Shop OBEX P1 Docs P2 Docs Learn Events
Solar Tracking Object - Page 2 — Parallax Forums

Solar Tracking Object

2»

Comments

  • jmgjmg Posts: 15,173
    edited 2012-01-01 12:29
    Regardless of which approach you use, the other factor to consider based upon the hardware is how often you move the panel. This "slew" rate can be be either continuous or only move periodically (when the angle is too big and power is being lost).

    I like the idea of simple timed-movements, but a heliostat will likely need a close to continual tracking.
    I would see the power-budget being close to run-time limited, so a long delay just means a longer eventual on-time - you have deferred the power-draw, not really saved much. (just motor rotor mass-acceleration, which will be small % )

    Couple to this that you have added a sensor (a real failure point), and we have the issue that an 'average user' could consider a 'smart' system actually broken, at a first glance. More support calls....

    The tolerances are so tight in heliostat's, I'm trying to think of a self-calibrate or feedback system.
    The best I can come up with, is a separate target on the tower, and any single mirror can be swept over that to check/correct effects like ground settling. (as well as confirm Mirror cleaning status, etc )
    Perhaps if every unit had a GPS installed, that might be enough to ball-park correct any swapped controller module, and then a calibrate pass would lock in the numbers ?
    .
  • KievlaninKievlanin Posts: 55
    edited 2012-01-01 13:30
    Well, it is looks like it will be beneficial to have a second axle (this can be called tilt axle).
    This additional power should be well enough to power the alignment mechnism itself.

    I am not very concern about the wind, this first prototype is to be relatively small with total supported PV panels area of being maximum of 4x4 feet.
    This will be a test unit for mechnical design, as well as electronic.
    Next unit will not be the same.
    I want to experimentally proove the economical advantage of having this unit on the yard.

    My recent experiments with actual PV panel proved importance of alignment presicion. As most power comes ONLY when panel looking directly at the sun.
    Any sugnificant misalignment could render the project as "money loser".
    The more power I can get from the panel the faster return could be.

    Of course if someone can install stationary panels on the roof (spending something in the range of $25000 per installation) and not count for the loss of potential electrical power...
    I can' t and I do not want to do it.
    10% of lost power means more then a month of (potential power) electricity a year.
    General rule is PV panel will pay for iteself in about 25 years... And I do no know is this with 7% loss or 30% loss? 25 years(!) to recover the cost...!
    No wander why I can't see a single PV panel on the roof within 10mile radius of Greenwood, SC...
    And this is in the area with almost 360 sunny days in a single year!

    People just do not see the advantage.


    Does anyone counted how much electricity is lost on non-aligning flat roof installations? Is it 30% or 50%.? Or even 60%?

    By the way if someone do not know (just one fact): USA is now exporting gasoline(!).
    And why? We did not reduced our supply of oil... We simply using less gasoline becouse we driving better cars. Instead of 15mpg we doing 20mpg or even 35mpg.
    We just do not need as much gasoline as we were needed 2 or 3 years ago!
    Imagine we can power 1 home computer per house in america.
    How many coal electrical turbines can be shut down?
    In 2009, the US Census Bureau reports all housing units totalling in at 130 million 159 thousand.

    Read more:
    http://wiki.answers.com/Q/How_many_single_family_homes_are_there_in_the_United_States#ixzz1iF9hRrxI


    My one laptop using up to 40watt/hour. This is a special 3D CAD design machine.

    Now: 130 000 000households * 40watt = 5 200 000kwt(!) = 5 megawatts This is only for one hour!

    The Three Mile Island Nuclear Generating Station in the USA has a rated capacity of 802 megawatts.
  • KievlaninKievlanin Posts: 55
    edited 2012-01-01 15:14
    jmg wrote: »
    I like the idea of simple timed-movements, but a heliostat will likely need a close to continual tracking.
    I would see the power-budget being close to run-time limited, so a long delay just means a longer eventual on-time - you have deferred the power-draw, not really saved much. (just motor rotor mass-acceleration, which will be small % )

    Couple to this that you have added a sensor (a real failure point), and we have the issue that an 'average user' could consider a 'smart' system actually broken, at a first glance. More support calls....

    The tolerances are so tight in heliostat's, I'm trying to think of a self-calibrate or feedback system.
    The best I can come up with, is a separate target on the tower, and any single mirror can be swept over that to check/correct effects like ground settling. (as well as confirm Mirror cleaning status, etc )
    Perhaps if every unit had a GPS installed, that might be enough to ball-park correct any swapped controller module, and then a calibrate pass would lock in the numbers ?
    .

    For the heliostat a precise tracking is much more critical then to PV Panel.
    For Solar Tracker we are seems to agree on 5 to 10 degrees misalignment is OK. Althought the more precise the better. But there is a trade-off. I do not want to keep alignment mechanism powered at all times.
    I am thinking about having it ON for 1 minute every 15 minutes.
    Then at the field testing maybe SLEEP time can be increased even more, to pass more energy to storage battery instead of extra alignment.
    This will depend on how much WATTS my linear units and electronics will "steal" from me....
  • Prophead100Prophead100 Posts: 192
    edited 2012-01-01 16:52
    The sleep mode combined with optimizing the pause between movements should go a long way. The propeller will a version of sleep mode that should work for you by temporarily stopping extra objects & cogs then slowing the clock to RCSLOW. This does two things, first it saves power and second it extends the length of time that the waitcnt command can be used to monitor the system without extra code. I have used this approach to drop the power use by 97% while waiting for a pin to go high indicating the system was needed.

    jmg was wondering why someone would use a two axis for a simple panel. That exactly what I am considering for a mobile panel (up tp about 8x8) on an RV/Trailer/Boat set-up where the direction to south would be changing periodically (I don't want to have to park the trailer any specific direction), the system must fold flatly, attachments must be keep low to the attachment surface and wind loading is a huge issue. The toe of the panel will by definition need to be as low as the mounting surface and the axis of rotation would need to swivel anyway.
  • Prophead100Prophead100 Posts: 192
    edited 2012-01-01 16:59
    P.S. You may want to look at running in RCSLOW mode anyway. Take a look at the datasheets for the Propeller chip and look for the power curves relative to process rates (e.g. MPS or clock speed) and number of cogs running. Use the full processing power when you need it (e.g. communication, status displays, storing data for the day, etc) then throttle back to low power when you don't (monitoring & slow calculations). You can also have the prop turn peripherals such as sensors, displays and other chips off when not needed as well.
  • Duane C. JohnsonDuane C. Johnson Posts: 955
    edited 2012-01-01 17:47
    Nit Pick mode on!

    Jmg wrote:
    The largest area, simplest mechanics, Solar trackers, have a long common horizontal axle.
    Then you have tilted axis units, but they can still use a robust/reliable axle design.
    We would rarely use a single horizontal N-S axis tracking mounts for PV panels hear in the temperate latitudes.
    They would be mostly relegated to the tropic latitudes.
    And never use a single horizontal E-W axis tracking mount.

    What I think you meant was a Vertical Axis tracking mount which rotates about the horizon or AZimuth.
    Generally, these are not very suitable for PV panels unless a secondary tilt axis is added which moves up from the horizon in ALTitude.

    The best single axis tracking mount for PV panels is a Polar Axis tracking mount where the axis is aimed at the celestial pole, close to the north star, and the motion is in Right Ascension or Hour Angle.
    A secondary axis can be added for seasonal motion. This is a tilt axis on the main polar axis and its motion is in DEClination. However this can add a maximum of about 7% energy per year. But if declination is manually adjusted 4 times per year you get about 5% of the 7%. The great majority of energy increase is in tracking the main axis in Right Ascension or Hour Angle

    Some cheat a bit with polar axis mounts for PV panels by tilting the polar axis. This is a bit simpler to make. I call these "Pseudo Polar axis mounts".
    Note! Pseudo Polar axis mounts are not suitable for concentrators which require True Polar axis mounts.

    Kievlanin wrote:
    My one laptop using up to 40watt/hour. This is a special 3D CAD design machine.
    This makes no sense. Laptops are rated in units of power, in your case this would be 40watts. Not watts per hour.

    Kievlanin wrote:
    Now: 130 000 000households * 40watt = 5 200 000kwt(!) = 5 megawatts This is only for one hour!
    No this would be simply 5 megawatts. Time is not required.

    Nit pick mode off!

    Most heliostats use vertical dual axis type tracking mounts.
    Howeve, there is another more obscure type.
    These are called Receiver Axis heliostat tracking mounts.
    Basically, the main axis is aimed directly in the direction where you want the light to go, at the receiver or through a window. The secondary tilt axis bisects the angle between the sun and the receiver.

    Duane
  • KievlaninKievlanin Posts: 55
    edited 2012-01-01 20:33
    Nit Pick mode on!

    Jmg wrote:

    We would rarely use a single horizontal N-S axis tracking mounts for PV panels hear in the temperate latitudes.
    They would be mostly relegated to the tropic latitudes.
    And never use a single horizontal E-W axis tracking mount.

    What I think you meant was a Vertical Axis tracking mount which rotates about the horizon or AZimuth.
    Generally, these are not very suitable for PV panels unless a secondary tilt axis is added which moves up from the horizon in ALTitude.

    The best single axis tracking mount for PV panels is a Polar Axis tracking mount where the axis is aimed at the celestial pole, close to the north star, and the motion is in Right Ascension or Hour Angle.
    A secondary axis can be added for seasonal motion. This is a tilt axis on the main polar axis and its motion is in DEClination. However this can add a maximum of about 7% energy per year. But if declination is manually adjusted 4 times per year you get about 5% of the 7%. The great majority of energy increase is in tracking the main axis in Right Ascension or Hour Angle

    Some cheat a bit with polar axis mounts for PV panels by tilting the polar axis. This is a bit simpler to make. I call these "Pseudo Polar axis mounts".
    Note! Pseudo Polar axis mounts are not suitable for concentrators which require True Polar axis mounts.

    Kievlanin wrote:

    This makes no sense. Laptops are rated in units of power, in your case this would be 40watts. Not watts per hour.

    Kievlanin wrote:

    No this would be simply 5 megawatts. Time is not required.

    Nit pick mode off!

    Most heliostats use vertical dual axis type tracking mounts.
    Howeve, there is another more obscure type.
    These are called Receiver Axis heliostat tracking mounts.
    Basically, the main axis is aimed directly in the direction where you want the light to go, at the receiver or through a window. The secondary tilt axis bisects the angle between the sun and the receiver.

    Duane

    Duane, I have measured the watts on my computer using "KILLAWATT" device. The computer was connected thru this device.
    Compaq NW8000 rated: 18.5v at 3.5amp by factory. Comes with 65watt power supply. Actual drain is about 40watt when not running large 3D assembly.

    Correct me if I am wrong: if I have a 100watt light bulb. How much energy I will use in one hour of use of this bulb?
    Going back to basics.
    Now if I have automotive battery rated at 75amp/h. This means theoretically(!) I can drain this battery for 75hours using 1amp/hour. Is this correct?
    I do know that we do not want to drain battery below 50%.
  • KievlaninKievlanin Posts: 55
    edited 2012-01-01 20:50
    When we are paying for electricity....
    We are paying for kw/hour.
    If I used a light bulb of 100watt for how much in kw I am paying?

    What if I used a 1000watt bulb?

    I just checked my bill. One kw cost me $0.10
  • Duane C. JohnsonDuane C. Johnson Posts: 955
    edited 2012-01-02 05:30
    Hi Kievlanin;
    Correct me if I am wrong: if I have a 100watt light bulb. How much energy will I use in one hour of use of this bulb?
    Going back to basics.

    You will multiply 100 watts times 1 hour and get 100 watt hours. (100W * 1hr = 100Whr).
    Now if I have automotive battery rated at 75amp/h. This means theoretically(!) I can drain this battery for 75hours using 1amp/hour. Is this correct?
    I do know that we do not want to drain battery below 50%.
    But your battery is not rated in amps per hour. It is rated in amps * hours.
    So,
    75Ahr / 75hr = 1A
    75 / 75 = 1
    A * hr / hr = A

    Just as you can perform math operations on the "Numbers" part of an equation you can do math operation on the "Units" parts also.

    It would make no sense if you use 75A/hr
    (75A/hr) / 75hr = 1A/hr^2
    75 / 75 = 1
    (A / hr) / hr = A/hr^2
    When we are paying for electricity....
    We are paying for kw/hour.

    Again, this makes no sense.
    What you are actually paying for is kWhr, not kW/hr.
    If I used a light bulb of 100watt for how much in kw I am paying?
    I just checked my bill. One kw cost me $0.10
    You don't pay for kW, (A unit of Power), you pay for kWhr, (A unit of energy).
    In this case one kWhr costs 10 cents.
    Let's say you run this bulb for 7 hours:
    ((100W / (1000W / kW)) * 7hr) * $0.10/kWhr = $0.07 or 7 cents
    ((100 / (1000)) * 7) * 0.1 = 0.07
    ((W / (W / kW)) * hr) * $ / (kW * hr) = $
    What if I used a 1000watt bulb?
    Let's say you run this bulb for 7 hours:
    ((1000W / (1000W / kW)) * 7hr) * $0.10/kWhr = $0.70 or 70 cents
    ((1000 / (1000)) * 7) * 0.1 = 0.70
    ((W / (W / kW)) * hr) * $ / (kW * hr) = $

    II am a stickler for the use of proper units when describing power and energy.

    An example:
    Many people use the term "Solar Energy" but this is incorrect, the sun only delivers "Solar Power".
    There is no such thing as Solar energy, it doesn't exist here on earth.
    The sun outputs mostly photons of light. This is power not energy since you can't store light over time.
    Sure, once the light hits your solar panel the photons are converted to heat or electricity. But this heat or electricity is indistinguishable from other sources of heat or electricity.

    Lastly:
    The term "per" means "divide", as in miles per gallon miles divided by gallons.
    The term "over" means "multiply", as in kilowatts over time or kWhr.

    Duane
  • KievlaninKievlanin Posts: 55
    edited 2012-01-02 08:23
    Duane,
    We are coming to the same conclusion.
    1kw bulb for 1hr will cost $0.1
    100wt for one hour will cost $0.01

    What are we argue about?

    I just trying to stress the point of importance of alignment to the sun and how to achive this.
    With the currect technologies on mechanical side I see no limits, except possible high cost.
    On this side there is so many variations it is hard to even imagine. Strength of construction is not a problem.

    Real challenge is to make it cheap, but w/out mass production it is hard to achive.
    This is why my approach is not to make it large.

    There is a need to allow people to start using Solar Energy for something small, like phone chargers, computers, DVD's. To have one wall outlet to be powered with "free" energy. People need to get use to it, like we now using Toyota Prius. (Most don't even know how it works) All they know it is 45mpg and they like it.

    And this should be something easy to assemble and connect by home owner himself. W/out the need of full installation crew.

    (Prophead100 is talking about mobile 8'x8' system). I think it is way to big for being mobile or managable by a regular person.

    System not necessarily need to be mobile, but it need to be easy to install & maintain w/out bachelor degree diploma.

    On electronic side (for me) is much more challenging. At this point we are getting close to high science. Astronomy, math calculation, timing.
    And I am not very good helper there.
    This need to be done by a team of engineers. This is what is called mechatronics.
    A unit must be developed, small, working from 12v battery. Reliability is super important. You want to install and forget about.


    This unit should be able to send signal to more then one Solar Tracker.

    I see several development stages there:
    1. Developing an algorithm/programming.
    2. Choosing electronic components
    3. PCB development
    4. Field testing
    5. Arranging all components in the small presentable box.

    Installer of Solar Tracker only need to be able to initally align the Tracker according to compass, and aligning it vertically. If it is a permanent installation.
    As for mobile mounting it is need to be slightly different.
  • KievlaninKievlanin Posts: 55
    edited 2012-01-02 08:39
    http://www.suntrack.es/index.html

    Question for all:
    Does anyone know the above "linked" company.
    They never answered my questions, but from what they discribed they already developed the system we are talking in this blog.
    I was never able to get the price from them.
    Looks like they are keeping this as a big secret.(?!) Why?
    It is also hard to get in contact.
    But the description of principal of operations is just what I can dream about.
  • Duane C. JohnsonDuane C. Johnson Posts: 955
    edited 2012-01-02 09:37
    Kievlanin wrote:
    We are coming to the same conclusion.
    1kw bulb for 1hr will cost $0.1
    100wt for one hour will cost $0.01
    Sure, I divined what you meant but that doesn't mean what you said was correct.
    What are we argue about?

    What I was trying to do was to show that weird units like "W/hr", "kW/hr", "A/hr", and the like make no sense.
    Or confuse "W", (power), with "Whr", (energy).
    I am just instructing the forum that units have meaning. Confusing units make it difficult to understand if one is talking about power or energy.

    I have had countless discussions with people only to find out they were confused about units which resulted in large errors in their calculations.

    Duane
  • jmgjmg Posts: 15,173
    edited 2012-01-02 16:34
    jmg was wondering why someone would use a two axis for a simple panel. That exactly what I am considering for a mobile panel (up tp about 8x8) on an RV/Trailer/Boat set-up

    Interesting examples, and for both these I imagine a turntable+Tilt design would be the lowest parked-profile, and strongest.
    but again you compete against fast-falling PV prices, vs complexity/maint/reliability, and for casual-usage PV, that can be more if a pain, than on a commercial installation.

    You can see this, comparing the Nov 2011 with Nov 2010 prices

    http://www.solarserver.com/service/pvx-spot-market-price-index-solar-pv-modules.html

    Shows some large price falls, even in the last couple of years - certainly makes the 5-7% gain over one axis marginal and even the ~40% gain
    over very-simple-fixed, is not many months price declines...
  • Prophead100Prophead100 Posts: 192
    edited 2012-01-02 22:34
    jmg wrote: »
    ...-falling PV prices, vs complexity/maint/reliability, and for casual-usage PV, that can be more if a pain, than on a commercial installation.

    You can see this, comparing the Nov 2011 with Nov 2010 prices

    http://www.solarserver.com/service/pvx-spot-market-price-index-solar-pv-modules.html

    Shows some large price falls, even in the last couple of years - certainly makes the 5-7% gain over one axis marginal and even the ~40% gain
    over very-simple-fixed, is not many months price declines...

    Thanks for the link. It is great to see costs coming down such that more PVs area will be an increasingly cost effective option where space and configuration will allow it.
  • Prophead100Prophead100 Posts: 192
    edited 2012-01-16 18:29
    Keeping in the spirit of how one can use the prop to aim a panel or relfect the sun.. ... I updated the zip file at the beginning of this thread to include a demo that traces a noon day Analemma and another that uses servos to aim a heliostat (mirror) to reflect the sun to a target.
  • KievlaninKievlanin Posts: 55
    edited 2012-01-17 05:32
    Keeping in the spirit of how one can use the prop to aim a panel or relfect the sun.. ... I updated the zip file at the beginning of this thread to include a demo that traces a noon day Analemma and another that uses servos to aim a heliostat (mirror) to reflect the sun to a target.

    Prophead,
    I am not a programmer, honestly.
    It is very hard for me to understand. I am "several floors below"....
    It is much easier for me to make a complete 2D/3D design of a log gripper then do what you do...We need to combine our knowledge and make a working Solar Tracker unit.
    I am still waiting for the machining components. Hopefully at the end of next week I will start final assembly.
    So far I have been experimenting with simple BS1 and photoresistors...
    Anyway, I am concentrating myself on something I can do better, mechanical design & machine parts.
    Once I have unit done I will need electronics/programming... to complete the unit.
  • Prophead100Prophead100 Posts: 192
    edited 2012-10-12 22:27
    FYI - The object ( http://obex.parallax.com/objects/807/ ) has been updated to include a method to calculate the approximate direct beam irradiation, indirect irradiation, and air mass consistent based upon time, location and elevation using formulas from Meinel AB, Meinel MP. Applied Solar Energy.; 1976 p. References are included in the documentation.

    Why? it gives you the option to estimate the solar strength at any time with consideration elevation ignoring clouds. If you use the solar trace demo to see when the sun will hit a particular location then you could integrate the values it will give for cumulative heat for a particular day. Its not included in the demo but its hidden in there with other functions such as methods to run mechanicals.... - Prophead100
  • Jim FelichJim Felich Posts: 2
    edited 2013-01-13 17:45
    FYI - The object ( http://obex.parallax.com/objects/807/ ) has been updated to include a method to calculate the approximate direct beam irradiation, indirect irradiation, and air mass consistent based upon time, location and elevation using formulas from Meinel AB, Meinel MP. Applied Solar Energy.; 1976 p. References are included in the documentation.

    Why? it gives you the option to estimate the solar strength at any time with consideration elevation ignoring clouds. If you use the solar trace demo to see when the sun will hit a particular location then you could integrate the values it will give for cumulative heat for a particular day. Its not included in the demo but its hidden in there with other functions such as methods to run mechanicals.... - Prophead100

    Prophead,

    I have been Goggling "calculated solar positioning" all day and stumbled onto your thread.
    It seems you and Mr. Johnson have been at this for some time.

    For my part, I have been developing (for my own use) an eight PV panel two-axis tracker using solar sensors for positioning.
    I am using a Microchip PIC18F4550 as the MCU for the solar tracker.
    I have decided that I also want to include the ability to calculate the proper position as well.
    This is a case of "belts and suspenders", but it will eliminate dealing with most of the edge cases that can occur with a solar sensor.
    These include: occluded sun, fog, solar eclipse, and such.
    These all expose the possibility of the tracker needlessly hunting for the sun, possibly with bad consequences.
    I see no problem using a Propeller to feed the calculated position to the Microchip primary controller.
    I also have an old GPS chip around so I can give the Propeller the correct date, time, latitude, and longitude data.

    To sum up, I want to thank you for your work and I will be monitoring this thread to keep up with your progress.

    One last thing: I did not see any information on the mechanical setup you are using for your demos and testing.
    Can you please post some photos and/or diagrams of your apparatus so that I may duplicate it for my own testing and verification?

    I thank you in advance for any help or advice you can provide.
    I, in turn, am glad to show you any of my work if you are at all interested.

    Sincerely,
    Jim
  • Prophead100Prophead100 Posts: 192
    edited 2013-01-13 22:19
    You can see the initial test rig at http://forums.parallax.com/showthread.php/138941-3_29_12-Presentation-Files-and-Topics. The code is fairly heavily documented so translation for a PIC should be straight forward. You could also look at the C version on the OBEX although it is less documented. Although it sounds like your familar with the PIC, you may find the Propeller able to do the work of the PIC as well. The code leaves plenty of space to also run stepper motors, read sensors, etc,,, Good luck.

    For full scale systems, I would look at Duane's Redrock Website.
Sign In or Register to comment.