Cheap, Open Source Heliostats

Hey Infinia! Use modular mirrors!

2011-10-09T07:51:00.000-07:00

You gotta love Infinia’s PowerDish. Solar thermal generating instant residential spec A/C. It is a better-sized version of the SES attempt at Sandia. And as someone from Detroit, I love their interest in leveraging the manufacturing expertise there.

But they should consider modular mirrors. Check the specs on the PowerDish: 1900lbs with a 4.7m/15’ diameter mirror (call it 16m² or 175ft²). That’s big. I’m betting the installed cost of the dish sans motor is a lot more than $1600 ($100/ m²). Plus, if it breaks, it’s off line, whereas with 20 modular mirrors, if one breaks, it goes off line and you’ve only lost 5% of your light collection capacity.

Mirror curvature (and wind shear) could be addressed with modular heliostats. Ignoring a 1m² hole in the middle of your parabola, you’d have 8 1m² identical parabolic-slice mirrors in your first ring and 16 1m² mirrors in your second ring. You only need two mirror shapes. Injection molded curved plastic with PVD reflective coatings anyone? Include a few holes to reduce wind strain.

And let’s face it, that engine is designed for household use (1φ, 240V A/C). What you really want is for 1,000,000 homeowners to buy it and install it at their house. Is the hesitation releasing the IP into the wild before you’ve scaled?

Longer term, it just makes more sense to decouple the light collection from the power conversion. The power conversion component is where the value is added. Light collection should be a commodity. A monolithic design which ties these functions together reduces the chance that the power conversion part will be a success. And I hope Infinia is a resounding success.

V 2.5, catching up

2011-10-06T06:27:00.000-07:00

The base: this new base sets the relations between the rods within the cinderblock and concrete instead of using angle irons that rest on top of the cinderblock. To make it, pre-cut holes in the pieces of wood above and below the cinderblock. Fit tubes into the holes for the shafts that the threaded rods move through, and wedge in the 1/2” rod which provides the main mirror support. Then, pour concrete around all of it and put the wood piece on top to help set positions as the concrete dries. The result is the two shafts are firmly held in position relative to the center support. This version used metal tubing for the shafts, but I think that tubing can be eliminated. PROs: cheap, sturdy, no custom parts; solid shaft path; wood creates a platform for motor mounting and simplifies prototyping. CON: the points may end up too close to each other and require too much force from itty bitty motors. If need be, there are larger CMU’s (concrete masonry blocks) but the one’s I’m using are extremely common and cheap.

Manufacturability: the new base is easier to make. There is a template which can be laid on to pieces of wood so you can pre-drill holes for the shafts and motors. Having a standard piece allows the rods to be fixed in place before concrete is poured and keeps them plumb and square. The wood provides a platform for mounting motors and other items. Later the wood platform could be plastic but for now wood makes it easier for others to play with the design if they wish.

Other: threaded rods are up-ticked to 3/8” to make the whole thing more sturdy. The way the motor interfaces to the platform and spur gear is simpler. I’m using store-bought clevis & pushrod pairs but this whole linkage (including the interface to the mirror) still needs improvement. My current thinking is to use a saddle joint for the weight bearing point which would allow movement in two axis but not rotation, but this is still a challenge.

Electronics: I’ll do a separate post when ready but briefly… Current approach is to put a motor controller at each ‘stat. Also, I’m driving the motors with 9V supply to make their work at the easier end of their power range so they are less stressed (motor failure is an area of worry) and also to combat voltage drop. Connections between the Arduino and each stat will be by Ethernet type cable because it’s fairly standard. Also, it’s mass produced and fairly cheap.

SketchUp: wrong tool for the job

2011-07-09T13:36:00.001-07:00

I've been trying to improve the center mount so that it prevents rotation. The plan was to add a raised tab to the ball, then fit that tab into a slotted socket (constraining rotational movement down to one axis) then fit that slotted socket into another slotted socket (which would add one more axis of motion). The result was supposed to be movement in two axis without rotation.

This seemed easy enough to model in SketchUp for printing on the Makerbot, and if it worked, I'd send it off to shapeways. Well SketchUp is really the wrong tool. Whenever you make hemispheres with volume (5mm thick walls) and try to cut notches through it, you get messy messy results. I finally got something to work by downloading this person's work, hand editing out a notch, then making a copy that was 5mm larger all around, and sewing them together to create a closed volume. It works but is really lame. Results below.

So guess I'll try this in Autodesk Inventor which I've completely forgotten how to use. And we're moving in two weeks which should subtract lots of spare time. Oh well. Stiff upper lip and all that.

Hello hall sensor

2011-06-09T21:59:00.000-07:00

Got the Hall Effect Sensor working. This is the code it's using. The other gotcha for me was wiring it up right -- took awhile to see the teeny tiny notations in the specs that tell you the wires are V, Ground and Signal read from left to right with the printing facing up. Feed signal wire into input 2 on the arduino (one of the two available interrupt pins). Also the Arduino sketch wasn't compiling until I put the pin information inside the (setup) section, but I think that's just me not understanding rules for declaring variables and such.

Here's what worked:

void setup()

{

pinMode(2, INPUT); //set the interrupt0 pin2 to input

digitalWrite(2, HIGH); // and enable it's internal pull-up resistor

Serial.begin(9600);

and so on.

This adds some useful options for position control on an otherwise simple gear motor.

Center mount issues and position sensing

2011-06-03T18:34:00.000-07:00

Not real progress to post, but a couple things I've been thinking of lately.

First, the center mount needs to be replaced. Much as I love how easy, cheap and sturdy it is, the problem is that a simple ball in socket connection allows rotation (undesirable!) in addition to free movement in the X and Y planes (desirable). Consider when the mirror is parallel to the ground. It can move (by wind or due to the play in the linkages) easily clockwise (or counter) and this puts lots of strain on the threaded rods. I've had them bend and had failures due to this. So I think that center point needs to be re-thought so it can't move except for in the altitude and azimuth axis. In addition I'm going to use purchased, metal clevis yokes with tight fitting push rods to reduce play/slop in the mechanical linkages. Also, I think I'll uptick the threaded rods to 3/8". After having it spend some time outside, it just isn't feeling sturdy enough. The other idea along these lines is to make it into more of a monolithic frame. I submitted a design to shapeways to see about doing this sooner than later, but because they charge by the cubic centimeter, it will cost upwards of $1500 to print. So I'll hold off on that for a bit.

Second, responding to a comment on the last post, I agree that it's very likely I will require additional positional information. Right now I'm leaning toward using some Hall Effect Sensors. These sensors put out a voltage whenever they are close to a magnetic field. This is the omnipolar one I got from Digikey. And here is an excellent write up of how to use interrupts on the arduino to integrate this sensor into a counter. The idea would be to embed a magnet in one arm of the pinion gear and count how many times it has rotated. A count of how many times the gear has rotated, that's a decent proxy for mirror position. There are bipolar sensors but I think they'd be a bit harder to integrate based on the comments section for this part.

Anyway, a short post but just wanted to lob that out there.

Connecting the Arduino to the heliostat

2011-05-21T13:23:00.000-07:00

Four months. Ugh. Progress is always so much slower than hoped or expected. But anyhow… After some busy time at work and home, I’ve spent the last several weeks trying to get the heliostat to be software controlled – leveraging off Gabriel’s excellent work.

First job was getting an RTC going. I used this one from Adafruit, which is functionally equivalent to the one Gabriel describes here. It’s pretty easy to get this working.

The driver took a little more time. I am using very small DC motors because they are the cheapest game in town in terms of $/torque that you can buy new. (Steppers would be great but I just haven’t found them cheap enough). The plan was to use the half bridge circuit referenced in this post last August. That circuit may still be a good idea, but because I have limited experience with electronics, I decided to buy my way out of potential problems with an off-the-shelf motor driver. The L293D is an H-bridge motor driver built into an integrated circuit, it’s the chip at the top of the breadboard in the pic. It’s widely used with the Arduino so there’s lots of support out there. I used this excellent write up from someone at NYU to help figure out how to wire it and for code examples. I bought the L293D for $2.95 from Jameco, it may be cheaper elsewhere. Also the person who did the NYU write up mentioned an alternate chip, the SN754410, which was $1.95. The L293D apparently has “output clamp diodes for inductive transient suppression” which sound useful, but I don’t know if it’s applicable to this application or not. (Here’s a thread on the topic). But even at $3/stat (each chip controls two motors) it’s in line with my current budget to stay <$100/m², and could likely be reduced. Efficiency and energy consumption when not in use also needs to be addressed.

What’s taken the most time and isn’t done yet for sure is modifying Gabriel’s code. I think I’m very close – it runs how it should based on the serial port messages it’s sending to the computer but I haven’t connected the ‘stat to it yet. I will post the code over at Gabrial’s forum. Right now it has lots of comments and is very messy. It uses the module he wrote for simple gears, modified a bit due to the different geometry of my ‘stat. Also, I took out the shift register part which is required for multiple stats, the module for multiple targets, and lots of the empty code for running additional ‘stats off the same Arduino. If the code works, my next step will be to add back the shift registers and set something up so it can control at least 4 stats. When it’s cleaner, I’ll post here but if you want it now, it will be on his forum.

So many plans, so little time. But, getting it running off code is an important step. I sure hope I can test the code on the 'stat before leaving for some travel on the 24th.

Linkage Improvements

2011-01-23T20:58:00.000-08:00

So here are the results of some attempts to improve the linkages and connection points. First is the center mount which supports the weight of the mirror. The close-up pic below shows the bottom half of a ball-and-socket joint. The four holes have capture nuts below them. There is an acrylic hemisphere (bought from here) stuck in there which will receive the ball. The braces are these. I got them because my first go was using a smaller ball (same as the ones used for the moving points) and the mirror slid off. This is an important connection which needs to keep a 1 meter square, 40+ lb mirror from flying off in wind so the new one uses a bigger ball (with 1/2-13 threads), a beefier connection point, and those braces.

Just to be clear, here's what it looks like with the ball embedded (with a bolt in it) and the second piece attached with two of the four bolts in. That second piece is tiered so it comes down to where the ball is starting to taper in so that you can't pull the ball out of the socket once it's in place. But, the plastic piece itself is not fixed to the wood except by the braces -- that may be required at some point. All the weight is from above so it's fine under normal conditions but if a wind really wanted to pull the mirror up it might not hold as-is. So far though it works well.

So in the pic below, the center mount is now in the background. The pic shows one of the linkages that goes on top of the threaded rods. As you can see, this is also a ball-and-socket joint at the connection point to the mirror. Using the same plastic balls except they're smaller: these have 1/4-20 threads.

Below the ball you see a brass sleeve over the threaded rod, then a wingnut, then another nut holding the wingnut in place. The purpose of these things is to prevent rotation of the whole linkage (it catches on something connected to the socket mount). It's unsatisfactory.

Below that is about 3" of threaded rod with an eyebolt. I wanted to keep the eyebolt 1/4-20 and got them at the local hardware store, so the eyes are large, which is why I'm using the 1/2" nut & bolt to connect to it. That whole bottom part of the linkage is essentially a clevis. It needs to 1) connect to the threaded rod, and 2) allow pivoting in one axis only. Since I have a Makerbot it was just cheaper and more flexible to print one. The functional reason for this is when the mirror is at an angle, the connection point will be a few inches offset from the threaded rod. The trade off for this is I get to use straight threaded rods which are off-the-shelf as opposed to using curved threaded rods which is how some people solve this (including a link to this astronomer's site who does long exposure photography because it's so well done, but I've seen folks on youtube use curved rods for heliostats too). Longer term it would make more sense to do this whole linkage from stock metal or plastic parts.

So this makes some progress from last go: 1) Linkages are better (still not ideal), 2) Support point is now a made piece rather than carpentry (meaning it can be mass produced in plastic -- though it can be pretty easily done with carpentry also), 3) It's got both motors in place now so I can start to play with software more easily. Software-wise, Gabriel Miller has had heliostat tracking working on Arduino for many months so I'll be starting with his code.

When it's satisfactory, I'll post files and assembly instructions. But if there's anyone reading who wants any of them, just comment and I can post 'em.

Cosine loss

2010-11-14T18:51:00.000-08:00

Cosine loss is the term for the energy you lose by not facing the sun's rays directly. For example, if you face a 1 square meter solar panel directly at the sun, meaning perpendicular to the incoming rays, you get all the energy available to that collection area (about 1kW). If the panel is at an angle to the sun, you get some fraction less of the available energy which is a function of the cosine of angle away from being perpendicular to the suns rays. See drawing below.

It's an obvious concept but under-discussed, probably because it results in lost power. If you think about it a bit, it explains a lot about configuration of solar collection systems. It is why all power towers are tall and the mirrors face South (in the Northern hemisphere). Since the sun's path through the sky varies across 47 degrees over the year, you'd want to put the target in the middle of that, so your cosine loss never exceeded 23.5 degrees. Cos(23.5) = .917, so you only lose about 8% energy in the worst case.

In the video below, my target is on a South facing roof with a 30 degree pitch. If there were solar panels on that roof, they would suffer least cosine loss when the sun is at a 60 degree altitude (measured from ground), around April 1 and mid-Sept -- that's when the runs rays would hit them dead on. Where I live, in mid-November, the sun is at about 40 degrees altitude (measured from ground). So a solar panel on my roof is about 20 degrees out of ideal this time of year, about 6% cosine loss. My heliostat was angled about 10 degrees, so actually my mirror suffered significantly more cosine loss than would a panel, about 60 degrees out of ideal, which is a 50% loss, given the target (assuming the target is perfectly hit).

The reason for bringing this up is that supplementing a residential solar panel installation probably provides the best return on investment for a cheap heliostat. That is, if you already have a solar panel installation, there is an excellent chance it is underutilized and a $500 investment in heliostats could deliver 10% returns in energy savings. It's not a particularly "efficient" use of the heliostat. The ideal residential installation would have a well placed target, ideally a solar panel that could handle multiple suns. Then again, power is as pure a commodity as there is, and 10% ROI is 10%. Still, it's worth keeping in mind.

_____________________

Some additional description:

This is an image from Power From The Sun (Chapter 2). Their description is this

"An instructional concept, and one often used in solar irradiance models, is that of the extraterrestrial solar irradiance falling on a horizontal surface. Consider a flat surface just outside the earth’s atmosphere and parallel to the earth’s surface below. When this surface faces the sun (normal to a central ray), the solar irradiance falling on it will be I_o , the maximum

possible solar irradiance. If the surface is not normal to the sun, the solar irradiance falling on it will be reduced by the cosine of the angle between the surface normal and a central ray from the sun."

Comments on v2.4

2010-11-13T11:46:00.000-08:00

It's great to be testing outside and today is a nice milestone. So, it's a good time for some frank pros and cons.

PRO:
- It works! (In a just barely kind of way).
- The motors are wonderfully cheap.
- The approach of supporting the mirror weight at a center point so you can use tiny motors to position is still potentially viable.
- There are plenty of opportunities for reducing costs via mass production.

CON:
- There's no software to speak of. The time lapse test was done with a fixed movement over a short time around solar noon.
- The linkages just stink. I mean really.
- The seasonal axis needs a brace, a motor, and a new linkage.
- The support point is still made by carpentry rather than a buy-and-assemble system. I have a system ready to try but it is untested.
- It needs some custom (cheap) motor controllers, AND a feedback system for the motors
- I'm really moving slowly on the project.

All in all, it is riddled with flaws but I'm pretty happy with it.

V 2.4 breaths the open air

2010-11-13T11:34:00.000-08:00

Another little realization

2010-11-03T07:43:00.000-07:00

The possibility of a computation engine on the internet puts a ceiling on control costs.

Meaning, if an Arduino-type microcontroller can’t calculate solar position, keep a reliable clock, issue motor controls AND do additional features (like go into a locked position if winds >30 miles/hour, or run test sequences to identify module failures, etc), then a computer sitting on the internet could do some of that work instead. The local microcontroller could be responsible for internet connection and i/o to the motor controllers. Given an internet connection (a big if), you could query an external server for time and solar position, motor instructions, or for more complex add-on programs. These servers could be government run or privately hosted. They would follow a defined standard a la SMTP (STP = Solar Tracking Protocol anyone?).

Not necessary, but possible.

Small epiphany

2010-10-26T21:55:00.000-07:00

I realized today that an operating ‘stat like v2.3 is a proof of concept for a design that would be built differently. If mass manufactured, it would probably be a plastic frame, motors and hardware rather than the current angle iron-based system.

The design questions that the proof of concept needs to address include:
1) Is the system mechanically practical?
Does the center point bearing mirror weight and allowing full X/Y motion work? Can the mirror be positioned as needed by the other two points? Can the other points provide the positional stability needed for a large mirror in wind? Are the current motors (small gear motors, not steppers or servos) usable? Does the system provide adequate accuracy? What mirrors would ultimately be practical for heliostats (lifecycle cost analysis)? Etc.

2) Is the system electronically practical?
Can an open source microcontroller & motor controller group handle a residential size array (2-20) of heliostats cheaply and reliably? Can a power distribution system work cheaply and reliably? Can additional sensors adding capabilities be added easily? (Example 1: heat detection placed at PV cells that cuts out stats above a certain temperature. Example 2: additional positioning sensors at the stats). Can the whole system be matured enough to make installation relatively simple? Etc.

3) Is the overall system viable?
What is the true final cost per square meter after mechanical and electronic corrections are incorporated? What types of installations are possible or attractive? What is the ROI in actual installations? Can installations be made simple enough so almost anyone could do it? Are there economic niches where the system could flourish? Are there installations where it could compete economically with conventional power generation? Etc.

The system is essentially a base and frame that holds the components in fixed relationships to one another paired with some microcontroller logic and directions on how to assemble and run the thing. It would be cheaper and more consistent to do this in a piece of blow molded plastic that fixes all the special relationships, provides mounts and protection for motors, wires and rods, can be filled with concrete or sand for weight, all while simplifying installation and improving reliability. Less modular, yes, but more practical.

The exercise of building it out of off-the-shelf parts is valuable. But if the exercise proves a success, the end product (what you could buy in a hardware store, or manufacturer yourself) will look different.

Version 2.4, little fixes

2010-10-26T21:37:00.000-07:00

Another configuration of motor mount. An issue with the other was that if the threaded rod drifted out of its path, it put stress on the motor. The function of defining the path for the threaded rod needed to be separated from the motor mount. Hence, the second angle iron 8" below the first. Nice thing about using angle iron is you know the holes are aligned relative to the center rod. You can also see another 1/2 cinder block which I added because the first angle iron was a bit whippy from being cantilevered out so far.

The eyelet nicely prevents the t-rod from turning, but one of the purposes of the in-line ball joint was to handle the angle change (and associated length change) when the mirror is at larger angles. The fix is hardware that pivots in only one axis, like this, but I can't find it cheaply. They're about $3 each in large quantities, but that's still pretty pricy. Working on it.

Parts list: $100 per square meter

2010-10-17T21:32:00.000-07:00

$1: Cinder block
Not what I used, but close enough.
$1: Concrete
I used about 1/4 of a 50 lb bag to make one stat
$3: 24” length 1/2”-13 threaded rod
Ballpark
$1: 1/2”-13 nuts (3), washers (4)
Cost is a guess
$9: 24” length 1 1/2" perforated angle iron and 12” length 1 1/2" angle iron
About $3/foot
$10: Motors (2)
$5 each in quantities >50.
$2: 1/4”-20 nuts, bolts and washers (say 12 each)
Guess
$10: Plastic parts
Guess. This would be in volume
$2: 12” 14”-20 threaded rod (2)
Roughly
$6: In-line ball joints (2)
$3: Phenolitic balls with 14"-20 threaded inserts (2)
I think I got these for 50 cents each, but using this for now
$35: Mirror, 1 square meter, Ikea
Controversial. I'm sure true 'solar mirrors' have coatings for strength and UV resistance, etc, so this may give too cheap a perspective. But, these are also retail. ESolar is working with Google to figure out how to make cheaper solar mirrors. Here's a general link on the topic.
$4: Wire
Guess -- .20/foot x 20 feet average per stat
$5: Arduino
Maybe it's cheating, but I'm assuming one arduino runs at least 6 stats
$3: Motor control electronics
Again, maybe cheating, but in any mature situation, these would be custom made for cheap.

In general, these are retail prices but without taxes or shipping. There are no costs for installation and profit. It's not meant to be perfect. In earlier calculations I've gotten into the $80 per square meter range and I still think that's possible.

My vision is something like an in-lawn sprinkler system. Central control is in the garage, 10-20 mirrors would be a typical residential installation. A home owner could do it themselves, or hire someone local who could install it all in a day or two and make $500 or so. 10-20 square meters of mirror isn't enough to power your house but if the ROI was 5%-10% it would attract some interest. A mid-tier application might be 100 heliostats generating about 10kW at peak, making the owner about $2K/year for a capital investment of $15K-$20K, so 10%+ return. A larger application would be an open source 'power tower' type configuration with thermal storage and power generation from steam turbines.

Note: For the figures above, the value of generated power assumes available solar energy of 1000 watts per square meter, with the conversion to electricity operating at 10% efficiency (100 watts per square meter). It assumes 5 hours a day of sun, 300 days a year, so this really only works for the US Southwest and similar climates. Power is considered to be worth .12/kW-hour -- whether used or sold back to the utility.

V 2.3, an update. Finally.

2010-10-17T19:24:00.000-07:00

Hola amigos, it's been a long time since I rapped at ya. The I video below is a quick update, and this post will provide detail.

So, the main news of the past months is the arrival, assembly and tweaking of the Makerbot. Assembling it is fun, but the tweaking takes awhile. And, it tends to break a lot so you have to fix it all the time. But, the MK5 plastruder, which arrived in September has greatly stabilized it and now it is a prototyping tool rather than a troubleshooting time sink. There is a great on-line community and I give the product, company and community a big thumbs up. They also just came out with a new product, essentially the Cupcake v2, which is called the Thing-O-Matic. The leap they are attempting with the thing-o-matic is automation -- the fabrication of many parts, unattended. Using the Cupcake made me re-think the design. The initial intent was to make something that others could assemble themselves with off the shelf parts. It's clear now that automated custom part fabrication will soon be commonplace and there is not a reason to exclude custom parts from a public domain/open source design. If custom plastic pieces can make the heliostat work better, assemble easier, facilitate upgrades, and simplify repairs and maintenance, they should be included.

So once the gears were done (pleasingly easy) the next element I focused on was the motor, mount, and threaded-rod component. The new motor is the GM17 from Solarbotics, mentioned earlier. It's got good torque and takes about 1/3 less power (253mA stall current at 3V vs 400mA) than the GM2 used previously. I wanted a mount that would fix the motor in place using simple connections to the angle iron. Another bracket holds the threaded rod in place relative to the motor.

First pic shows the mount and the motor. You can see an embedded notch which receives a nut that holds the motor mount to the angle iron. Easy to do on a Cupcake and it makes for a simple installation.

The motor fits into the mount like this...

Then the whole thing mounts to the angle iron like this. You also see the top bracket, which holds the spur gear down (so it doesn't float up the rod) and provides the pathway for the threaded rod.

It's not perfect but I like the direction. Motor is kept securely in place without having to drill into the metal of the angle iron, it's fixed with simple screws. The motor also keeps it's place relative to the spur gear and the threaded rod. Installation takes a couple minutes.

This was the piece that convinced me that plastic pieces are a net positive even though you can't easily buy them at the store. I'll upload the files. Once they're mature enough, anyone with a makerbot could make them, or, there are many on-line production shops that could fabricate them. It should be noted that in some cases the parts are made to accomodate constraints in the build capabilities of the Cupcake. C'est la vie. Ultimately all the parts (except the heavy stuff like cinderblock, concrete and a mirror) could be bought on-line as a kit. There is nothing proprietary so costs will be low.

Next, a parts list and posting of source files. It's not there yet. This is first assembly, it needs many adjustments, and many improvements are tempting. But, it shows some promise and there is now (or will soon be) enough detail that others can take the design and run with it if they wish.

Version 2.3, first assembly

2010-10-17T18:10:00.000-07:00

Command and Control

2010-08-10T19:50:00.000-07:00

While waiting for Makerbot to release their new plastruder, one's thoughts turn to other steps of the project...

The biggest part is software -- getting a sun positioning algorithm to run on arduino, then modifying it to accomodate multiple heliostats pointing at a common target a la Gabriel at cerebral meltdown. But, WOW. I just looked at his site to add the link and he's done a lot since the last time I visited. He's working on an arduino platform now and already has sun tracking working and, he has it working for multiple heliostats and multiple targets.

Next issue is how to cheaply run lots of motors off the same hardware that is doing sun tracking. Gabriel has a solution for this too, here. I've got it a little easier in that my motors only use 400mA at 3V. I can probably use something like a ULN2803, that could power eight of my motors no problem. A board with 4 of those plus the shift registers and some nice connectors for getting wiring out? Probably $5 in parts to control 32 motors = 16 heliostats. All in for a 32 'stat array? Arduino ($30) + time clock ($20 -- cheaper than GPS) + motor drivers (say, $30) = $80 = about $3/stat. As long as it's under $5/stat, it's acceptable.

Anyway, despite some Makerbot delays, seeing other people's work (there are lots working on this) is reinvigorating. More to come.

UPDATE: after a little homework, I found this article, a great review of motor controls. It provides a circuit using a ULN2003A chip and points out that you need to add DPDT relays otherwise you can't move the motor both directions. It also says a very cheap way to do it is with a circuit called a half bridge made out of only a very few, cheap parts. The only reason not to use it is it requires 10mA to signal, but the Arduino provides that and shift registers appear able to pass it through also. It's high efficiency too so is certainly worth trying especially in a high production environment. It's designed for 3V motor that draws 300mA which is very close to the Solarbotics motor's specs except that it draws 400mA at stall.

And, checking back with Solarbotics, they now have this motor, which is very similar to the one I'm using but is even more efficient, producing almost the same torque at 3V but with a stall at 253mA. So the half circuit would work as designed. That pairing is attractive longer term.

Homemade gears

2010-08-02T18:03:00.000-07:00

So these...

are the first heliostat pieces made from my Makerbot.

Not perfect, but consider this:

Gears are made in SketchUp.
Here's the plugin you need.
Export to stl file with this.
Convert to GCode. I'm rolling SkeinFox
PRINT.
Not what you want? Alter it and re-print. One hour turnaround.
Make em any size you need.
Choose the teeth count you need (I may go fewer teeth as this number, 16, is starting to exceed the precision of my printer).
Cut out a center hole to match your motor shaft.
Intent a hexagon so you can capture a nut (the right one). Not so easy to buy on-line eh?

All in all, it's very different and much better than buying them made by someone else.

There are a few other parts to make. The Makerbot takes some work to configure right (there's a new plastic extruder coming out in a few weeks that may improve things), and that slows things. But, it is progress.

What could you do with a cheap heliostat?

2010-05-16T22:46:00.000-07:00

Lots of things of course. You could use it to provide light and warmth to a room in winter like this guy does. You could provide additional light on an existing installation of solar panels (with a sensor to remove it if those panels reached high temps as redrok suggests). You could heat pool water (OK, parabolic, very different, but it’s too great a blog not to link).

But, like many others, I’m interested in solar as an alternative to extracted fuels like coal and oil. So I think mainly about three applications (in order of interest):

1) A residential use of concentrated solar on PV cell.

2) A quasi-residential use of concentrated solar thermal paired with storage (and converted to electricity by stirling engine or micro steam turbine).

3) Commercial power tower applications

I focus (ha ha) on the first of these because it seems the most plausible. No one sells cheap (consumer, single junction) solar panels that can handle the high temps of multiple suns. But I think (hope) that is more because there is no market for them than because there are technical constraints to creating them. A cheap heliostat might create a market for a consumer level solar panel that could handle higher temps. So there may (or may not) be a market for a solar panel that could sustain 10-50 sun concentrations. Further, it would be easy to make the target high in the air (like on a roof), which minimizes the physical danger of that level of solar concentration.

1000 watts per square meter is the canonical number of solar insolation for calculations. 50 times concentration (think of 50 heliostats shining on 1 square meter solar cell – it needs a lot of space) is 50,000 watts per square meter. 50,000 / 10,000 (number of centimeters per sq meter) = 5 watts per square centimeter. 100 square centimeters (a 4” square) would be 500 watts per 4” square. It’s hot, but not that hot. It seems possible that in an outdoor application, passive heat removal (or, a small fan) could work. It’s useful that you are immediately adjacent to an infinite heat sink (the great outdoors). You want to avoid water cooling because it carries complications and cost. Incidentally, that configuration (50 stats), with a 10% efficiency of conversion to power, in the US Southwest (300 days of sun/year, 5 usable hours/day) generates 7,500kW-hr per year (50 x 1kW x .1 x 300 x 5). People in the US use around 11,000kW-hr each year. Your mileage may vary, as they say on the inter-googles.

Residential applications are appealing because they’re cheap – many individuals would happily self-finance if it returned 10% and they had the space. And, the money is spent largely on installation and maintenance. That means you are paying people, in your community, instead of firms with large capital requirements.

One could make a good case that residential applications won’t have much impact on climate change. But, it proves out the technology and is interestingly disruptive in the same way wi-fi was. If it works at a household level, it could be expanded to “community size” or larger.

Speaking of community sized applications, I think option 2 – CST with thermal storage – is the more sensible approach in a larger picture. If you have a cheap way to collect heat, cheap thermal storage has a large multiplier effect. That super expensive 20% efficient 25kW Stirling engine can now run for 24 hours a day instead of just 5-6 hours, so it amortizes way better. And, of course, it addresses the fact that we sometimes use power at night.

Look, I hope for magic technology bullets as much as the next guy. But at this stage, everything is possibility, a clean energy reality doesn’t exist. And, I believe that the more competitive the clean energy options are, the faster they will move the most viable options into place. My project isn’t going to slow anything down. If I am very fortunate, it will speed someone else up (or cheapen them).

So that’s some of what you might do with a cheap heliostat. Perhaps someday it will be worth writing up more thoughts on option 2. But, far more valuable is to just develop a cheap ‘stat and see how people use it.

Frustration and evolution

2010-05-14T20:38:00.000-07:00

I’m very frustrated at myself for not moving faster on this project. Very, very frustrated. That said, I want to note a change in design direction.

The original plan was to build it entirely out of parts that are mass manufactured and could be easily purchased. I’ve changed that stance. The mirror mount and probably the points where the threaded rods connect to the mirror are simple but not easily bought. So I’m going to make em. Yep, my Cupcake CNC has shipped. After I screw around for weeks on assembly and then months on learning how to model stuff (boy, I hope that was a joke), I’ll be in the mix. If I can make some plastic parts that make it easier to build and more functional, I’ll do it. If it’s worth it, I can find a way to cheaply produce them and sell it all as a kit on adafruit, solarbotics or the like.

Gonna have to switch to glass mirrors

2010-04-19T20:57:00.000-07:00

Here's a picture of the reflection made by the 2' x 4' piece of mirrored acrylic I've been using.

Here's a picture of an ordinary, glass mirror. About 2' square. Yeah. The acrylic isn't really going to work.

So, back to glass. Ikea sells this for $8 on their website, $10 at my local store. It's about 16" x 48". I was planning on using two, which gets very close to 1 square meter. So glass costs are $20/meter square. Which is much cheaper than the acrylic. The downside is only weight, and perhaps breakage. Ikea also has these for $5, which is $1.25/ft2 = about $15/sq meter.

Version 2.2

2010-04-19T20:13:00.000-07:00

The best is the enemy of the good so I'm just going to stop procrastinating and post. Here's a couple pictures of current status. First an overview against a yellow wall.

Next, the same again, closer and from below. It's easier to see the linkages from here.

And, a close up of where the motor interfaces with the threaded rod.

So the changes in this version are...

1) The way the motor connects to the angle iron is simpler and better. I laid it out so the threaded rod would go through one of the round holes, let that determine the position of the motor, then drilled holes to mount the motor.

2) It uses a tee-nut instead of a gear. The version in the last post used a 1/4" bolt that was clamped to the motor shaft. That kept falling off so for this I drilled a hole through the motor shaft and put a #4 machine screw through it. It tended to catch its threads on the tee-nut, so I took a dremmel to the end of it to smooth it and that helped. In retrospect (or for a next version), it might be easier just to pre-drill a hole in the shaft then drive a nail through it. The tee-nut is a rare 6-prong version. That works well because it only has to be turned 60 degrees per motor rotation (it's more forgiving than the 4-prong tee-nut -- you can put the motor farther away). I like the tee nut for being cheap but it makes the interface a little finicky. Longer term I'm tempted to make real gears (maybe with this!), because if it ever went to production, it'd be easy to have gears manufactured cheaply.

3) The piece that keeps the threaded gear from riding up (top cap?) is done more neatly. It's an L-bracked that is bolted to the back of the angle iron and then the L comes over the top so the threaded rod goes through it. Works well, though it's one more thing to make. It probably needs one at the bottom of the rod also to help define the threaded rod's path as perpindicular to the angle iron.

4) The way the threaded rod links to the mirror actually works (and it's still not super stable). When I ran the motors to move the threaded rod and tilt the mirror in v2.1, it would get out of alignment. The addition of a second in-line ball joint fixed that issue for now.

5) There's a brick supporting the long angle iron. This wants to be another 8" high cinder block but I have not gotten to the store for that.

So it's better, but still not where it needs to be before it's worth documenting and field testing.

Cheap gears

2010-03-09T21:41:00.000-08:00

Gears aren't as cheap as they should be as Erik Rossen noted oh so long ago. So, I was trying to use tee-nuts as a crude gear that would be moved 1/4 turn by a rod perpindicular to the shaft of the motor. It works, and it's cheap, but I'm not sure if I'll incorporate it or not.

Here's a pic.

The the element perpindicular to the motor shaft is a bolt. It's attached to the motor shaft by clamping it to the shaft with a couple of nuts. There's yellow tape on the end 1) cus it wanted to be just a little bit longer and I didn't have a longer bolt and 2) the end needed to be smooth so the threads of the bolt didn't catch in the edges of the tee-nut.

In other news I was looking for a good, mass manufactured piece of plastic that could be used for the mirror frame (adding stiffness) and as a replacement for the perforated angle iron. The cost of the metal pieces is a bear. Larger diameters (2"? 4"?) of PVC pipe that was ripped lengthwise might work well and be a better price point, but it adds some complexity to acquiring the material cus you can't just buy it ready to use. I'll investigate and report back but I'm trying to get a version built that's decent enough to post a parts list with costs and instructions on how to make it.

Hello world, I'm version 2.1

2010-02-28T15:54:00.000-08:00

It works! Here's a pic:

Little changed from the v2.1 posted below, except the motor is now attached. The wire hanging off connects to the Arduino Motor Shield. The motor turns the threaded rod, which moves up and down easily and the mirror tilts as expected. Not bad. One difference is the mirror is spaced off the in-line ball joints by 3" screws which allows the straight threaded rod to interface with a mirror at 30 degrees. Also, there isn't a motor on the altitude (seasonal) axis yet -- it's fixed for now.

Immediate next steps are to build a v2.2 and document it:

Use one angle iron and plan holes around it.
Having the angle iron be a small piece was useful for first assembly but is unnecessary at this point. The right way is to use a single piece and lay out the mounting holes so the motor shaft and threaded rod are aligned with existing holes (to reduce assembly hassle). Also, lay out the holes in the mirror to align precisely with these.
Lay it out with cheap gears that could be used in production.
Add another cinder block to provide additional support under the angle iron.
Set up wiring so the device can operate on it's own 100' from a motor controller which has power but isn't connected to a computer.
Misc.
- Improve motor mount
- Improve precision of mirror holes
- Be more organized about parts and assembly sequence
- Weatherize a bit
- etc.

And then larger steps are 1) document a parts list, costs, assembly and how it operates (power requirements, # rotations, speed, etc) 2) start in on programming -- at least something basic, 3) consider applications.

But, all in all, I think it's a very cheap heliostat hardware platform. We'll see what develops.

Movement!

2010-02-22T16:30:00.000-08:00

At last. I finally got the GM2 motors to turn a threaded rod.

Description (from the video):
Here's a short video of the GM2 motor moving the threaded rod up and down. You can see the GM2 motor in white turning a metal pinion gear which then turns the black spur gear. Glued to the spur gear (above it in white) is a threaded spacer which is what turns the threaded rod. Below the spur gear is just a 1/2" nut used as a spacer to get it to the height of the pinion gear. The perforated angle iron above both gears is to provide a top so the spur gear doesn't float up and out of contact with the pinion.

The GM2 motor is controlled by an Arduino with a motor shield. At the moment the sketch for it just turns it for 4.5 seconds, pauses 1 second, then repeats. The motor completes a rotation in about 1.5 seconds, and the pinion to spur is about 3:1 so 4.5 seconds should turn the spur gear about once. The threaded rod is 1/4-20 so one rotation of the spur gear will move the threaded rod about 1/20th of an inch up or down. The arduino and motor shield allow scripting of motor speed and pauses between running versus stopping the motor which should provide pretty precise control of the speed the threaded rod moves up and down to enable sun tracking. The next step is to connect it to the heliostat. I'll add detail on the blog, at www.heliostats.org.

A little more detail and comment:
It's mounted on a one foot length of perforated angle iron. I'm not sure if the angle iron is the best idea because you are tempted to try and use the existing holes even when they don't quite line up to where you need holes. But, mounting it all on a shorter piece that is then connected to the cross bars works well. Once the mounting details are stable, I'd want to make a template and use a drill press to fabricate many of them at once. As a one off (and I don't have a drill press), it was awkward. I broke a few drill bits. Ultimately it could also be a machined piece.

The GM2 has holes for mounting, so I clamped it to the angle iron and just drilled through it's mounting holes. Those holes drilled easily. Source some long (1") #2 nuts and bolts for mounting. The next problem was attaching the pinion gear. I had purchased some pinions and spur gears from my local hobby shop making sure they were all the same pitch (48 as that seemed most common). Then bought these which Solarbotics offers as an option to the GM2. I mounted that and then screwed through the pinion into it. Not perfect. A next step here is to find a good source for cheap gear pairs and then find and describe a mounting method. Maybe something like this for $1-$2.50. The exact tooth count or ratios between the two don't matter much as that can all be fixed in the programming for the motors anyway.

Anyway, once the motor and pinion were mounted, that set the location for the spur gear. I marked that and drilled it but it happened to fall between holes in the perf angle. Here's where I broke some drill bits. This probably wouldn't have a been a big deal for someone with a drill press or more experience but eventually I hacked it out. What you can't see because it's inside the 1/2" nut acting as a spacer is a nylon bushing (example) for the 1/4" rod. There is also one that is visible that sits on top of the threaded spacer pointing up. The purpose of these is to make it easy for the threaded rod to move up and down. Also, the spur gear needs to be able to spin easily. If it can't spin, it will basically act like a screw head and grind into the surface of the angle iron until it stops. If it can spin, AND if the threaded rod doesn't turn, then the threaded rod will move, which is what we want.

That brings us to two other elements. First the angle iron on the top that is spaced off the bottom angle iron with a bolt on each side. This serves as a ceiling so the spur gear can't rise any further -- since it can't go up but can spin, it forces the threaded rod up through it.

Next, the long bolt hanging down to the left of the spur gear. That, combined with the L bracket fixed to the bottom of the threaded rod, is to keep the threaded rod from spinning. If the threaded rod can spin, then it will just turn in place. Only when the threaded rod can't turn will the turning of the threaded spacer cause it to move up or down. In this picture from Gary Seronik's write up, you can see the threaded rod is fixed to the top platform with the cap nut on top and jam nut on the bottom of the platform.
In our case, we can't fix it to a platform because the mirror (which moves) will change angle. Gary solved this by having a curved rod but I want to avoid that. So the idea is that the L bracket (which will be replaced with something simpler) will slide up and down that long bolt keeping the threaded rod from spinning. This isn't perfect but it's the best I've got for now. Also, because I had previously used the strange interface plate up by the mirror to accomplish this, those can now go away, replaced by good old in-line ball joints.

Next steps:
Next steps are 1) complete and document the hardware, give it a simple program, and see how it fares in weather, 2) plan programming on the Arduino, AND wiring and power for larger arrays of heliostats.

The goal is for the microcontroller to run a suntracking program and provide direction to multiple heliostats. (Gabriel at cerebralmeltdown has done this. Here's his solar tracker program which controls two heliostats each with up to 10 targets.) Lady Ada's arduino motor shield, which I use, will control 4 DC motors, but that's only 2 heliostats. That's not enough to provide significant power generation if for no other reason than it doesn't amortize the electronics across a large enough number of heliostats. Eventually, I need something that could do at least 10-20 'stats, which is 20-40 motors. The reason for so many is you just don't get that much power out of them. Assuming you could get a solar panel built that could accept, say 10 suns, and operated at 10% efficiency, that's only about 1K-watt/hour max generation. That's maybe 1500kW-hr/year if you live in the Southwest. Average electrical consumption in the US is roughly 10,000kW-hr, so that only gets you 15% of the way there. Now if it pays for itself, it's still worth doing, but 2 heliostats doesn't really cut it. You need a bunch.

Well, that's enough for now. More when it's connected up and operational.

Version 2.1, status update

2010-02-16T21:44:00.001-08:00

Progress is slow. The way of the world perhaps. But there is progress.

Amateur Astronomers to the Rescue
Amateur astronomers are cool for too many reasons to recount. But one reason is they often post about their DIY star trackers which use threaded rod and fixed nuts to move their telescopes. Gary Seronik provides a great example of work here. He uses a gear motor (you read that right – not a stepper or servo) to move his camera at a constant rate so he can take pictures of the stars with exposures lasting several minutes. And though the precision required for long exposure photography exceeds that of sun tracking, the method is very reusable. He uses a curved threaded rod, but I prefer straight to keep the materials simple. Using a straight rod means we will have to vary the speed of movement, and also have to deal with a changing angle where the threaded rod connects to the mirror.

The next inspiration I took from Mr. Seronik's project was to try a straight gear motor. Something that began to bother me about the motors I'd planned to use is they are repurposed. An important part of my project is that it be easily replicated (and improved) by others at price points which meet the original project goals, so a new motor is preferred. Solution? This little baby offers almost 50 in/oz of torque for $5 (in quantity). That's about as cheap as you can buy torque. And, it takes 1.5 seconds to turn a revolution. Pretty slow, which is ideal for solar tracking, and it will be slowed further by moving it through a worm gear. I suspect this could be moved for 3-6 seconds every minute or two and that would provide adequate resolution over a day. They finally motors arrived yesterday.

So here's how things look right now:

V 2.1 updates include a simpler mirror frame (which may not work out), cross bars upgraded to angle iron for stiffness, a simpler mirror mount (I struggled with this for weeks, have many ideas, but I think a simple in-line ball joint will do), and the beginnings of connections from the cross bars to the mirror.

Here's another view:

The connections at the top are meant to provide a fixed point for the threaded rod that is linked to a ball joint which connects to the mirror. It works ok for now but will probably be changed.

Anyway, I wanted to get something up so prove I had not fallen off the edge of the world. Will begin playing with these motors and hope to post on movement soon.

Version 2, promising but not there yet

2010-02-01T18:37:00.000-08:00

I hoped to get to a 'hello world' level of operations on version 2 prototype this weekend but didn't make it. But, got a new base and mirror mount and those work as well as could be expected.

A new base
Jettisoned the unistrut approach in favor of something cheaper. It's uglier than v1, which is a good sign. It's a 8" x 16" cinderblock, something like this, filled with concrete ($4 for a 60 lb bag). Here's a pic.

Put the 1/2" -13 threaded rod into the cinder block and add wet concrete mix. A few rocks gets the rod reasonably plumb. Then I added the cross bars you see -- these are pieces of strap metal you'd find in the building/fence section of a hardware store. The long pieces is maybe 2' and the short maybe 1'. Little detail here cus I think I'll change them for angle iron in the next go around -- they're too whippy without an angle and don't have enough holes to mount other things to. Showing two pics because I poured the first amount, then added the metal straps and bolted em, trying to make them close to a right angle (later the long one will be an east/west axis, the shorter a north/south). Completed, it weighs about 62 pounds (28 kilos), which is nice since it's purpose is to anchor the heliostat so it doesn't blow away. Could make it heavier by filling the other side with concrete. It's a sturdier (heavier) and cheaper solution than v1 so it's a keeper for now.

Mirror mount
Using the 'gyroscopic drink holder'. Here are a couple pics.

The nice thing about the threaded rod is that is easy to hang things off it. For the moment, a piece of angled metal creates a horizontal base and the mirror mount is fixed to it with double stick tape. Inside the basin of the 'drink holder' is a piece of 1" wood wrapped in duct tape. I cut it pretty close then wrapped it in tape so it sort of wedged into the bowl. That is later mounted to the mirror frame.

As noted below, I believe this mount is an important design element. Ideally it bears most/all of the mirror weight, but other than that, easily allows tilt in any plane. A pin point would do this, but the mirror would fall off it, so the mount also wants to bear weight off to the sides. So, bear all gravitational (axial) load, and also bear sideways (radial) loads. This has got to be a typical bearing, and I hope to find it. Meanwhile, I'm looking into circles of bearings that a sphere could rest within (like the pics below but upside down). For the moment, I'm using the gyroscopic drink holder. I've bought a lot of nesting bowls lately but have not found an easy way to recreate it. Still, even if that were necessary (which I doubt, there should be a simpler more commonplace bearing), it would not be expensive to make in plastic.

All together now
Here it is with the mirror on it. First straight, then tilted. What's causing the tilt is just a 9" strap of metal hanging off the end. That is good because it shows very little weight is required to tilt the mirror, which means motors can be small.

So next steps are to mount two motors on the underside of the mirror. Attached to their shafts are 1/4"-20 threaded rod. This rod will then thread into an un-moving nut, so as the motor turns the rod, it will pull the mirror down or push it up relative to the fixed nut. Basically it's a worm gear on the cheap. I was excited to find these stepper motors for $2.75 each as they already had a threaded rod on them. But it turns out its like M7 or M8 with a .8 or .7 pitch -- not so easy to find coupling rods to 1/4"-20. The problem here is that as the angle of the mirror changes so will the angle of the motor shaft into the fixed nut, so it requires a universal-joint or some such accomodation. Gotta think about this to keep it cheap (hinges with springs?).

Still, as a hardware platform it's coming along and it cheapened up significantly with this v2. More progress on motor integration soon I hope, then the action moves to the software side.

Mirror Experiment & Housekeeping

2010-01-24T20:02:00.000-08:00

I sort of did the mirror experiment. Bought the ultra cheap mirrors from Oriental Trading Company. Pretty disappointing. If a good mirror reflects 95%, I'd say these reflect 70% at best. They're cloudy plastic. Nevertheless, in the 15 minutes I had to do it I measured 5 suns (ignoring cosine loss). Outside ambient temp, about 65F. Temps measured with an IR temp gun something like this.

Readings:
0 suns: 65F (white plastic, in the shade)
1 sun: 74
2 suns: 76
3 suns: 81
4 suns: 85
5 suns: 93
Experiment very far from perfect, needs to be done with more time and care, better mirrors, more mirrors, hotter ambient temps. But, temp differences were less than I expected.

Also, this blog now resolves to www.heliostats.org. I owned the domain (you can see older material there by dropping the www prefix) so figured why not class things up a bit. My wife did it as her contribution. Thanks J!

Mirror Mount

2010-01-24T19:50:00.000-08:00

I'm convinced that the specifics of the mirror mount are important. A simple, efficient mount which bears almost all the weight of the mirror and includes a counterbalance system to minimize the force required to move (tilt) the mirror is the key. If well done, the motors which move and hold the mirror in place can be minimized.

Motors are one of the key cost variables on the project. My target motor is a cheap low torque stepper. Cost goes up fast when more torque is needed. I'm also hoping to integrate a worm gear to trade speed for torque. Also worm gears have a certain holding force even at rest. I hope to use that to hold the mirror in place under ordinary conditions, so power is used only when it's moved (once every few minutes). Since power generation is the ultimate goal, the system should be miserly with power.

I'm investigating two mount directions at this point. First is the ball and socket, second is the "gyroscopic drink holder", both pictured below. I like the "gyroscopic drink holder" better. It was literally sold as a device you could attach to your dashboard with double stick tape then put a hot cup of coffee on your dashboard while you drove to work. It would keep it level so it didn't spill while you drove around corners and over hills. Comedy. My friend found it at a garage sale.

It's not a gyroscope but nested curves with ball bearings in between. You can see in the picture plastic ribs on the lower bowl/curve create six pie wedge sections, one for each ball bearing. I'm pretty sure there's a also a rib ring that keeps the bearings from rolling down to the bottom. So the top bowl moves smoothly over the bottom bowl. Then, the red screw is attached to a weight which hangs below the second bowl. The weight is the counterbalance so the top bowl always wants to stay level. To make it, I'm trying to find some nesting bowls, put some ball bearings in between and hang a counter balance. The mirror would go on top facing directly up, but could tilt in any direction, the counterweight pulling it back to level.

The ball and socket is easier to get parts for. I bought some ball thrust bearings (the kind used for lazy susans, not the radial motion type) from VXB. First pic is how the bearing would look on top of a sphere, next pic is how the mirror frame would sit on top of the bearing on a sphere. It's easy enough to buy the spheres, and this bearing type would work on them. This is a second choice at the moment because even with a counterweight, once the mirror moves past rest point, the mount itself makes it want to keep going. Whereas with the bowl style mount, when out of rest position the dynamics of the mount itself make it want to go back to rest. More potential energy at the rest point of this mount as my high school physics teacher would say.

Anyway, just posting as an update. Need to make something and post pics and how to's. Meanwhile, working in the background on sourcing motors.

Should also note that for v2, the configuration is an array of mirrors reflecting light up to a target that is raised but in the center of the array -- more like Ausra in the layout. Whereas v1 was more for an array all to one side of a target, reflecting back toward the sun -- more like a classic power tower. The difference is because how I guess retail solar (especially lightly concentrated PV) might play out. So v2 has a rest position for the mirror at something close to the horizontal, where the rest position for v1 was something close to vertical.

Version 2

2010-01-18T15:09:00.000-08:00

After dithering about whether to make v1 operational or work on version 2, I'm going to do some work on version 2. V2 design has been going for awhile in my head so may as well get it rolling in real life.

The core design issues to work through (using small motors to their best advantage and a cheap, simple hardware design that leverages software) are the same for either version, so there isn't much lost by resolving them in v2 instead of v1.

Version 2 in a nutshell ("Help, I'm trapped in a nutshell"):

Base: $20. Same concept as v1: unistrut or a similar component based system.

Mirror mount: $10. This could be a universal joint, ball and socket platform (this and this or just this), or something equally simple, cheap and effective.
This is a ball and socket part of a monitor mount that seems ideal, but I couldn't find a component part:

Or, here's a sweet gyroscopic drink holder from the 80's that would work great and probably be cheap. I just can't find/figure out how to make it cheaply.

I just spent hours looking for a good mirror mount on the internet with some leads but nothing perfect yet.

Motors: $20 for two. Ideally these are small, sealed stepper motors with worm gears easily integrated. Well, anyway, a boy can dream.

Mirror and frame: $30. Roughly 1 square meter of mirrorized acrylic or whatever else works.

Microcontroller, motor controller and wiring: $10 (amortized over 10-20 mirrors, so $100-$200 total).

Installation and profit: $20. This is more or less a random number.

V1 used the a motor to rotate the mirror around a center mounting shaft. V2 also has a central shaft, but it's just a weight bearing point (no motor there). There are two other points of connection to the mirror, one on the N/S axis, one on E/W axis, with a mirror attached to each point. Between the three points (one static, two moving), they define a plane which reflects the sun to the target.

I'll figure a way to post some drawings to make the design clearer. Or better, just post some pics as it gets built.

$100

2010-01-18T15:06:00.000-08:00

This is a picture of what my dad gave me for Christmas to "use this for your special garage project". Nice that it should buy exactly one heliostat. I will probably save it until it does.

Lightly Concentrated PV (cont)

2010-01-13T07:04:00.000-08:00

Following up on Monday's post, here's an excellent tutorial on PV provided by the US Dept of Energy. It discusses band gap, temperature effects, doping, etc. About 50% of sunlight that hits solar panels is outside the band gap and could perhaps be filtered by sheet films like these, though there are some off-axis issues with most filters.

In other news, Oriental Trading Company has 6" square mirror tiles for cheap. I'm still in.

Playing Around With Lightly Concentrated PV

2010-01-11T21:51:00.000-08:00

The application I've been thinking about lately is concentrated light on solar panels. Usually, when you are talking about concentrated solar for PV (photovoltaic), you mean high concentrations (100-500 suns) focused on a few square centimeters of high efficiency (and expensive) triple-junction cells. I'm looking at much lower levels of concentration, say, 10 suns, concentrated on something very close to a normal solar panel.

According to backyard experimenters, typical solar panels seem to put out 1.5-2 times their rated output under concentrated light but the higher temps generated by multiple suns can ruin the panels. Though reflecting light onto heliostats is done frequently, I haven't seen the basic info of how hot surfaces get under 2x sun, 4x sun, 10x sun and so on -- preferably under different weather conditions. I'd like some data on how hot things get under multiple suns. Given this, one could pair mirror configurations to solar panels operating parameters.

We know too that solar panels waste most of the of solar energy that hits them since they're only sensitive to specific light frequencies (wikipedia et al). Ideally, one could concentrate, say 10 suns, on a PV panel. Prior to the panel, there would be filtration films to eliminate IR or UV rays and so reduce the heat load. The ideal configuration would be 10 suns filtered precisely by a film down to the small band of frequency that the panel uses. If 80% of the heat were filtered out (overly optimistic? yes.), then you're down to 2x sun. Tweak the panel to handle high heat and higher currents and go for it.

Admittedly, this is an experiment undertaken with plenty of ignorance There are multiple effects to worry about if you actually focused 10 suns on an off-the-shelf solar panel for a chunk of time. Overheating the panel say, or size of the wiring at the back of the panel. Would the panel need to be excessively doped to produce the desired current? Sure, who knows.

But for all it's flaws, I still want to focus 10 suns on a brick wall and measure the temperature. Then, add a filter and see how much it helps. I'll let you know the results. Then, back to heliostats.

Why build a cheap heliostat?

2009-12-29T04:39:00.001-08:00

Heliostats have been around a long time. Wikipedia gives the usual good introduction. Building them is not an engineering challenge, it's done all the time. Building them cheaply is a challenge and it's one where success contains huge potential benefit.

This Sandia Labs study (section 6.1) estimates that if heliostats can be made for under $100 per square meter, solar power could be made in a power tower configuration for 6 cents/kW-hr, which is competitive with conventional wholesale power generation. Right or wrong, a hundred bucks per square meter (of mirror) is a sensible target. It's a nice round number that represents roughly the state of the art.

Is there any reason to expect it's possible to make working heliostats that cheaply? Several companies (eSolar, Asura, CoolEarth Solar,...) all seem to think so and some of them may have already succeeded. So it's not unimaginable. Is it reasonable to think internet hobbyists could do so? Well, the last 40 years have brought plenty of advances to simplify small scale innovation -- the internet (for shopping and sharing information), global manufacturing and distribution that lets you buy things cheaply, huge advances in the power and availability of microcontrollers & motion control (robotics, motors,...), a recognized practice of public design (open source software and hardware) -- the list is long. So, possible or not, it's hard to think of an easier time to try it.

Further, heliostats created for commercial power towers are only one configuration. A retail configuration has different design requirements and price points. Retail power costs about twice wholesale power, so from that perspective, there's more money to work with. On the other hand, a house in the USA spends on the order of $1,000/year in electricity costs (if gas heated and not including an electric car). That's pretty low -- less than what some people (like me) spend on things like phones and internet access. On the third hand (or other foot), retail systems also cost much less to install. So if they make economic sense to an individual, people can pay for it themselves, without raising commercial capital to build a large plant. Commercial or retail, how you are going to use the heliostat -- the specific configuration -- becomes integral to the design pretty quickly so there is more to say on this topic.

Anyway, that's the short case for why this project is worth trying.

V1: frame and mirror

2009-12-09T21:38:00.000-08:00

The frame was made from 1 1/2" aluminum angle. It was assembled by my friend who is handy. I provided consultation and beers. It is 4' by 2', so the long sides are 4' pieces, the short pieces are 2'. It was cut with a hacksaw and screwed together with self-tapping screws. Here's a pic.

There are a lot of holes in it because the original plan was to use bolts which could be adjusted to push against the mirror at strategic points and bend it into parabolic like shapes for tighter focal points. It didnt' work that well. A 12' piece of aluminum angle is about $11, though I used scrap I found.

Mounted to the frame was the mirror, which was a 4' x 2' piece of mirrored acrylic. Here's a site that has it for $34, I think this is where I got mine. In the end it was fixed to the frame by two bolts in the center and another two bolts in the middle at the top and bottom. When drilling the holes for the bolts, we cracked the mirror. To avoid cracking it, use a high drill speed but drill slowly, don't rush. If doing it again, without trying to make a bunch of standoffs to shape the mirror, I might just do a hole in the middle and the four corners or something more sensible. Use washers against the mirrored edge.

Here's another pic of the center.

The t-strap I added at the end, that's where the altitude adjustment bolt attaches. The other two holes are used to fix the frame to the hinge. Placement of these is important because you want the supporting pole to fix to a point close to the middle of the mirror so it's well balanced (minimize work required by motors). Find center line of the frame, then off-set those holes so that the hinge pin falls along the center line.

V1: movement

2009-12-07T19:31:00.000-08:00

Precise and controlled movement of a mirror is the bread and butter of a heliostat. All the heavy lifting was done by the folks over at servocity.com. As mentioned below, I bought my way out with this. Though too expensive for a production solution, this piece solves several problems.

First, it provides a motor mount and mechanism for suspending the motor to the shaft. I bought the 90 degree arm to attach it, see the pic. The 90-degree arm can be fixed to the gearbox on the side. Bought small (M6 maybe?) nuts and bolts to fix it on there. Reserve the screws that came with it to fix something else to the large gear. Attached to the other end of the 90-degree arm is a 3" t-strap (Stanley) which had existing holes which aligned with the arm and the threaded rod. The T strap is fixed to the threaded rod with a bolt above and below and serves to hold the motor fixed in a given orientation so when the gear turns it can move the mirror rather than just spinning the motor mount around the center rod.

The second (and best) thing the Gearbox does is integrate a gear (to step down the motor) and a mounting platform with ball bearings which simplifies transferring the rotation of the motor shaft to the mirror. The mounting platform includes a hollow shaft which the 1/4" threaded rod passes through. Note that when assembling, there needs to be a nut already threaded onto the rod that the T-strap will rest on. Next, slide the T-strap/motor frame onto the threaded rod, but after the T-strap passes on, screw another nut on, and screw it down as the T-strap and then gear shaft threads over the rod. (Just order of operations -- otherwise you can't get get the 2nd nut on after the motor is fitted over the threaded rod).

Next step is to add hardware to attach the mirror. Buy 1 1/4" perforated angled iron (Home Depot or wherever) and cut a piece about 2.5" long. The piece I got alternates hole sizes. The circular hole is exactly 3/8" and so it fits snugly over the shaft provided by the servocity gearbox. Around the 3/8" hole, drill 4 holes using the 90 degree arm as a template to show hole placement and size. The drilled angle iron looks like this.

Use the screws that came with the gearbox to screw the angle iron to the gear (those holes are threaded). I only got two screws to thread, probably because my holes weren't well aligned. Also, because the 90-degree mount is 1/4" thick, the screws don't go flush to the top of the angle iron. If this part of the design sticks around it'd be worth adding washers and nuts as spacers.

Next, the hinge. This was a normal, 3" door hinge from the hardware store. The idea is that the bottom part of the hinge is fixed to the angle iron and the top part of the hinge floats free. The mirror frame will be attached to the top part of the hinge so it's still free to pivot to account for the sun's altitude (seasonal movement), but the main motor will move the fixed part of the hinge for daily (azimuth) movement. (As shown it's attached at only one point -- it really needs to be two points because as-is there's too much play).

Here's another pic from the back so you can see the hinge is free to pivot. What this pic also shows well is the big, expensive ($40!) modification I did to the hinge. I paid a welder to weld on an in-line ball joint to the center portion of the hinge. It shows in the pic as the blackened portion at the center of the hinge and hanging down from that is a shaft. That shaft receives the 1/4"-20 threaded rod. At first, I had the rod just rest on the hinge, but the problem with that is that single point is bearing most of the weight of the mirror. Any misalignment caused it to really want to fall off. The ball joint allows it to pivot (daily movement) but otherwise holds the hinge (and so the mirror frame) tight and bears the weight better.

Here's a pic of the in-line ball joint. I got it from midwestcontrol.com. Here's the link. I had the guy cut off the bolt portion before he welded it.

Note that everything is based off 1/4"-20 once you get above the 1/2" frame part. This allows you to buy a variety of bolt sizes, washers and nuts and have some flexibility putting it together.

OK, so the two things remaining to do are attach the mirror frame and add a part for altitude (seasonal) orientation. The next pic gives a view of the mirror frame attached.

The mirror frame is attached by the two top nuts/bolts. Pre-drill two holes in the mirror frame to align with the existing holes in the hinge. Then fasten with the usual 1/4"-20 nuts and bolts. This picture is from an earlier version so you also see that it used two bolts to fasten the lower 1/2 of the hinge to the angle iron. Pre-drill the holes through the hinge for this. This was a better method than the final version, it had a lot less play in it. Also, this was before the ball joint was welded to the hinge, so you see it resting on the head of a long bolt which was serving as the top threaded rod.

Some of what's above got changed when the mechanism was added to fix (and adjust) the altitude of the mirror. To get that, I added another angle iron, which can be seen in the next pic.

OK, apologies, I missed taking a side angle which would have made this clearer. This picture is taken from the other side, ignore the motor position as that changed. Remember, the pic directly above is an older version, instead, go two pics up to where you see my hand. In that pic, you see the green hinge is on the inside of the angle iron (between the angle iron and the shaft). So the second angle iron is same size as the first, and is attached to it on its vertical face like a mirror image. You are taking two L's and putting them back to back to form a T (upside down).

Once you are oriented within the pic, you can see the last piece added. This is another ball joint (hasn't had it's bolt portion cut off). At the bottom, a good old long 1/4"-20 bolt holds it to the 2nd angle iron. At the top it's fixed to another 3" T-strap which is fixed to the frame. By screwing this bolt up or down, it changes the height of the mirror relative to the 2nd angle iron and so changes its angle of altitude. The ball joint is needed to handle the pivot that happens at that point. The ultimate plan is for this to be a worm gear controlled by a second motor. For now it's hand adjusted.

Well that's it for this post. This is most of the value. But, more posts to follow on motor control, application, what was good or bad about this design and preparing for v2.

V1: the base

2009-12-07T19:14:00.001-08:00

OK easy part first. The base is unistrut, because it's easy to work with and tough. I bought a 10' piece and cut it (er, my friend cut it) into a 4' piece and two 18" pieces. You can buy the channel (long pieces) at Home Depot or the like. The fittings you can buy at plumbing supply stores or order special pieces on-line.

Here's a close up of where the channels intersect. There's a X-fitting which my plumbing supply didn't have so I used a T and a two-hole piece.

You can see a nut where the 1/2" rod goes through the cross, there's a matching nut on the underside of those metal fittings to keep it tight. Also, the nuts you see going into the fittings are coarse 1/2" made by unistrut, they mate with unistrut bolts with springs.

Cost for the 10' piece of channel was $22.70, the T fitting was $12, nuts and bolts $5.40. So, $40 for unistrut. Too expensive really but OK for prototyping. If this were a larger implementation, say 10-20 mirrors, it could be made into a grid so as a system it could be stiffer. To cut costs, smaller channel pieces (13/16") could be used, or maybe a cheaper, plastic substitute. Also, the L and T pieces are pricey and in a grid it might be possible to use flat 3-hole pieces. All in, $20/ device (per square meter) for an easily installed base might be the target.

The threaded rod was a 2' piece of 1/2" -13 from Osh. This costs about $1.50/foot. It would be better to cut it down to 1' for stiffness (the mirror doesn't require 2' for clearance), but leave it at 2' when first building -- it's easier to get under it to work.

Here's a pic of where the 1/4"-20 rod is coupled to the 1/2". The 1/4" isn't stiff enough to use for the whole length, but 1/4" is needed to fit through the hollow shaft of the servo gearbox. The 1/2"-13 to 1/4"-20 reducer coupling nut is an oddball. I found it at tannerbolt.com, here's the link. I had to buy a box of 50, but they're 50 cents each in quantity. I have 49 left. The bolt you see in the picture is what the gearbox will rest on. Will tackle that in a next post.

Version 1: overview

2009-12-06T12:53:00.000-08:00

Here's a picture of attempt #1 (hereafter v1). This is an overview, I'll address each section in separate posts.

The base is made from uni-strut. Coming vertically up from the unistrut is a 2' piece of 1/2"-13 threaded rod. There's a coupling nut which steps the 1/2" rod to 1/4"-20 rod.

The motor and gears are this, which is a servo and gearbox by servocity.com. I'm using the 8.6:1 ratio which has metal gears. This threads over the 1/4" rod and is held in place by bolts on the rod itself.

The pivoting mechanism (daily and seasonal) is a door hinge. The hinge is bolted to a piece of perforated steel angle, and the steel is screwed to the gear. The steel also attaches to the mirror frame. Movement of the gear moves the hinge for daily (azimuth) movement. For seasonal (altitude) movement, the other portion of the hinge can pivot.

The mirror is mounted to a frame made of aluminum angle. The mirror itself is a 4' x 2' piece of acrylic fixed to the frame.

Movement is controlled by an Arduino board with a motor shield.

OK, now will try to do breakouts of each component.

Cheap Heliostats Project -- Hello World

2009-12-06T10:47:00.000-08:00

This blog is to document my attempt to build a cheap heliostat and share ideas so others can build on and extend them. There is a community of people working in an open source approach to solar power generation and this site is my attempt to join them.

I'll probably do some future posts on heliostats generally, what "cheap" means (sub-$100 per square meter), how heliostats can be used and so on. But for now, most of the people that would run across this site and care to read it would know all that. So, I'll just comment on design directions and move on with practical posts.

My ideal design would be assembled from mass produced parts and require little or no manufacturing or special tools to keep costs low and increase accessibility / modification of the design. It would a modular design so components of the system could be easily upgraded. It would contain no patented elements. I lean toward design for retail users, and a system that could be mastered fairly easily so barriers to entry for installation are low. Not all these goals will necessarily be met, but just so you know where I'm coming from. And now, on to more practical things.

Incidentally, this blog picks up from the website heliostats.org.

First Post

2009-12-01T20:53:00.001-08:00