©±°²³µ¼½¾÷ø
|
 * LED3X Solar Tracker A simple, accurate, low cost, single axis electronic solar tracker based on using green LEDs as photovoltaic light sensors. |
Go back to Red Rock Energy.
I use Electronics Workbench Personal Edition.
This program includes circuit design cad, simulation, and auto routing PC layout. I also use PSPICE at Unisys which has more capabilities but costs 6 times as much and needs to be renewed yearly.
apc
AP Circuits
Canada
Prototype Printed Circuit Board Fabrication Since 1988
twinstar
Twin Star Inc.
Excelant 1 day turn around policy for prototype quantities. And their local to me.
4"x10" $100
10"x10" $175
What a deal!
trc
TRC Circuits
These guys make boards for me when large quantities are needed.
olimex
OLIMEX Ltd.
89 Slavjanska St., P.O.Box 237, Plovdiv 4000
Bulgaria
Printed Circuit Board (PCB) Prototypes
expresspcb
ExpressPCB
They give you a proprietary PC board cad program that allows quick turn around board processing.
The program is easy to use if not overly simplistic.
Printed Circuit Board (PCB) Prototypes
far
FAR CIRCUITS
Printed Circuit Boards
Thy make double sided non plated through service. No email!
Custom boards may be made for $.50 square inch or $4.00 minimum per board for single sided, etched, drilled and solder coated. Custom double sided non-plated through holes, etched, drilled and solder coated boards may be made for $.75 per square inch or $6.00 minimum per board. A clear film negative is required for custom boards. Film from camera ready artwork is $8.00 per shot based on 8" x 10" film per shot. Quantity pricing available. Far Circuits reserves the right to change the pricing based on the complexity of the board.
I generally design circuitry that uses surface mount soldering techniques.
This link has a wealth of information about surface mount package shape and connection information.
Prototype of an 8 Bit Experimental Bit Serial I/O.
I needed a universal I/O device for my computers. I had several requirements:
| 1. | The ports may need to be as much as 2000 ft from the computer. |
| 2. | Ideally the I/O ports need to be easily expandable from the same control output. |
| 3. | Speed should be medium. Capable of up to several thousand bytes per second. |
With these requirements in mind, and some others, I decided on a variation of National Semiconductors "Micro Wire"(TM). In general this is called a "bit serial" interface. The interface is composed of shift registers connected in series. Each port is in a chain. Data bits are sent down the chain through the shift registers for output. Input data bits are similarly sent down the chain to be inputted into the computer.
My standard bit serial I/O consists of 6 wires. The wires for the standard version use an RJ11-6 6 conductor flat telephone type cable. I also have a simplified 4 wire version. This 4 wire version is predominantly used for output devices only. Pins 5 and 6 are not used. This version eliminates the data input channel and has no self powering capability.
Bit Serial Pin Designations.
| Pin # | Color | Name |
| 6 | Blue | Interface power |
| 5 | Yellow | Serial Data Bit In |
| 4 | Green | Serial Data Bit Out |
| 3 | Red | Serial Clock |
| 2 | Black | Data Strobe |
| 1 | White | Ground |
In some cases the printer port can supply all the power needed to run the interface.
IO
Bit Serial Interface Schematic.
Diagram of RJ11-6 Connector.
RJ11-6 Interface Data Input RJ11-6 Printer Port Adapter
RJ11-6 Loop Through Connector
-------------------------- --------------------------
| ------------------------ | | ------------------------ |
|| W B R G Y B || || B Y G R B W ||
|| h l e r e l || || l e r e l h ||
|| i a d e l u || || u l e d a i ||
|| t c e l e || || e l e c t ||
|| e k n o || || o n k e ||
||-- w -- | | -- w -- |
|| | 1 2 3 4 5 6 | | | | 6 5 4 3 2 1 | |
| --- --- | | --- --- |
| | | | | | | |
| ---------- | | ---------- |
-------------------------- --------------------------
View from the Cable side View from the Cable side
Cable to Plug Arrangement.
When choosing or building the cables that connect the computer and I/O boards together make sure they are the strait through type. In other words the cable has NO TWISTS in it.
________________ 6 Conductor < < ________________
| | | Blue 6| Flat Cable > > |6 Blue | | |
| | | Yellow 5|_______________< <___|5 Yellow | | |
| | | Green 4| > > |4 Green | | |
| | | Red 3|_______________< <___|3 Red | | |
| | | Black 2| No Twists > > |2 Black | | |
|_|___|___White_1| < < |1_White___|___|_|
> >
______________ 6 Conductor < < ______________
|___| RJ11-6 | Flat Cable > > | RJ11-6 |___|
_| PLUG |_______________< <___| PLUG |_
|________________|________________> >__|________________|
\ \ No Twists < < / /
\ \ > > / /
\ \ < < / /
-- > > --
Connection Diagram.
-------------
| |
| Computer |
| |------+ <-- Parallel Port Adapter
------------- |
|
------------- |
| I/O Board |------+ <-- Interface Data Input
| Highest |
| Addressed |------+ <-- Loop Through Connector
------------- |
|
------------- |
| I/O Board |------+ <-- Interface Data Input
| Middle |
| Addressed |------+ <-- Loop Through Connector
------------- |
|
------------- |
| I/O Board |------+ <-- Interface Data Input
| Lowest |
| Addressed |XXX <-- Loop Through Connector
------------- ^---------- Terminator
The Terminator, XXX, is built by connecting the Pin 4 Green and Pin 5 Yellow wires together in an RJ11-6 plug in the last I/O device in a string.
iooperation
Bit Serial Sequence of Operation.
| 1. | Put one bit on the data bit out line. |
| 2. | Bring the data clock line high and then low. |
| 3. | Repeat until all the output data is shifted to the registers. |
| 4. | Bring the strobe line high and then low. The strobe line does double duty. It latches the output data into the output registers and drivers. It also clocks input data into the input shift registers that will be shifted in later. The first input bit is already present on the computer serial data bit input line. |
| 5. | Bring the data clock line high and then low. |
| 6. | Input each bit in turn. |
| 7. | As a check of the output continue to read in a copy of the output bits. |
| 8. | As a check of the input reread the input twice. |
The software was written in Basic. It directly "Bit Bangs" the printer port.
There are 2 type of registers used in the system.
1. CD4021BC is an 8-stage parallel input/serial output shift register. This register is used for inputs.
Fairchild Semiconductor data sheet for the CD4021BC
Texas Instruments data sheet for the CD4021B
2. CD4094BC consists of an 8-bit shift register and a 3-State 8-bit latch. This register is used for outputs. Note, I don't use the 3-State feature.
Fairchild Semiconductor data sheet for the CD4094BC
Texas Instruments data sheet for the CD4094B
Of course the program needs to have copies of the external registers.
The faster the computer the faster the data can be transferred.
This picture shows the RJ11-6 glued down. The next revision has the connectors soldered down in the normal surface mounted way.
Prototype of Bit Serial Quad Satellite Actuator Driver.
I needed to drive satellite dish type actuators for positioning components of my heliostats. I decided to design the interface such that it drives the most common, read cheap, types of these actuators. These generally have either limit switches or overrun clutches to prevent damage to the drive train. Another trait is the use of a single counter switch.
The satellite actuators can draw upwards of 15 Amps when in overrun. This takes a robust transistor. I have chosen 30 Amp P and N channel power MOSFETs in a true high voltage CMOS H bridge configuration. The implementation can drive 4 actuators continuously without overheating, not even warm, not even in hot weather.
The driver consists of an "exclusive or" gate driving the bridge. One input is from the computer software and the other is from the feedback actuator read switch. The job of the exclusive or gate is to apply power to the actuator motor driver until the condition of the feedback switch is the same as the movement bit from the computer. This process allows the computer to move the actuator in small single steps.
The direction the actuator travels is determined by enabling pairs of transistors in the H bridge. One pair for movement outward and the other pair for movement inward.
The computer keeps track of the position and the direction of the last movement. The computer repeatedly sends direction and movement commands to the four drivers. This action causes smooth and even movements to all actuators. The software makes all the actuators move in a manner such that they will all arrive at the required position simultaneously. This in effect makes the rate of movement of the actuators that need to move a small amount less than the actuators that need to move a large amount. My software can drive more than one of these driver boards so any number of actuators may be used. Neat huh!
actuator
I've been compiling some data about the currently available linear actuators. I'm presenting the data here.
I am also working on a simplified driver based on a PIC16C715 micro controller by MicroChip. It will be small and have over current protection.
Power Transistor Terminals.
The arrows in this picture point to examples of power transistors used as output terminals. The terminals are composed of two power transistors mounted vertically and back to back. Since they dissipate so little power they don't need heat sinks. The machine screw is used to clamp conventional spade terminals between the two transistors.
I use hot melt glue to hold the nut in place.
This picture shows the RJ11-6 telephone wire connection. There is also another RJ11-6 connector on the other end of the board. The second connector is used to connect to other boards. If this board is the last in the string a dummy data turn around jumper, or terminator, is used.
This picture shows the RJ11-6 glued down. The next revision has the connectors soldered down in the normal surface mounted way.
sunpath
carter demartile peterthinks tag
Sunpath diagram.
* SunSeeker Project
David Carter's PIC based solar tracker.
Carter uses the a method that acquires the sun by alternately panning in azimuth then elevation.
* "Rich DeMartile" <rich_demartile@prodigy.net>
* has a schematic of a solar tracker and mount based on a pair of CdS photo cells.
* "Peterthinks" <Peterthinks@hotmail.com> has made a solar tracker using RC servos. The system has a tracker based on BEAM technology. The beam circuits powered the RC servos. The tracker used only the power of the sun to move.
* The Analog Guy Solar Trackers:
ST2-48V5A SINGLE & DUAL AXIS SOLAR TRACKER 56V 5A MAX
ST2-12V DUAL AXIS SOLAR TRACKER 18V 0.5A MAX
One of my competiters.
 Electronics |
 Front View |
 Stepper Drive |
 Rear View |
LED7 Solar Tracker Schematic
relay
cds
Cadmium Sulfide Relay Tracker Schematic.
CdS1
This is about the simplest tracker I know of. It uses a Radio Shack 275-249A. Adjust the sensitivity of the CdS cells with a Sharpie permanent marker as described
bellow, in the Chace tracker. The picture tells it all.
This tracker is not as accurate as the electronic tracker but quite sufficient for use with PV panels.
Top. Normal operation between limit switches.
Middle. The left limit switch has opened to stop movement to the left. To move to the right again the diode conducts current that allows movement to the right.
Bottom. The right limit switch has opened to stop movement to the right. To move to the left again the diode conducts current that allows movement to the left.
Sellect a diode or rectifier rated at the maximum motor current plus some margine. Also the voltage should be at leat 100V and preferably 200V.
Needles to say, the limit switch must operate before the mechanical limits are reached. If the mechanical stop is reached before the switch the motor can draw quite high currents and can destroy the solar tracker.
LED1Bingo! This got me to thinking.
They generate quite a bit of voltage. The green ones generate about 1.65V, some as much a 1.74V. Not the piddley .55 volts of a silicon PV cell. How is this so? Well, it turns out green LEDs are made from Gallium Phosphide, a semiconductor with a much higher bandgap voltage.
I thought I had invented the use of LEDs as PV cells as I had never heard of this effect before. Well, after some investigating I found a number of references to this. The guys that had done the most work in this area were the people form the "BEAM" project. They make tiny solar powered robots and some used LED photo sensors.
I had been using a very low threshold MOSFET in a TO-92 package, BS107P. The threshold is about 1.5V. If I put two LEDs back to back, one fighting the other, the one with more light intensity wins. I thought I could use this to switch the MOSFET. And it worked.
By using one LED as a sort of power supply and the back to back pair connected from it to the MOSFET gate the circuit is complete. (This I have not seen elsewhere.) My implementation uses three power supply LEDs, aimed East, Up, and West. The sensor LEDs are aimed about 90 degrees from each other and at about 45 degrees either side of up. Of course the easterly pair will be a little to the east and the westerly pair a little to the west. This makes the center have a dead zone where tracking stops.
The circuit is quite sensitive. It brings the panel back to the east just after sun rise. The accuracy is quite good. You can calibrate the sensor by bending or aiming the LEDs a bit.
LED2Note! R4 and R5 are used to force parking when it gets dark. If parking is not desired don't use R4 and R5. Parking may not be desired in low power consumption applications.
Also, the parking resistors, R4 and R5, reduce sensitivity a bit.
LED5S5VI've tested the circuit on voltages from 3V to 21V. With some component changes it should be useful to 63V in a 36V PV panel system although I haven't tried this yet. With higher voltage and the use of heat sinks on the bridge transistors much higher currents should be possible.
The parts cost is very low. Parts cost estimated using Digikey prices. Ok, you can get stuff from the surplus stores but I will stick with Digikey.
1. 2N2222 NPN transistor 4 @ $0.21 = $0.84
2. 2N2907 PNP transistor 4 @ $0.21 = $0.84
3. 91 Ohm 1/2 W resistor 2 @ $0.06 = $0.12
4. 5 KOhm 1/4 W resistor 2 @ $0.06 = $0.12
5. 22 nF capacitor 1 @ $0.08 = $0.08
6. LED Green Lumex SSL-LX5093LGT 2 @ $0.12 = $0.24
Total = $2.24
$2.24, is this cheap enough?
led5s12v
LED5S12V Simplified LED low power tracker.
LED5S12V
1. 2N2222 NPN transistor 4 @ $0.21 = $0.84
2. 2N2907 PNP transistor 4 @ $0.21 = $0.84
3. 750 Ohm 1/2 W resistor 2 @ $0.06 = $0.12
4. 47 KOhm 1/4 W resistor 2 @ $0.06 = $0.12
5. 100 KOhm 1/4 W resistor 2 @ $0.06 = $0.12
6. 22 nF capacitor 1 @ $0.08 = $0.08
7. LED Green Lumex SSL-LX5093LGT 2 @ $0.12 = $0.24
Total = $2.36
led5connections
LED5 Connections
led5s12vplc
LED5S12V Modified for use with a PLC.
LED5S12VPLC
This version can be used as a sensor input to a PLC, Programmable Logic Controller. This circuit does not have a parking function. Parking and Reverse Inhibit functions are best performed in the software of the PLC.
2l003
Grainger 2L003 Gear Motor
This is a 12VDC gear motor from Grainger's.
This controller works well with the 2L003.
This motor is an off the shelf motor from Grainger's. Stock number 2L003. It's rated for .45 RPM at 50 In Lbs. The motor current is less than 100mA at 12V and about 50mA at 5V.
tamiya
Tamiya has a number of Model Gear boxes.
The motors supplied are rated for about 3V. They draw a bit to much current for the 2N2222-2N2907 driver transistors.
lamble
LED5
Stewart Lamble built this version.
He subistuted BD135 (NPN) and BD136 (PNP) transistors.
LEDFAST1. R7/R8 form a voltage divider to produce a voltage of 1/2 of VCC and applied to the non inverting inputs of the OpAmps.
2. The OpAmps are setup to have a gain of 1000X through R3/R1 and R4/R2. Capacitor C1/C2 limit the high frequency response of the circuit to prevent oscillations.
3. The LED sensor circuits need to be high impedance so are isolated from the gain resistors through resistors R5 and R6.
4. I use large 10mm "Green" LEDs with clear cases. They are made by Lumex, but all normal LEDs can work. (Don't use the White LEDs as they are not normal types.) The LEDs act as small photo voltaic generators. Since LED1 & LED2 sensors are connected back to back the sensor that has the greater light intensity expresses its voltage over that of the sensor with lessor light intensity. Imbalance in the light on the sensors produces a differential voltage which is amplified and presented to the motor. As the light approaches balance the motor differential voltage approaches zero resulting in no motor current.
5. The LM324 has an output current drive capability of >10mA. Transistor pairs Q1/Q2 & Q3/Q4 form unity voltage gain emitter follower current gain amplifiers. With power transistors that have a gain of 100 the motor drive current can be about 1A or more.
6. VCC can be from about 6V to an absolute maximum of 32V. Other OpAmps can be used for a greater VCC range.
7. Note! Limit switches are required in the motor circuits. See:
How Limit Switches Operate.
8. There has to be a down side though. This is a true analog circuit that drives the output transistors in a linear manor therefore power is wasted when slowing the motors. Heat sinks may be required.
9. This circuit is not generally suitable for use with normal high efficiency solar tracking applications. It is best suitable for school projects.
LED3Note! I have replaced the LED3 with the much more capable LED3X series of solar trackers. See below.
This circuit uses power MOSFET drivers and is designed to operate satellite dish linear actuators, however most any DC motor can be used. The power drivers are capable of delivering about 10 amps of peak current, maybe more. When better transistors become available this current can be increased. The drivers operate the actuators in pulses of about .3 second every 3 seconds or a 10% duty cycle. This eliminates the needed for a heat sink on the transistors. Neat huh!
I haven't decided if 10% is the best duty cycle to use. Less will make the tracking slower but, we don't need speed anyway. I will determine this when I get better weather. Slow tracking speed helps in partly cloudy condition. This prevents the tracker from making unnecessary movements when clouds move by.
No electrical adjustments are required. The LEDs can be mechanically adjusted for optimum tracking performance by aiming them after the circuit board is mounted.
led3shadow I have used a band of metal about .5" in width at about 6" from the LED3. If the LED3 is used for E-W tracking the band is oriented N-S. Conversely, if the LED3 is used for N-S tracking the band is oriented E-W. The shadow device is not particularly critical. For instance, I have used black electrical tape on the weather dome and it worked well.
To improve accuracy, ie. with concentrators such as troughs or dishes, a blocking shadow can be placed in front. The shadow just covers the two inner LEDs when aimed at the sun. Similar to the shadow on the Chace Tracker.
| Power Supply Voltage | 8 Volts to 22 Volts inclusive.
The 8V minimum is specified to prevent damage to the MOSFET power drivers. The damage is due to operating them in the linear region with a load. This causes excessive power to be dissipated in the MOSFET with a resultant damaging temperature rise. The 22 volt maximum is defined by the voltage tolerance of 24V protection zener. This zener protects the power MOSFETs from seeing damaging breakdown voltages. During testing I had several failures when operating from a car battery while the alternator was running. It was determined that the alternator was producing voltage spikes in excess of the 30V breakdown specification of the MOSFETs. |
| Load Current Continuous | 5 Amp resistive.
The power MOSFETs are rated at over 10A at 25°F. A conservative derating of 50% is prudent especially in hot weather conditions. |
| Load Current Intermitant | 10 Amp intermitant at 10mS width once per timing cycle.
The Power MOSFETs have an absolute maximum current rating of 30A, but this is with ideal conditions where the temperature is 25°F and very fast gate rise times. The LED3 has a relatively slow gate rise time and may be operated at quite high temperatures sue to the weather. I think 10A at about 10mS is adequate for normal tracking applications. I should note that the satellite dish actuator I use normally consumes about 290mA of current at 13.8VDC. This actuator is capable of driving a 15' dish with 1500lb of force. You don't need a high powered drive, just a slow forceful one. |
| Operating Temperature | -40°F to 185°F or -40°C to 85°C |
The circuit board and sensor assembly of the Chace tracker.
The circuit is not water proof so a protective plastic dome is needed. I have used 2 liter clear plastic soda bottles. They last a long time, at least 5 years for one I have used, probably a lot more. This circuit will fit into the 20 Oz. size. The plastic bottle chosen needs to have a round bottom, the type that comes with the black plastic bottom which is removed. The type with the molded in feet don't work very well as they diffract the light too much.
Glass jars are even worse optically if looking through the bottom.
Any one out there with a good idea for a better weather dome?
In Minnesota I have snow that builds up on the top of the dome. This snow is quite reflective and can confuse the sensors as to the correct direction for the brightest portion of the sky when in the sky is overcast. It's easy to just brush the snow of but this is not always timely. I have experimented with putting black opaque paper inside the dome to eliminate the light from the snow. Another paper light blocker can be put under the sensors to stop the reflected light from the snow on the ground.
dome
Dome on Chace tacker
This weather dome is made from a 2 liter pop bottle.
jiff
I have tried and like using plastic Jiff Peanut jars
 Jiff peanut butter jar. |
 Jiff peanut butter jar with only the lid. |
 East View |
 West View |
 Assembled |
 Top View Without Dome |
 View From Below |
 Long View  The Base |
 LED3 in Fruit Jar |
 PV Panel |
 PV Panel |
 Panel with LED3 |
 Back of Mount |
 Power Controller  Battery Pack |
 LED3 in Fruit Jar |
 Overall View |
 Closer |
 Mount and Connections |
 Wearher Dome in Upper Corner |
 Closeup of Weather Dome |
 LED3 |
gary
Gary <gposavad@home.com> has suggested making the dome from Lexan. He wrote:
It's fairly easy to make a perfect dome out of Lexan.
Make a 2 piece mould out of whatever you can find that's the right size.
Heat up a piece of Lexan in your oven at about 200F for about 5 min.
Put it on the mould and press and release.
Trim off the excess after it cools.
If you want really good optics leave room in the mould for a piece of flannel on each side.
Scratches are polished out with toothpaste.
I Got this from EAA Sport Aviation to make nice wing tip strobe light covers.
mrea2001
Demo of my tracker at the 2001 Midwest Renewable Energy Fair
LED3 Schematic
Layout
Molex Connector| LED Pair 1-3 | LED Pair 2-4 | H-bridge Left | H-bridge Right | Movement |
| West | West | Low | High | Move West |
| West | East | Low | Low | Stopped |
| East | West | High | High | Stopped |
| East | East | High | Low | Move East |
Note! Unlike most H-Bridge drivers where both sides are always driving the load this circuit has independent drives for each side. This allows the motor to stop when both sides are high or low. The motor moves when they are different.
In addition, both sides have the top MOSFETs turned off most of the time until the clock enables movement by periodically pulling R9 and R11 low.
Q3 and Q4 form a gated level shifter. To see how this driver works think of the bottom end of R9 and R11 being grounded when the top transistor is to be enabled.
Also the driver for the left side are inverted from the driver on the right side. This is a method for returning the tracker to the east, for parking, if R2 and R3 are used. This way Q1A will be enabled and Q2B will be on when in the dark.
parking
This circuit uses power MOSFET drivers and is designed to operate satellite dish linear actuators, however most any DC motor can be used. The power drivers are capable of delivering about 50 amps of peak current, maybe more. When better transistors become available this current can be increased. The large power MOSFETS, 72A, when operated at low duty cycle or low currents eliminates the needed for a heat sink on the transistors. Neat huh!
The duty cycle is adjustable from 0% to nearly 100%. Idealy tracking from stop to stop should be 10 to 30 minutes. The duty cycle helps to slow down the motor drive speed. Less duty cycle will make the tracking slower but, we don't need speed anyway. Slow tracking speed helps in partly cloudy condition. This prevents the tracker from making unnecessary movements when clouds move by.
led3xshadow
I have used a band of metal about .5" in width at about 2" from the LED3X sensor. If the LED3X is used for E-W tracking the band is oriented N-S. Conversely, if the LED3X is used for N-S tracking the band is oriented E-W. The shadow device is not particularly critical. For instance, I have used black electrical tape on the weather dome and it worked well.
led3xremotesensor The remote sensor comes can be configured in several flavors. The PC board is configured in 2 halves. Each half is an individual single axis sensor. For single axis use the board is cut in half or dual axis if left whole. Depending on which components, positions, and jumpers installed all the configurations can be obtained.
led3xremotesensorschematic
In cloudy or overcast weather the tracker seeks the brightest part of the sky. At night it moves to the easterly parking position. The parking position puts the panel at a steep angle so night time snow doesn't accumulate as much in the winter. A westerly parking position is obtained by mounting the tracker upside down and reversing the leads to the actuator motor.
led3x
LED3X LED Sensor Electronic Tracker with H-Bridge Drive.
LED3X Solar Tracker
The LED3 using surface technics was just to hard to assemble in a timly manor as sales were increasing. I needed a circuit that could be built with mostly through hole components. At the same time, there were other features and enhansments that I thought were needed. This was the impetus for the LED3X series of solar trackers.
To improve accuracy, ie. with concentrators such as troughs or dishes, a shadow blocker can be placed in front. Similar to the shadow blocker on the Chace Tracker or like this:
OK, this is an example of a dual axis version.
The single axis version uses a strip of metal.
Cool weather dome!! As I recall the dome is polycarbonate and from the oil or water traps on air lines and regulators.
LED3X Remote Sensor
1. Single axis
2. Dual axis
And parking or no-parking on either axes.
Dual Remote Sensor Schematic
| Power Supply Voltage | 10.5 Volts to 44 Volts inclusive.
The 10.5V minimum is specified as the under voltage point. Less voltage protects the power mosfet in the H-bridge driver circuitry. The 44 volt maximum is defined by the voltage tolerance of the 51V protection zener. This zener protects the power MOSFETs from seeing damaging breakdown voltages. |
| Load Current Continuous | 9 Amp resistive.
The power MOSFETs are rated at over 70A at 25°F. A conservative derating of 50% is prudent especially in hot weather conditions. |
| Load Current Intermitant | 20 Amp intermitant at 1S width once per timing cycle of 60S.
The Power MOSFETs have an absolute maximum current rating of 72A, but this is with ideal conditions where the temperature is 25°F and very fast gate rise times. The LED3x has a relatively slow gate rise time and may be operated at quite high temperatures due to the weather. I think 20A at about 1S is adequate for normal tracking applications. I should note that the satellite dish actuator I use normally consumes about 290mA of current at 13.8VDC. This actuator is capable of driving a 15' dish with 1500lb of force. You don't need a high powered drive, just a slow forceful one. |
| Operating Temperature | -40°F to 185°F or -40°C to 85°C |
led3xforsale
LED3XS24Vc3 For Sale
Please go to this page:
http://www.redrok.com/led3xassm.htm#led3xforsale
power relaydc1 relayac1 relayac2 manual Since I do the assembling variations can be easily made for such things as a change in the timing of the oscillator for other duty cycles or to disable the parking feature. I am currently selling the assembled units without the parking feature.
assemblyled3x led3xenergy The daily energy consumed is quite small. My actuator draws about 290mA and can go from stop to stop and back in about 5 minutes. So: A very small 5W or even a 2W PV panel and small gel cell lead acid battery is a good combination for the power source to run the system.
I intend that this circuit will supplant the Chace tracker as its simpler to build and adjust.
For two axis tracking two circuits are needed. However, for PV panels the second axis only adds about 5% on the average and may not warrant the added expense.
economics Using the NREL data: Or a single axis north-south tracking panel at lat. +15 deg. Lets do it in June: Or a single axis north-south tracking panel at lat. -15 deg. Of course your location will have different results. For instance I just did it for San Jose, CA and got an increase of only 14% in both June and December. Clearly this is not the best place to do solar tracking.
Here are the 4 maps that I used. The full set of maps from NREL are here.
I don't have a cost for the tracking mount but the actuator I'm using costs about $140us and my tracker is $35us. PV panels now are about $5us/W so a 100 watt panel costs $500us. Using a system of 3 panels this is $1500us.
Since the tracker delivers at least 40% more output, in December, it would be fair to say that the tracker has an equivalent value of $600us in a 3 panel 300W system. I can't see the tracking mount would cost $600-$140-$35=$425 more than the equivalent stationary mount that has to hold at least 4 PV panels. Of course, the comparison is even better in June with the 100% increase.
Today there are many obsolete C-band satellite dishes. These have polar axis mounts which are almost ideal for use with arrays of PV panels. Remove the dish and install racks to support the panels. They are well designed and very strong. They have all the needed hardware including the motor actuator. Besides they are often free for the asking.
I am convinced that the best and strongest mount for PVs at ground level is one based on the design example of:
poulek aps powerlight szymanski tripod It's made from PV plastic sewer pipe. There are some compromises to this example. Improvements Note! I no longer sell the LED4 as it was replaced by the much more capable LED3X. Ok, the LED3X doesn't have a charge controller but this isn't really required to run the tracker.
The design philosophy for this tracker was to make it more robust and self contained. I included a charge controller for rechargeable batteries intending them to be NiCds to run the actuators. Since the charger is there it might as well be capable of running a large PV panel with lead acid batteries.
Note! this charger is not intended to charge the expensive system batteries. Leave that job to the high end smart charger such as those by Trace, Outback, and others.
It can also run a very small PV panel to charge the local NiCds or NiMH batteries. These batteries work well over a wide temperature range.
I will test the ability of the sensors to be at the end of a long cable. I will test this using 1000' of 4 wire telephone cable. 8 wire Cat5 cable should also work with the extra wires driving the actuators.
With the use of long Cat5 cable the main board can be inside a heated battery room. Of course you still have to run the heavy gauge main charging wire pair to the battery room. The advantage of having the batteries in a heated room is temperature compensation won't be needed.
Some have expressed an interest in driving high powered loads beyond the capabilities of the H-Bridge driver transistors. To this end I developed several high powered driver circuits, (actually their almost the same circuit as in the relay trackers).
RelayDC1
Relay circuit that uses DC relays with 12VDC coils. The DC motor in this case is a permanent magnet type that is reversible.
RelayAC1
Relay circuit that uses AC relays with 12VDC coils. The AC motor in this case is a capacitor run type.
RelayAC2
Relay circuit that uses AC Solid State relays with 3 to 32 VDC control inputs. The AC motor in this case is a capacitor run type.
led3manual
led3xmanual
LED3XManual
Some have expressed an interest in adding a switch to manually move the array for test purposes. One of these circuits, when added between the tracker and actuator, will allow manual movement. The first circuit has a fast and slow position. The second circuit is slow only but simpler. The third circuit is the nicest.
Assembly & Opperating Instructions for the LED3X series. Plus a lot of pictures and examples.
The circuit draws about 15mA when idling. It can operate at temperatures to -40F or up to 158F. Its designed to operate from a 12V or 24V lead acid power source. I generally recommend using a 36V actuator on 12V.
((.015A * 24hr/day) + (.290A * 5min/day / 60min/hr)) * 13.8V = 5.3Whr/day
Or about 5Whr/day, which is pretty small.
Many have said that it makes no sense to use a solar tracker with PV systems as it is cheaper to just add an extra panel for every three. To this I say bunk.
My link.
http://rredc.nrel.gov/solar/pubs/redbook/redbook_index.html
I find that in Minnesota a single axis tracking PV panel will have a 40% increase in output in December and a 100% increase in June.
In Minnesota:
A PV panel with 15% efficiency
in December tilted to your latitude plus 15 degrees.
http://rredc.nrel.gov/solar/old_data/nsrdb/redbook/atlas/
2 to 3 kWh/m^2/day lets say 2.5 solar
2.5 kWh/m^2/day * 15% = .375kWh/m^2/day electric
3 to 4 kWh/m^2/day lets say 3.5 solar
3.5 kWh/m^2/day * 15% = .525kWh/m^2/day electric
.525 / .375 = 140%
This shows 40% improvement with a tracker in December.
4 to 5 kWh/m^2/day lets say 4.5 solar
4.5 kWh/m^2/day * 15% = .675kWh/m^2/day electric
8 to 10 kWh/m^2/day lets say 9 solar
9 kWh/m^2/day * 15% = 1.35kWh/m^2/day electric
That's a 100% improvement with a tracker in June.
nrelmaps
Click the map!
traxle
* Traxle
Poulek Solar
by Martin Poulek
Gary A. Werner, President GWM Corporation
Polar mount with tracker.
The tripod support arrangement is very strong with light weight construction.
* APS Solar's "Tilt Tracker".
Similar to Poulek Solar's Traxle, with a clockwork tracker. The second axis, the tilt axis, is adjustable for a few percent improvement in total output. The actuators are hydraulic.
* PowerLight
A bigger version, similar to Poulek Solar's Traxle
* Martin Szymanski's variation of a Tripod Tracking Mount
TriPod Mount
This is my version of Poulek's mount. I call it the TriPod Mount. This mount is cheap and crude but easy to make, especially for demonstrations. It's easy to build and knock down. Its quite portable.
A proposed Stationary Receiver Dish on a Tripod Mount.
A proposed Stationary Fresnel Lens on a Tripod Mount.
1. The legs are fitted with end caps.
2. The foot caps are drilled and have long bolts protruding through to be pushed into the ground.
3. The support legs have "I" bolts fitted to the top caps.
4. The main polar axis leg has a large bolt that is fitted through the "I" bolts of the support legs.
(The Primestar mount doesn't allow the pipe to go all the way through so I made a compromise and have this large top bolt mounted to the dish support. Although the Primestar mount has a method to manually adjust for declination.)
5. The polar leg needs to rotate. To do this side pipes are glued on to the polar and one of the support legs.
6. The satellite dish linear actuator is fitted to the side pipes.
7. To set it up just adjust the support legs until the polar tube is aimed at the pole star. This is due north and angled at exactly your latitude.
8. To make sure the thing doesn't blow away I have a heavy nylon cord, ( not in the pictures ), tied from the top of the polar leg down to the earth. I used a screw into the ground dog leash stake.
9. Don't bolt the dish and PV panel to tightly. This will allow some manual adjusting.
1. The actuator can only rotate the polar leg about 150°. A better way would be to have a pulley and cable to get the rotation to 270° or so.
2. The Primestar mount is not ideal. A better mount would allow the polar leg to protrude fully.
3. A longer polar leg would allow more rotation without interfering with the supporting legs.
1. For larger equipment one should make the tripod from steel cold water pipe. This would make things much sturdier.
2. Make the polar rotating device using a large pulley and cable to allow more rotation.
led4
LED4 LED Sensor Electronic Tracker with Simple Solar Charge Controller.
LED4
The "LED3" light sensor works very well. It is sensitive, accurate, and cheap. A couple of years ago I had devised a simple solar charge controller I called "Shunt 2". I thought it would be a good idea to combine them into a single PC board called the LED4.
 LED4 Schematic |
 LED4 Layout |
led4connections LED4 Pinout |
ledsensor01 Alternate LED4 Sensor Schematic |
| Separate LED Light Sensors | The signal sent back from the sensor can be sent a considerable distance since the current flow is minimal. However, static electric discharge could be a problem if the sensor is more than about 6 feet or so from the main board. If long distances are needed please Email me for advice. |
| Under voltage Shutdown Protection | Shuts down the actuator motor drive when the battery voltage drops below about 10½ volts. |
| Built In Charge Controller | Designed for use with nominal 13.8 volt battery systems. The charger is designed for controlling 13.8 volt 125 watt PV panels. There is no provision for either temperature compensation or multiple charging strategies. |
| Larger Power Transistors | The power MOSFET transistors are much more robust than in the LED3. It uses IRF5305 and IRLZ44N transistors from International Rectifier. The IRF5305 P-channel MOSFET has an on resistance of about 60mOhm and the IRLZ44N N-channel MOSFET has an on resistance of about 20mOhm. |
| No need for a heat sink | A 125 watt PV panel can deliver about 9 amps of current. The maximum wattage dissipated by the IRLZ44N shunt transistor under these conditions is about 1.6 watts continuously. The Schottky Barrier series protection diode is rated for about 9 amps. The H-bridge transistors are pulsed at about a 10% duty cycle. I'm rating the actuator maximum pulsed current at about 16 amps. If you want to continuously drive the actuator keep it below 5 amps. The IRF5305 has a maximum current rating of 110 amps and the IRLZ44N has a maximum current rating of 160 amps so if the currents are kept to below my system rating no heat sinks will be needed. |
| Built with Through Hole Components | Through hole construction allows the average home brew guy the ability to easily put the LED4 tracker together either as a kit or scratch built. I can supply the PC boards separatly for $9.00us. |
| Operating Temperature Range | -40°F to 185°F or -40°C to 85°C |
concentratortemperature A number of people have asked the question: Let's use this example: Lets review: 12' foot C-band satellite dishes have a focal length of about 10'. At least mine does.
The sun has an image diameter of about 1/2 degree of arc. At 10' this perfectly focused diameter would be: So firstly, you can't get to a 1" diameter.
"How can you estimate the temperature of the various smaller diameters?"
The simple answer is you can't. Temperature of a receiver is only loosely related to X. The more important factors are: Another way to help in the understanding of receiver temperature is it is a kind of balance. The receiver, insulated on the back side, radiates into a hemisphere. The dish appears to cover a portion of the hemisphere. And this is true. However larger receivers have other losses such as: What are you trying to do? sphericalcollectormath Another way to use this information on our system. martins
Calculations of Concentrator Temperature Rise.
"How hot will my concentrator get?"
or:
"How much concentration will I need to get to a certain temperature?"
"A twelve foot dish focused to a 6 in., 3 in., or one 1 in., diameter receiver.
How can you estimate the temperature of the various smaller diameters?"
X factor is the ratio of the captured sunlight area divided by the radiating surface area of the receiver. Assuming the back side is well insulated.
( 2 * tan( .5deg / 2 ) * 10' = .087' * 12"/1' = 1.04"
That's for a perfect parabola. C-band dishes aren't nearly that accurate. So maybe the diameter is about 3".
3" is about (144"/3")^2 = 2304X
6" is about (144"/6")^2 = 576X
1. Rate of heat removal.
For instance if the heat removal is infinite then
the temperature would be the same as ambient.
2. The selectivity of the surface of the receiver.
Good selective surfaces such as black nickel oxide
are pretty good and will attain a higher temperature
due to not radiating as much heat back to space.
3. A limiting factor is the surface temperature of the sun.
Generally considered to be about 10000F or about 5600C.
4. The quality of the reflectivity of the mirror.
I usually assume this at 80% to 85% regardless of what
the manufacturer of the Mylar films say.
The ratio:
(dish angular surface area) / (a hemisphere) * 10000F = Temp
This implies that a receiver temperature is independent of X.
1. Convection loss
2. Process heat removal
Generate steam?
Steam receivers don't have to be particularly small as the temperatures are limited by the steam system.
See our experimental receiver above.
This dish has a no thermal output temperature rise of about 1100F with a concentration of about 20X. We estimate this is about correct for the dish angular area.
* Here is an Excel spreadsheet to do the math.
http://www.redrok.com/sphericalcollectormath.xls
1100F = 0% efficiency. ( No output of steam.)
50F = 100% efficiency ( Perfect cooling of the receiver.)
Other efficiencies are approximately linear between these.
837F = 25%
575F = 50%
312F = 75%
Actually the lower temperatures will have higher efficiencies because convection losses are reduced.
Does this make sense?
<Martin Szymanski's> variation of a Tripod Tracking Mount
assemblyled4
Assembly Instructions for the LED4.
opperatingled4
Operating Instructions for the LED4.
Chace1Some guys on the net have expressed a desire for a simple analog solar tracker to operate their concentrator projects. Jeremiah Chace sent me a schematic for such a tracker. His circuit was based on a Cadmium Sulfide, CdS, photo cell with a relay output. There were a couple of problems with the circuit. After some discussions with him I made some improvements and here are the results.
The basic operation is essentially that of 2 separate photo sensor circuits in pairs. Each separate sensor circuit has a control pot to set it's sensitivity for bright sunlight.
It is important to have a thorough understanding of the operation of the circuit in order to be successful in making the correct adjustments for acceptable operation. One must do the adjustments on a clear cloudless day. At the very least wait until the sun is between clouds.
CdS cells have a wide variance in the resistance they exhibit in bright sunlight. Generally the resistance is too low for this type of circuit. I have found that the CdS cells can be modified to have higher resistance by painting the cells with a black marker. I use a Sharpie black permanent marker. I paint this on all sides and edges. If you get too much on it can be removed with fingernail polish remover or acetone.
The correct amount of blackening is when the adjustment pot is approximately at mid setting when the circuit is tilted so about half of the cell is illuminated by sunlight and half covered by the light blocker. Repeat this painting of the cells until all 4 CdS cells are adjusted.
I have found a helpful tool to make the adjustment easier. Just use a small mirror to reflect sunlight onto the CdS cell. By covering one cell with the finger and reflecting light onto the other cell the tracker can be moved at will. This allows the cell sensitivity adjustments to be easily made.
The painting procedure would not be very easy to get accurate were it not for the pot. The settings seem stable but I haven't had very much time on the circuit. I suspect that the most problem would be the fading of the black marker. If this happens I suppose a different more permanent light filter coating could be devised. Possibly Parsons Black which is carbon lampblack in spar varnish.
P.S. I have had about 4 months of operation on the tracker and have not seen any problems.
There have been days where no bright solar radiation falls on the sensors the tracker just stays in the last position that movement occurred. The panel is not at the optimum orientation to gather the maximum amount of energy from this diffuse radiation during these times.
In the long run this is not a problem. Of course the amount of energy is less than optimum during cloudy times. The gain due to tracking greatly exceeds the dim time loss.
chaceshadow
Sensor and shadow light blocker.
Each CdS cell of a pair is on each side of a central light divider with a top light blocker. The light blocker is just wide enough to block both CdS cells when the sun is dead on center. In this example the light divider is made of rubber and the light blocker is an old aluminized 5.25" floppy disk write protector. The materials used are not critical. They need only be light opaque. The light divider should be black and non reflective. The blocker could be made from reflective aluminum foil to prevent heat buildup.
Note! The CdS cells are highly angled in such a way that they are generally aimed at a portion of the sky that is away from the sun. This helps to get the platform moving in the right direction when far off track.
If either CdS cell is uncovered by the light blocker it will conduct heavily and the associated power driver will move the actuator in such a way that the CdS cell is again covered.
When neither CdS cell is exposed to bright light the actuator is prevented from moving. This leaves the tracker at the last moved to position. Little power is expended searching for a new position when in dim light.
The light blocker is supposed to never let both CdS cell be exposed to bright light. However it could happen. If this condition occurs the actuator is prevented from moving or hunting. This saves power by moving only when necessary.
This schematic is for one axis. The second axis is the same as the first and uses the other half of the LM339 comparator.
When the sun is behind clouds the actuator is prevented from moving or hunting. When the sun comes back out the tracker will move toward it again.
When the sun sets the platform remains aimed to the west. In the morning the easterly aimed CdS cell becomes active and the platform positions itself back to the east.
In my example I have the light dividers about 2 inches long. If a tighter angular tolerance is desired the light dividers can be made longer.
Overall Layout.
The layout of the circuit has the CdS cells mounted on the eastern side of the board. The east/west, right ascension, pair is on the left. The north/south, declination, pair is on the right. The output power drivers are in the foreground.
I have mounted the power MOSFETs in such a way that they form the output terminals. Another example is pictured in my heliostat I/O board shown here: Terminals.
The PC board is all single sided using mostly 1/4 Watt resistors and capacitors mounted in surface mount fashion. This prototype construction technique is fast and easy to build. If the board is to be built in larger quantities a through hole board could be build with the parts in exactly the same location as the surface mount version.
(Any one want one? Email me at: <redrok@redrok.com>)
Here is my 1998 US cost breakdown in single quantities.
IRF9Z34N $1.82 * 4 = $7.28 Digi-Key
IRFZ34N $1.25 * 4 = $5.00 Digi-Key
LM339N $ .49 * 1 = $0.49 Digi-Key
resistors $ .02 * 20 = $ .47 Digi-Key
Pot D4AA15 $ .22 * 4 = $ .88 Digi-Key
CdS Cell $ .35 * 4 = $1.40 Jameco
Total $15.42
A true surface mount board using conventional, tiny, surface mount parts would make the board about 1/2 the present size. All the components are available in surface mount form except the CdS cells. Even the power transistors can be had in surface mount.
I don't think that the smaller sized board layout justifies the higher cost of the surface mount components.
Weather Dome.
How do you like the high tech weather dome? It is made of a 2 liter polyethylene plastic pop bottle. This plastic is pretty good in the UV of sunlight. Of course it can be easily replaced if it becomes clouded. Use the bottles with the black plastic base and remove it. The fancier Coka Cola bottles with the molded in feet don't work very well optically.
I think a 1 liter bottle would work just as well and the wooden mounting would be smaller. I haven't tried a 1 liter bottle yet.
The dome is not sealed from the weather in the picture. I will drill a hole in the wooden mount for the wires. The tapered sides to the wooden block will seal the dome to the block. The interior gets hot enough to vaporize any condensed moisture.
Experimental Goofy Demonstration Mount.
I needed a platform to test the circuit. I usually don't want to publish an untested circuit. I have had experience with circuits that don't operate as expected the first time. I sent Jeremiah a prototype, (Rev. A2), of the circuit and PC board to try out on his dish. Unfortunately he is moving his shop and couldn't try it out right away. I then decided to build a simple mount that I had been thinking about for a while.
This is a simple design meant for use with PV panels. It uses 2 satellite dish linear actuators. One to move the platform east and west about +-75°s in right ascension. The other to move the platform north and south a minimum of +/-23.5°s in declination.
The main pier is slanted to the south at an angle of 90° - the local latitude, (45°s for me). The reason the pier is tilted south is to align the right ascension axis to the polar axis which is the bolt that holds the ascension member to the pier. A second brace will be installed under the pier for support if needed.
Right Ascension / Declination Head.
This is a close-up of the mount head. The declination mechanism is controlled by the second satellite dish actuator. It tends to move the declination platform around the axis on the ends of the right ascension mechanism.
This picture shows the arrangement of the two satellite dish linear actuators. These actuators have adjustable limit switches. These switches are required by this circuit both to limit the travel and stop power from being wasted when at the limits.
Actuators with limit clutches are unsuitable for this application. While they do stop when the limit is reached, their motor continues to consume power. This is disastrous when running on a storage battery. The actuators, which consume over 10 amps, will drain the battery in a short while.
The magnitude of the declination movement is maximum when at the extremes of the right ascension and minimum in the middle of the travel range.
This design leaves a lot to be desired.
| 1. | It has limited travel. |
| 2. | The declination angle changes in a complicated way with the movement of the right ascension angle. |
| 3. | It probably is not very strong when at the extremes of travel. I used it only as a convenient means of testing the analog controller. |
Brad's PV Solar Tracking Panel and Controller.
bradtrough
This is one of Brad's patents for this mechanism.
Brad's Trough Collector.
This is Brad's patent for the solar tracking controller.
shunt
charger
Solar Charge and Diversion Controllers.
Most solar charge controllers are of the shunt type. They are easy to build and work very well. More complex controllers are the Maximum Power Point Controllers.
I will describe several circuits that are based on a Zetex ZM33064 computer voltage monitor and reset circuit. While there are a number of companies and models of voltage monitors on the market I chose the Zetex ZM33064 because of its low cost, accuracy, low power consumption, but mostly for the low, 20mV, hysteresis characteristic. Most reset circuits operate in a similar manner to the Zetex ZM33064 and could be substituted in the circuits.
I believe that power controllers based on voltage monitor circuits are significantly reduced in complexity due to the high integration of several power controller characteristics in a small 3 pin device that looks like an ordinary TO-92 transistor. In come cases it and a power MOSFET transistor are the only active devices in the circuit.
shunt1 This is the simplest charge controller that I know of. OK, a really big zener is simpler but I don't know where to get one that can do the job. I have built several of these for different purposes and they work OK but I don't recommend there usage anymore because more advanced circuits exist today.
Shunt1 Of course you will also need a 25 watt heat sink for the transistor.
I don't know why but Digi-Key dropped their bipolar power transistors. You can get any NPN power transistor capable of dissipating 25 Watts possibly from Radio Shack.
This is a version of an amplified zener diode with a Schottky diode used to prevent power loss through the panel at night.
Here's A better Solar Shunt Regulator. It can be made as large as you want. It's not as "simple" but it works better.
Shunt2 Is this cheap enough?
How the Conventional Shunt Charge Controller circuit works.
This shunt regulator is based on a Zetex ZM33064 computer under voltage reset circuit.
The voltage monitor is an integrated circuit in a 3 pin package that in its normal usage is used to reset a microprocessor. The IN pin would normally be connected to a 5V logic bus.
When the 5V bus voltage drops bellow 4.6V the open collector OUT pin is pulled down toward ground.
In my circuit I use this output to switch the gate of a logic level power MOSFET.
When the voltage is bellow 4.6V the transistor is turned OFF and the PV panel is allowed to supply power to the battery until the voltage goes above 4.6V. The Zetex ZM33064 has a hysteresis of about 20mV which is degraded to 100mV in the circuit. This helps to prevent the power MOSFET from going into linear mode and dissipating power in the transistor. When in regulation the transistor will switch On and OFF at a rate dependent upon the capacity of the battery and power available from the PV panel.
The zener and resistors are used to divide the battery voltage down to the 4.6V range for the voltage monitor.
Don't skimp on the power MOSFET. You might ask why use a transistor capable of passing 41 amps in a circuit designed for only 5 amps. The answer is in the cost and complexity of the heat sink. This transistor has only .022 Ohms of ON resistance. This power dissipation at 5 Amps is: Of course if the full 41 amps is passing through the transistor then: You might ask if there is a problem with the direct shorting out of the PV panels? The simple answer is NO! Now for the more complex answer. Think of it this way. when the panel is delivering it's peak load at a nominal efficiency of 15% much of the other 85% is dissipated as heat in the panel. If the panel is either open circuited or shorted then no power is delivered to the load and 100% of the influx is dissipated as heat in the panel. The shunt resistor, if selected properly, will have a voltage, not 0 volts, across it and dissipate nearly the same power that would have been delivered to the battery. The result is that the temperature of the panel will remain more constant.
OK, so this is only a minor point but I wanted to explain it thoroughly. I suspect that environmental conditions, such as clouds passing by, vary the temperature far more than that caused by any power regulator. Besides, the manufacturers say there isn't a problem here anyway.
I used the high current 41 Amp MOSFET because when operated at low currents it will dissipate only negligible power and prevent over heating.
The optional heater resistor can be used to heat something such as water or the house in winter. It's not needed. Normally the transistor dead shorts the panel. This is not an error. It's standard practice for solar panels to be shorted in this way. There is no harm to the panel doing this.
If you do use a large heat sink on the transistor this regulator is capable of much higher currents. Possibly as high as 41 Amps. (I haven't done this myself yet.)
The MOSFET power driver circuit probably should be improved for very high currents during regulation. The best improvement would be to use a D flip flop between the voltage monitor and the transistor gate. This flip flop is clocked at about 100 Hz. This will cause the transistor to have a controlled and predictable gate pulse.
Here's the Best Solar Shunt Regulator. It's more complicated but it works better because it's temperature compensated.
I have adding temperature compensation to the basic shunt PV regulator. Temperature compensation is needed when the battery pack is stored in a location where the temperature is not controlled, such as outside and near the PV panel, in hot or cold weather.
Depending on the manufacturer, the charge voltage is adjusted by measuring the battery temperature. The specific value is mainly dependent on the internal RESISTANCE of the battery. This resistance is basically dependent on the CHEMICAL ACTIVITY in the battery. This chemical activity is dependent on the TEMPERATURE.
The construction methods can also affect the change in internal resistance vs. temperature.
A Simple Solar Shunt Charge Controller.
Simple Shunt Charge Controller.
These parts can be obtained from: Digi-Key or Farnell.
5 Amp 40 Volt Diode $0.63us SB540CT-ND 518177
12 Volt 1 Watt Zener Diode $0.26us 1N4742ADICT-ND 368660
1 Amp 50 Volt Diodes $0.60us/10 1N4001DICT-ND 365117
10 Amp 25 Watt NPN Power Transistor Q1 $1.00us approximately
7.5 Ohm 50 Watt Power Resistor RL $3.22us FVT50-7.5-ND
Total $5.17us
shunt2
A Better Solar Shunt Charge Controller.
Shunt Charge Controller.
These parts can be obtained from: Digi-Key or Farnell.
5 Amp 40 Volt Schottky Diode $0.63us SB540CT-ND 518177
5.1 Volt .5 Watt Zener Diode $0.21us 1N5231BMSCT-ND 368970
8.2 Volt .5 Watt Zener Diode $0.21us 1N5237BMSCT-ND 369020
41 Amp 55 Volt Logic Level MOSFET Q1 $1.73us IRLZ44N-ND 637488
4.6 Volt ZM33064 Voltage Monitor $1.74us ZM33064C-ND 633318
1 K Ohm Trim Pot R2 $0.22us D4AA13-ND
2.0K Ohm .25 Watt 5% Carbon Film R1 $0.28us/5 2.0K E BK-ND
24K Ohm .25 Watt 5% Carbon Film R3 $0.28us/5 24K E BK-ND
Total $4.86us
With Optional Heater Resistor
2 Ohm 100 Watt Resistor RL $12.45us AVT100-2.0-ND
Total $17.31us
7 * 7 * .022 = 1.1 Watts. The transistor will get warm but not excessively and without a heat sink.
41 * 41 * .022 = 37 Watts. This is significant and needs to be heat sinked for this amount of heat.
damage
Damage Your Panel With A Shunt Controller. Not!
Will this shorting damage the PV panels?
I have talked to several PV panel manufacturers. All have
said that there are no detrimental effects to their panels.
One of the reasons that you might want to use the optional heater resistor instead of directly shunting the panel is that the panel will run slightly cooler, which is a good thing.
shunt3
The Best Solar Shunt Charge Controller.
| °C | °F | LM50 Output | -.0050 V/(°C * Cells) | -.0040 V/(°C * Cells) | -.0025 V/(°C * Cells) |
| 50°C | 122°F | 0.900V | 12.34V | 12.75V | 13.07V |
| 25°C | 77°F | 0.750V | 13.80V | 13.80V | 13.80V |
| 0°C | 32°F | 0.500V | 16.23V | 15.55V | 15.02V |
| -20°C | -4°F | 0.300V | 18.17V | 16.95V | 15.99V |
| -40°C | -40°F | 0.100V | 20.12V | 18.34V | 16.96V |
Shunt3
Temperature Compensated Shunt Charge Controller.
These parts can be obtained from: Digi-Key or Farnell.
5 Amp 40 Volt Schottky Diode $0.63us SB540CT-ND 518177
5.1 Volt .5 Watt Zener Diode $0.21us 1N5231BMSCT-ND 368970
5.1 Volt .5 Watt Zener Diode $0.21us 1N5231BMSCT-ND 368970
41 Amp 55 Volt Logic MOSFET Q1 $1.73us IRLZ44N-ND 637488
4.6 Volt ZM33064 Voltage Monitor $1.74us ZM33064C-ND 633318
LM2902 Dual Operational Amplifier $0.49us LM2902M-ND 400002
-40°C to 125°C Temp Sensor $1.47us LM50CIM3-ND 630937
1K Ohm Trim Pot R7 $0.22us D4AA13-ND
500K Ohm Trim Pot R8 $0.22us D4AA55-ND
2.4K Ohm .25 Watt 5% Carbon Film R3 $0.28us/5 2.4K E BK-ND
2.7K Ohm .25 Watt 5% Carbon Film R5 $0.28us/5 2.7K E BK-ND
10K Ohm .25 Watt 5% Carbon Film R4 $0.28us/5 10K E BK-ND
20K Ohm .25 Watt 5% Carbon Film R6 $0.28us/5 20K E BK-ND
91K Ohm .25 Watt 5% Carbon Film R1 $0.28us/5 91K E BK-ND
200K Ohm .25 Watt 5% Carbon Film R2 $0.28us/5 200K E BK-ND
Total $7.26us
With Optional Heater Resistor
2 Ohm 100 Watt Resistor RL $12.45us AVT100-2.0-ND
Total $19.71us
How's that?! It's cheap and has temperature compensation too!
The bread boarding went well. I now have the circuit working. The LM2902 buffer Op Amp was required because the LM50 has a high impedance output. This output is not designed to drive even very low power powered loads.
The LM50C is a new generation SOT-23 part from National Semiconductor. It reads temperature from -40°C to 125°C. The interesting feature of this sensor is that it outputs 10mV per °C referenced to -50°C.
Vout=(10mV/°C*Temp°C)+500mV for the LM50C
| Temp°C | Temp°F |
