Author Topic: Heliostat Project  (Read 33752 times)

Gabriel

  • Administrator
  • Hero Member
  • *****
  • Posts: 649
    • View Profile
Heliostat Project
« on: January 31, 2009, 05:35:16 PM »
Here is a heliostat that I have been working on over the past few weeks. So far it seems to be going pretty well. I just got it set up for the first time, and it seems to be working. I was only able to try it out for about an hour and a half before the sun went behind the trees because it was late in the day. I will have to set it up early next morning to get a good run out of it.

I have a couple of picture attached to this post. If you look at the first one, you can kind of see the sun in the upper left corner. You can also see the reflection from the mirror on the cooler.

For now, I only have a small mirror attached for testing, but I will be putting a much bigger one on later.

The heliostat is designed in such a way that I should be able to attach more of them together in a row. So it will only take two stepper motors to control several heliostats.

I will continue to update this thread periodically as I get more finished.

Thanks for reading!
Gabriel


d2day

  • New Member
  • *
  • Posts: 5
    • View Profile
Re: Heliostat Project
« Reply #1 on: February 13, 2009, 12:28:04 PM »
Nice job Gabriel...I am impressed with what you have accomplished.  :) :)

I have been looking into trying to track the sun in the same way you are doing it. I have a PRT Shopbot CNC, and am wondering if after I use the G code converter in Shopbot's control software to get your code converted, that the SBot control software would work for a sun tracker. I noticed that you have an interface worked out but that it requires Linix to run; I am trying to eliminate another learning curve. 

 I appreciate any help you may be able to offer, also not sure how SBot may react to use of their software for non Bot projects; even though it is for personal use.

digitize2day


Gabriel

  • Administrator
  • Hero Member
  • *****
  • Posts: 649
    • View Profile
Re: Heliostat Project
« Reply #2 on: February 13, 2009, 06:06:30 PM »
Yeah, I've been working on this project a lot lately trying to get it finished so I can start experimenting with it. I swear though, Murphy had this project in mind when he made his law.

I've redesigned the heliostat since I took the picture because, along with several other reasons, that version didn't seem stable enough.

I've gotten the new version put together and was able to test it out some earlier today. It's working fairly well, but I need to make some adjustments to the program because it's not quite keeping up with the sun like I'd hoped. It takes several hours before that starts to really be a problem though.


I tried out the Sbot control software that you mentioned to see if I could figure out how it might work. I've never used it before, but it seems pretty straight forward.

I had a problem converting the g-code to shop bot code. For some reason, it didn't convert the pauses correctly. What I did to fix it was change the program so that it would output shopbot code instead of G-code, which fixed the translation problem.

 When I run it in the shopbot program though, it doesn't stop to pause like it is supposed to. I think this is just because I don't have a shopbot plugged into my computer. I have attached the file I attempted to run for you to try. If it pauses for 60 seconds like it should, your half way there.

-- I also attached a test program which outputs the shopbot code for you to play around with if you want. Please note that this program doesn't track the sun like the one I have on my site but instead reflects it towards a single spot. If you get it to work and need the other one, let me know -- Program removed because it has been replaced by a newer version


The next thing you need to do is to see if there is a way to change either the gear ratio or maybe the lead screw's TPI in the shopbot program. If you can do that, you should be able to "trick" the program to make one inch instead represent one degree.  I tried to locate it myself but couldn't find it anywhere. Maybe you know where it is?

Using the shopbot program like this is pretty neat. You'll have to let my know if you can get it to work.

If you have any more questions, let me know.
Thanks
Gabriel
« Last Edit: April 19, 2009, 05:10:17 AM by Gabriel »

d2day

  • New Member
  • *
  • Posts: 5
    • View Profile
Re: Heliostat Project
« Reply #3 on: February 15, 2009, 11:09:44 AM »
Hey Gabriel,

The code looks good that you generated, I have run it in "preview mode" and I have the same indication you do. The preview mode is ignoring the pauses, I believe that is normal and you are right about the code needing to be run on a live machine. I appreciate the modification in the .rb file to convert the code to .sbp that will make the process much simpler.

Shopbot does have a way to "calibrate" the amount of movement, it is found in the "Values" under "UV". Shopbots indexer for rotary "stuff" can be set in degrees but not the xy or z axis. Not sure but it sounds like that "calibration" of the units that it outputs to the motors is the problem with your controller EMC2. I updated my system and had a very similar problem with it at that time. It would run a small part file with no issues but larger more complex part files lost steps. Shopbot eventually worked out the issue. SInce the total movement is so great the potential to loose "steps" becomes an issue. It may help to install bearings on all the movement points to help with binding. I plan to use a 8 foot sat dish and make my own mount like you have. I am not as far along with with my project as you are, I had struggled with the actual code to convert the sun position. (No code no need); I believe you have that part right and I am going to proceed on to the the actual building of the mount for my dish now.

Sorry I can't run a  file to check the entire code "movement" out because my machine bed is 4'x8' and some parts of the code would cause movement that would "crash" my machine. I have had the machine for a long time though and know the code will work on the controller. I will check to make sure it does the "pause" and let you know how it does.

I do need to get code that will track the sun...if that is not too much trouble.

Thank You very much for your help, I had almost given up being able to accomplish what you have done already. If there is any way I can help please e-mail me with your comments. I check e-mail a couple of times a day. ;D

d2day

Gabriel

  • Administrator
  • Hero Member
  • *****
  • Posts: 649
    • View Profile
Re: Heliostat Project
« Reply #4 on: February 15, 2009, 08:56:08 PM »
Hey d2day,

I'm glad that I've been able to help. I know that it was about eight months ago that I first started looking into this whole sun tracking business. I couldn't find what I was looking for, so I had to start from scratch, which was a lot of work. I decided to start this website so that other people wouldn't have to do the same. Knowing that I've helped at least one person makes it all seem worthwhile.  :)

You may be right about EMC2 losing steps. I'll have try and rig Mach3 to see if that fixes it. Although, I've used EMC2 before and I don't think that I was losing any steps, but it's hard to remember because it was so long ago and I didn't really know what I was doing at the time.

 I've actually changed the design of my heliostat again yesterday because the design I was using made it really difficult to calibrate EMC2 so that moving it one inch would equal exactly one degree, which also made it difficult to troubleshoot the problems I was having. With a heliostat, accurate movements are pretty crucial. It's not a big deal to be half a degree behind when you're pointing something directly at the sun, but ,when you're reflecting the sun to hit a spot say thirty feet away, half a degree makes a pretty big difference.  :o

Anyway, I've attached the changed ruby files. Everything seems to be in working order. I'm not sure how the shopbot software will like the pauses for the restofweek and restofyear because they are some pretty long numbers. (For example PAUSE 59.9339710834736) I ran it on my end though, and it seemed like it was working. Those two ruby files are still pretty experimental because I haven't had a chance to run them for more than a day with a sun tracking machine hooked up. You might just want to leave those alone for now until I can test them more thoroughly.

I know I've said this else where, but be careful with the sun tracking programs. There are a lot of equations in the code which have the potential to give an answer which ,mathematically, is technically correct , but completely wrong for our purposes. It's hard to catch all of these anomalies because the time is always changing and also because people live in different locations. I only bring that up because you are making solar concentrator AKA "Death Ray".  ;D


I hope everything goes well with this project.
When you get it finished, I most definitely would like to see a picture.

If you need anything else, just let me know.

Take Care
Gabriel

Attached Program has been removed because a newer better version is available.

« Last Edit: April 19, 2009, 05:13:35 AM by Gabriel »


d2day

  • New Member
  • *
  • Posts: 5
    • View Profile
Re: Heliostat Project
« Reply #5 on: February 16, 2009, 05:55:36 AM »
Thanks Gabriel.. I will post some info and pictures when I have them.

Gabriel

  • Administrator
  • Hero Member
  • *****
  • Posts: 649
    • View Profile
Re: Heliostat Project
« Reply #6 on: March 02, 2009, 04:58:43 PM »
I have completely rewritten the sun tracking program with much better calculations over the last week. It is for this reason that I haven't updated the site in awhile as I have been working on the program rather obsessively. It is a straight forward enough process, but the book I used as a reference, Astronomical Algorithms by Jean Meeus, is meant for astronomers, and I don't know much about astronomy. I got stuck in several places and had to do a lot of research on the subject to figure it all out.

Another thing that caused some confusion was the fact that the internet sites I used to double check the output  are WRONG. Arrrrrrrrrg!   ???  :)

Suffice it to say that it's a long story, but it all worked out.

The program isn't quite finished, but I figured that I would upload what I have so far just so I can say I've gotten something done today. I still want to add in compensation for light refraction in the atmosphere  and I haven't yet added in adjustment for the time it takes the stepper motors to move from sunrise to sunset yet. Plus, I want to do some more double checking.

And I need to write the documentation for it

And I still haven't gotten the heliostat working yet. (Turns out I've been trying to do it wrong the entire time. There isn't a straight forward way to do it with the current program. Also, ignore all picture of my previous designs. They are wrong!)

And I want to make a GUI for the program

And (one day) I want to make it possible for the program to run the stepper motors directly instead of through a machine controller.

And I want it up so it can point at different windows automatically when one room gets too hot.


Anyway, if anybody wants to take a peek at the new sun tracking program you can. It will output the G-code, but there is no adjustment for the time it takes the stepper motors to move back and forth, so the machine itself will slowly fall behind. It will probably take a day(/week??) or two before that happens though depending on the design of your sun tracking machine, so you can still try it out.

I'll try and get it finished soon, but I've fallen behind in other areas of life and need to do some catching up. It might be awhile.

Thanks!
Gabriel
« Last Edit: March 02, 2009, 05:03:46 PM by Gabriel »

d2day

  • New Member
  • *
  • Posts: 5
    • View Profile
Re: Heliostat Project
« Reply #7 on: April 18, 2009, 09:29:56 PM »
Hey Gabriel,

Sorry I haven't been on the forum for a while but I have been busy picking up the parts I needed to complete my project and working on the everyday things that get in the way of the fun stuff. I have most of the mechanics of the dish mount complete and have found a suitable power supply, the motor drives/gear boxes. I am still struggling with the final gear ratio but I believe that I will proceed on with my best "wag" at it.  More to come later with photos.

D2day

Gabriel

  • Administrator
  • Hero Member
  • *****
  • Posts: 649
    • View Profile
Re: Heliostat Project
« Reply #8 on: April 19, 2009, 05:06:24 AM »
Hey D2day what's up

Glad to here that you're making progress with your project.

I've attached the new sun program modified to output shopbot code.
If you have any trouble with it, let me know and I'll try to fix it.

Take care,
Gabriel

Davetech

  • New Member
  • *
  • Posts: 47
    • View Profile
    • Davetech's Projects
Re: Heliostat Project
« Reply #9 on: April 22, 2009, 11:52:53 PM »
Greetings Gentlemen,

I too am interested in this project. I have an electric tankless water heater which I wanted to preheat water for. I installed a holding tank and built and plumbed a "hot box" solar collector. Circumstances dictated that the best place to mount the collector was in an area that is shaded half the day; however, I have mirrors. The system is now working, with manual aiming of the mirrors.

I wanted to find a way to build a heliostat that would not require the use of a computer or even a microprocessor. I have seen a good sun tracker that uses green LED's as sensors and thought I could modify it to act as part of a heliostat.

I spent last night breadboarding a simple circuit that uses an LM339 dual comparitor and two potentiometers. I figured if the sun tracker turned one pot, the comparitor would be out of balance, turn on a motor and turn the mirror and the other pot until the pots were in balance again and the motor would turn off.  I hoped that with the proper gear ratios I would be able to keep the mirror reflection stationary.

I got it to work, sort of. But hysteresis in the circuit seemed to make it take more and more movement on the second pot to get them back in balance. I'm not even sure if the pots were linear or logarithmic, I just dug them out of the junkbox and saw they had the same resistance value.

By then it dawned on me that this was only going to be the first of many snags I would run into by trying to do it this way. I decided to go ahead and learn to use a picaxe.

I've done a bit of simple programming in Basic years ago, but my total experience with the picaxe and gcode consists of a few hours of reading here and there. I got on eBay tonight and ordered a board/kit from England which has a picaxe and motor H-bridge on it. So it will be a while before I can get my grubby hands on it.

So in the mean time I thought I would see if there was any gcode heliostat programs available on the net... I googled +gcode +heliostat and there wern't too many choices that popped up. But this forum was right at the top so here I am!

I'm surprised that there seems to be so many people building sun trackers (judging from youtube, instructables, etc) and yet so few tackling the heliostat. Anyway, this is my all consuming state of mind now and the thing that keeps me awake at night with all the little wheels turning, and I'll probably be checking in fairly often.

Davetech

Gabriel

  • Administrator
  • Hero Member
  • *****
  • Posts: 649
    • View Profile
Re: Heliostat Project
« Reply #10 on: April 23, 2009, 07:01:44 PM »
Hello Davetech and Welcome to the Cerebral Forum

Sounds like you've done quite a bit of work already. I'm not overly familiar with electronics, but I think I understand what you were trying to do with the LED's. From what I've read on the net, it is indeed difficult to get it to work correctly with that method.

I've also been interested in setting up a microcontroller to take the place of the computer simply because it uses a lot less power. That, and it can be placed directly on the heliostat. Right now though, I've been trying to get a larger heliostat design put together first so that I will have a decent amount of sun power to play around with.

I have attached the heliostat program. I haven't actually released it on the site yet because I haven't had much of a chance to test it, but you are welcome to check it out. Do be careful though.
UPDATE: Attached Program has been removed as it is now out of date. Follow the link for the most recent version. http://www.cerebralmeltdown.com/projects/heliostatprogram/default.htm

So, are you going to try and translate it into Basic for the picaxe, or were you just looking for this out of curiosity?
There is actually a heliostat program that I found fairly that is already in Basic. The link is http://www.green-life-innovators.org/tiki-index.php?page=Sunalign

I haven't tried it myself, but it might work well for you.

Anyway, I have to go. Which is probably a good thing because I have a tendency to ramble.  :)

Keep up the good work
Gabriel
« Last Edit: May 24, 2009, 03:38:22 PM by Gabriel »

Davetech

  • New Member
  • *
  • Posts: 47
    • View Profile
    • Davetech's Projects
Re: Heliostat Project
« Reply #11 on: May 07, 2009, 09:05:31 AM »
Well, I've made some progress.  I built a mock-up sun tracker that uses LED's as sensors. I used gears and a dc motor out of a junked printer and breadboarded a control circuit using a single BA6209. it is capable of driving the motor either direction depending on which pin goes high. It will follow my LED flashlight around the room and will track the sun. Bright sunlight has enough ambient light reflected that it makes the motor chatter all the time, so I'll either solve that with a hood or might need to interface it with a component with a threshold. I think if the sensors are mounted on the mirror assembly, it should work fine but I'd need a second one to control the Z axis. Getting tons of rain down here now though.

My Picaxe 18x w/ motor control kit arrived yesterday. I've assembled it, built a diy serial download cable, programmed it and have it working flashing LED's. Now it is just a matter of interfacing it to do something useful. Myriads of possibilities! 

I've downloaded PCB Artist (freeware) and gotten used to using it. Its autorouting is a great timesaver! Laid out and etched a board with two PBL3717A's on it that I think will control the X,Y and Z axis. The Picaxe will control this board. The RKeducation board came with an L293D on it but this board I made will be beefier I think.

So I'm making progress on two fronts.

Gabriel

  • Administrator
  • Hero Member
  • *****
  • Posts: 649
    • View Profile
Re: Heliostat Project
« Reply #12 on: May 07, 2009, 07:24:34 PM »
Yeah, it's raining here too, and it is most certainly slowing things down as far as building heliostats goes.

Sounds like you're definitely making good progress. When you get it finished, you should put it on your site to show it off. (Assuming that you haven't already and I just missed it.)



Gabriel

Davetech

  • New Member
  • *
  • Posts: 47
    • View Profile
    • Davetech's Projects
Re: Heliostat Project
« Reply #13 on: May 10, 2009, 06:25:38 PM »
Although I've been successful at driving stepper motors with the picaxe, the KISS principle dictates that I just use regular 12vdc motors for this (especially since I have a box full of them from junked VCR's). I also have several quad pole, double throw relays (military grade!). A single one of these relays can reverse the directions of the motor using only a single output from the picaxe if ya know how to wire it.

Its gotten down to mechanical construction now and I have gathered volunteers from my junk box. The pics are of dead VCR gears, motor, head cylinder assembly (good bearing) and a threaded rod with 2 nuts brazed to a straight copper fitting. The flat belt and gears will slow down the rpm's and the threaded rod assembly will be my linear actuator.






Now I need to make a housing to mount the stuff on and to make bearing ends for the threaded rod (after I remove the end threads). I decided to fire up the foundry and make some aluminum/zinc alloy sheet metal. I've done it in the past when I poured molten aluminum on a thick steel I-beam. Then I just cut it into sheets. Thin but strong.  Well, 2 attempts and 2 failures today. The first time the metal bubbled. I had forgotten to heat the steel up to remove any trace of moisture from its surface. The second pour resulted in too thick a puddle, which cracked upon cooling due to shrinkage. Then I remembered that I should have tilted the I-beam a bit so the metal would run down it, producing a thinner sheet.  Lessons learned.  Well, I had gotten enough smoke in my eyes for today and decided to leave attempt number 3 for another day.



Gabriel

  • Administrator
  • Hero Member
  • *****
  • Posts: 649
    • View Profile
Re: Heliostat Project
« Reply #14 on: May 19, 2009, 08:10:49 PM »
Wow, you're really moving along. I'm going to have to raid a few of the old/broken VCRs that I have lying around the next time I need some DC motors.

I also like how you attached the threaded rod, which is something that I can rarely get right.

And sorry it took me a couple of days to reply. I had some work I needed to take care of.

Thanks for the pictures and I can't wait to see what comes next!
Gabriel