Home › Blogs › Dockbot's blog

Sun Tracker/Solar refrigeration

Tue, 06/09/2009 - 02:46 — Dockbot

I dont' know how much feed back I'll get since I'm an inexperienced blogger but, I'm posting this in the hopes that the people on dorkbot who are more experienced than myself (that would be everyone) will review and critique this idea. In other words am I wasting my time?

I'm planning to make a solar deep-freeze that would not be grid-tied. This project is in the very early stages but, I plan on using an Arduino to track the sun to point a parabolic trough with a fresnel lens over it heating a pipe full of calcium chloride and pure ammonia. This array would also have PV panels to charge the servos' batteries. (I'll have to distill the ammonia. Fun.) Heating this pipe forces the ammonia out of solution where it will be condensed by a liquid to liquid heat exchanger and be gravity fed to a holding cylinder in a deep freeze. At this point the trough would roll 180 degrees away from the sun to reflect radiant heat. The pipe will then cool off which would allow the internal pressure to drop and the ammonia to recombine with the calcium chloride pulling heat away at the same time. I am looking at additional ways to cool the rotating/sun tracking pipe between cycles. Maybe the liquid to liquid heat exchanger or a separate one with a pump. I guess the heat pipe would need a water jacket. I'm not worried about that part.

By the way these ice makers are nothing new. They already exist but, they only cycle once a day with the sun. I am simply trying to make one that cycles several times a day. So it could be much smaller. Here's an example of a single cycle ice maker:

http://seasteading.org/seastead.org/localres/misc-articles/solarice.pdf

After this idea is torn apart properly I will continue to post details. For now I am mostly looking for a sanity check. I think this would be a really fun and useful project.

Thanks for reading! Hope to hear from you soon.

Dan


AttachmentSize
solarice.pdf346.54 KB

Comments

Sun, 06/14/2009 - 20:06 — mark gross

I've been thinking about

I've been thinking about doing a sun-tracker-logger type of thing to measure the motion of the sun from day to day using some a PCB I made for a robot that includes a dual H-bridge and a Tinsy++.... Perhaps we could reuse some of these parts to help make your thing.


Mon, 06/15/2009 - 04:20 — Dockbot

Hello Mark and thanks for

Hello Mark and thanks for the reply. I should have mentioned that I won't actually be in Portland until next fall. I'm very excited to get there. On the electronics side, I am starting from 'hello world', so any help would be greatly appreciated! I will gladly replace any parts that would be used for the project.

I was thinking of doing something along these lines. One TSL230R light sensor in the center of the array and one more extending in each horizontal axis of the array. Then write a program that controls 2 servos via an Arduino to keep all three frequencies equal? I'm sure there is a better way. I just wouldn't know yet. I'm still gathering the parts for the evacuated tube and designing the absorbent pipe/water jacket.

Let me know if I'm off base. By the way that's really cool that you just have extra PCB's laying around that you made.

Thanks!

Dan


Mon, 06/15/2009 - 05:41 — Dave_Riness

analog comparator & 2 solar cells

Hey Dan -

Thanks for posting the link, learned something new about solar cooling today! I built a simple sun tracker years ago by using two solar cells mounted at 90 degrees to one another (think two triangles back to back - the edge of the joint would be parallel to the solar collection pipe) - the edge where both cells meet is pointed towards the sun - when the output of the solar cells is equal, you are pointed directly at the sun. Whichever cell has the lowest voltage is the one that is pointed "away" from the sun - you can use an op amp circuit to create a simple analog computer - the output drives the motor (forward or reverse) depending upon which solar cell (A or B) needs to be pointed towards the sun. You could also use an analog to digital converter on an Atmel microcontroller - either way works fine...


Mon, 06/15/2009 - 19:59 — Dockbot

Hey Dave! I'm already

Hey Dave! I'm already learning a lot from these responses. Was your tracker a single or dual axis tracker? I'm assuming that for a dual axis tracker I would need more than 2 photo diodes. http://www.sparkfun.com/commerce/product_info.php?products_id=8940 Or, could the rest could be resolved with programming? I'm planning on using an Arduino even if that's not the most pragmatic approach because this is forcing me to learn C++ (this seems a little like sorcery for a mechanical guy like me) and I may decide to include, some failure detection, and an E-stop with some kind of ability for the system to alert me. Maybe twitter and WiFi. I guess that would be a lot of code. But, it seems like most of that has been done and I could probably get away with compiling and editing existing code. If I need more memory I guess I'll pick up a Sanguino. I suppose I should learn to build my own! If this project ends up lacking robustness or the possibility for single point failures is too great I'll refer to this post again. This is going to be a drawn out project but, I'll post any updates and add photos as they happen. I'll come up with a better way to document everything and post it when the project is complete.

Really appreciate the input,

Dan

Sun, 07/05/2009 - 01:05 — shah

solar tracker

I want to built solar sun tracker as you have already did, I am sure that must be working perfectly, can you send me schematic and some photogrpah,

thanks

Shah alternateenergi@yahoo.com


Thu, 07/09/2009 - 23:48 — Dockbot

Hey Shah, I haven't built a

Hey Shah,

I haven't built a sun tracker yet but, I will post photos as I hit each new milestone. The bad news is that I won't have the heat exchanger complete until October so, please be patient and I'll try to keep everyone informed of what worked and what didn't. Also, as I post comments or photos let me know if you think you know of anything that we could improve. It's amazing how often another set of eyes can spot things I've missed.

Dan


Thu, 06/18/2009 - 20:40 — tap3w0rm

Sun!

My imagination floats to the following idea while thinking about your tracker. I have set something up like this in the past. It was kind of a pain but totality worth it to watch it work.

Get yourself 5 panels and create a cross like this: ignore the 0's i needed them to space out the diagram correctly

010

234

050

from the side This is a cross section it should look like this. Quite a bit tilted for 1 5, and 2,4.

/ - 1

| - 3

\ - 5

Now as the user above said you monitor the panels 1 2 4 5 in 2 sets 1,5 and 2,4.

you connect whatever controller you are using to the 2 sets and try and balance the output. If the left panels output drops then rotate the rig right until the output is the same. You would also do this with the up and down set. The panel 3 would need to be the highest output in the set if 1,2,4,5 are all balanced and 3 lower than any of them your off by 180 degrees. This is assuming your have this rig mounted on a full 360 rotation platform. Adding panel 3 simply allows the rig to self center on a 360 setup.

One of the things I ran into was you need to allow a threshold of difference between the panels I would recommend about %5 before it moves the rig. Unless you like your motors to burn out within a day because its always moving. Also your samples need to be double checked. The thing should not spaz out if someone walks by it or a bird flys over. If it thinks there is a change and needs to move then set a move bit then check again in 20 seconds to see if it still needs to move.

Thats just my 2 cents. Maybe I over complicated it but I have done this before using a application written in VB and some serial output. It is kind of fun to just move everything around up down left and right then uncover the device and watch it find the sun only for it to realize it's off by 180 and flip around to correct it's self. This type of setup could even be mounted on a movable device such as a motor home or trailer. It would be able to track the direction of the sun while moving especially if you mounted it in a dome of some kind. You could also use something like this with known time as a compass. As long as you know the time and date. And where the sun is suppose to be you should be able to tell a person which way north is.

We were going to make a dodecahedron at one point just for fun that was run by a PIC microcontroler that was totally self powered and mounted inside a hamster ball that would roll around looking for the sun. it would know the time and know when it became futile to search for it. This device was going to be able to move when a shadow came. It was going to be put inside a large marked off area with walls and keep track of it's position and when the sun was completely gone would return back to the center and align it's self to where it expected the sun to come back up.

Tap3w0rm-


Sat, 06/20/2009 - 00:24 — Dockbot

Great info! I think this

Great info! I think this will save me a lot of time. It's still going to take a long time though. Is it worth trying to write a program that would only use 3 panels/photodiodes? My tracker is going to require some weird code since it will have NH3 and H2O lines running from it so I won't be able to do any more than 360 deg. movement. Thanks for the suggestions about thresholds and time checking! I didn't know what would be reasonable parameters for that. I may have to get with Mark or others who have learned to make PCB's because I like your idea for placement of panels for sun tracking but, I'd like to have them all on one board.

Even though some (ok, a lot) of these posts are over my head they're forcing me to learn and try to catch up so that's good and really appreciated.

The dodecahedron sounds really cool! I'm imagining it wandering around Makerfaire asking people to get out of its light.


Sun, 08/09/2009 - 00:30 — Dockbot

Concept Link

http://commons.wikimedia.org/wiki/File:Gas_Absorption_Solar_Ice_Maker.jp...

Now lets make it cycle more than once a day so it can be smaller!

Wed, 08/19/2009 - 22:46 — Dockbot

I'm straying from dorkbotpdx's theme but, still cool.

http://www.treehugger.com/files/2009/08/ice-bear-at-mec-burlington.php

More complex and non solar but, still an efficient example of ice making.

Ice is not the best way to store energy but, when you are using energy to cool something it makes more sense. Check out what this company was able to do with this idea. I think it still uses a compressor and vacuum pump though but, at least it does so at off peak times.