. All mirrors in a field of
heliostats move in exact synchrony. I.e. you only need
to track one mirror. All others then "move" in exactly
the same angular amount in 2 axes.

The placement of the heliostats is completely arbitrary
in 3 spacial dimensions. Of course, some locations have
advantages over others. Generally they are mostly placed
on the northern side of the target in the northern
hemisphere.

OK, there are some preconditions:
1. All the heliostat mounts must be initially installed
   so the axes are all accurately aligned parallel to
   each other.

2. Each mirror should be circular but a square is close
   enough. If not circular there is an error called
   "Sagittal Error" and is maximized with Oval or
   rectangular mirrors. The effect is the shape of the
   reflected image is similar to the shape of the
   mirror but it rotates and distorts throughout the
   solar day.

3. All linkages, or the electronic equivalent, must be
   made similar to the parallelogram mechanism so the
   mounts all move synchronously. (Stepper or servo
   motors are useful here.)

4. Before any synchronous movement each mirror must be
   initially preset so the sunlight is reflected to the
   target.
   Note! This initial mirror setting can be used to
   aim some of the members of the array at different
   targets. These different targets can be placed
   completely arbitrary in 3 spacial dimensions also.
   Of course, some locations have advantages over
   others.

This works because after the initial mirror adjustment
the basic job is to bisect the angle between the sun
and the target. For any angle movement of the sun the
mirror will move exactly half this movement. Since the
movement of the sun observed by each mirror is the same
for them all they all move the same amount.

This is best done by having essentially a mounting base
that contains the movement mechanisms. The mirrors are
then mounted on top of the bases and have mechanical
adjustments to aim them initially at the desired targets.
The idea is that the bases have the synchronous movements
with all the redirection in the mirror mounting.

The true bisector angle can be broken down to a pair
of movements. One for each axis.

There is another heliostat array, I call them
"Arc Heliostats", that works with a single point tracker.
This one is obscure and I don't think any were built.
See my patent page and do a browser search on this page
for "Arc Heliostat" to see some of the patents. See:
http://www.redrok.com/neat.htm

In summary, single point tracking for heliostat arrays
are possible. In fact it's the norm, even on the large
arrays. Sure there is a computer that drives the stepper
or servo motors but just as described above only one
tracker function needs to be done for the whole system.

Duane
                                      Receiver
  Powered by             \  \     \        |*|
 Thermonuclear   Solar Energy from the Sun / |
Energy (the SUN)           \  \     \  /  /  |
Red Rock Energy             \  \     /   /   |
Duane C. Johnson   Designer  \  \  /  \ /    |
1825 Florence St  Heliostat,Control,& Mounts |
White Bear Lake, Minnesota    === \   / \    |
USA    55110-3364                  ===   \   |
(651)426-4766        use Courier New Font \  |
redrok@redrok.com     (my email: address)  \ |
http://www.redrok.com  (Web site)          ===


redrok

  • New Member
  • *
  • Posts: 3
    • View Profile
    • Red Rock Energy
Re: Heliostat Array Project
« Reply #25 on: September 15, 2009, 08:44:20 AM »
Hi Gabriel;

Your heliostats are in a straight line which is a
simplification from the arbitrarily placed mirrors.

I think I see in your pictures that the AZimuth bars
are not exactly parallel. They should be, so the
AZimuth angular movements are exactly equal.

Then, the Azimuth initial mirror setting can be made.

Now the problem:
Since each mirror is mounted directly to the hinge
each hinge axis is aimed in a different direction.
When the mirrors tilt in ALTitude this causes the top
of the mirrors to move laterally different amounts.
The greater the tilt the greater the difference. This
causes the length of each tilting cable to be different.

The solution:
1. Make the parallelogram exactly parallel.
2. Make the mirror hinge axes exactly parallel. To
   do this, substitute plywood for the mirrors then
   mount the mirrors on the plywood with a means
   to redirect the reflected light to the target.
   Here is one possible method. See William Beaty:
   http://www.redrok.com/main.htm#beaty
   And many other methods.

Yes, this is one more bit of complication but in
the long run its really a simplification.
(Besides, if these things are not done each mirror
 must be independently tracked.)

Its easy to adjust things to be exactly parallel,
strait, perpendicular, in alignment, or horizontal.
Each of these adjustments are made independently.
Forget about the reflected light until the final
mirror redirection adjustment.

Note! The heliostats don't actually have to be in a
straight line. The only requirement is the mirrors,
parallelogram mechanism, and targets be on in the
same plane.

Duane
                                      Receiver
  Powered by             \  \     \        |*|
 Thermonuclear   Solar Energy from the Sun / |
Energy (the SUN)           \  \     \  /  /  |
Red Rock Energy             \  \     /   /   |
Duane C. Johnson   Designer  \  \  /  \ /    |
1825 Florence St  Heliostat,Control,& Mounts |
White Bear Lake, Minnesota    === \   / \    |
USA    55110-3364                  ===   \   |
(651)426-4766        use Courier New Font \  |
redrok@redrok.com     (my email: address)  \ |
http://www.redrok.com  (Web site)          ===

Gabriel

  • Administrator
  • Hero Member
  • *****
  • Posts: 648
    • View Profile
Re: Heliostat Array Project
« Reply #26 on: September 15, 2009, 09:55:16 AM »
Hello RedRok,

I guess you got my email. :) Thanks for coming over and checking everything out.

I think that I have everything set up on my heliostat array exactly the way you said. It's working fairly well, but still not perfectly. I guess whatever is causing the drift has nothing to do with the fact that the mirrors on the left and right are angled since you are saying that it should work.

Something else must be messed up.

Right now, I think I'm just going to give up on trying to work out the bugs involved with linking the mirrors together and instead skip to controlling the mirrors separately, each with their own set of stepper motors. That way, I can change the targets for the mirrors via the software instead of having to go out and manually readjust them.

That's pretty much been my ultimate goal all along anyway.

My next goal is to see if its possible to control multiple heliostats with just a single two axis driver board. This should help bring down the cost of the electronics. Basically, what I want to do is to use MOSFETs to turn the different sets of stepper motors on and off when the computer tells them to.

So, if I had three heliostats, that would be three pairs of stepper motors controlled with just one 2-axis driver board.

If it's time to move heliostat #1, a MOSFET would be told by the computer to allow power to flow from the driver board to the first pair of stepper motors.

If it's time to move heliostat #2, a different MOSFET would be told by the computer to allow power to flow from the driver board to the second pair of stepper motors.

It would be the same process for heliostat #3 and I suppose several more heliostats if need be.


I don't know a whole lot about electronics, so there might be a flaw in my plan. That is my general idea though. Do you, or anybody else for that matter, know of any similar examples that someone else might have tried?

The one problem that I see with it so far is that several MOSFETs might accidentally turn on all at once allowing the current to flow through several pairs of stepper motors which would probably ruin the driver board. The easiest way that I can see for controlling the MOSFETs is through the computer's parallel port. The parallel port tends to have a mind of its own though, so it is quite probable that this behavior could be what ends up ruining the driver board. I guess there would need to be some sort of protection against this sort of thing built into the circuit.

Another idea I had would be to use an Arduino or another similar micro processor. The computer would first send the instructions to the Arduino via a serial connection and then the Arduino would be what actually controls the MOSFETs and driver board. That way, any random signals sent out by the computer, when it's turning off or on for example, won't have any effect on the heliostats.

The Arduino also has more outputs than a parallel port, which means that it would be possible to control more heliostats. If I do end up using one, I want to see if its possible to control the heliostats entirely with the Arduino. That way, I don't need to use a power hungry desktop or laptop all the time. I'd only need the computer if I wanted to change targets. It also should be pretty easy to set up a wireless connection with the Arduino so that I can reduce the number of wires that go out to the heliostats.

So yeah, I have ideas.  ;D
I haven't had much time to work on it over the last few days, so obviously I needed to vent off some of my thoughts. Any suggestions from anybody with ideas of their own would be welcome. I'm basically just making it up as I go along at this point because I haven't a clue what I'm doing.  :D

Gabriel

  • Administrator
  • Hero Member
  • *****
  • Posts: 648
    • View Profile
Re: Heliostat Array Project
« Reply #27 on: September 15, 2009, 10:14:38 AM »
Whoops! you posted while I was writing my own post.

I'm not sure which pictures you're referring to, but I think the one you are looking at is the one where I have just two mirrors set up. Those most definitely didn't work. They were mostly just a quick trial to figure out how I might be able to connect multiple mirrors with just one set of stepper motors.

If you're looking at the pictures where I have three mirrors connected together, they might look off because the mirror in the middle was accidentally placed about an inch forward from where it was supposed to be.

For that heliostat array, I did make sure to align everything correctly. Something else is probably messed up though. It could be that the stepper motors are losing steps or a pulley is misaligned or something similar. I don't know. It takes such a long time to test these things. You only really have 2 or 3 tries per day before it gets dark. That, and clouds don't help one bit. :)

Thanks!
Gabriel