Solar windowsill info
Rsadeika
Posts: 3,837
A couple of weeks ago I started working with a 12V solar setup. I tried to budget this at $200 or less as a start, but it seems like it is a nickel and dime process that is getting ready to pass my budget limits. The system consists of a 12V 50W solar panel, PWM solar controller, 35Ah SLA battery, 10 Amp 12V to 5V regulator, Activity Board/RPi combination and the nickel and dime stuff like wire, connectors, nuts and bolts, ..., etc.
I chose the 12V system because the essential parts for this are readily available. If you go with something other than a 12V system you will probably have quite an interesting time trying to get the system to work like it is expected. For me their is no "green" philosophy or "saving the earth philosophy" involved, just trying to see if the sun can do some work for me, other than giving me a sunburn or stroke.
My early observations, I have to start thinking about implementing devices to replace the power hungry Activity Board and RPi. Yes, I was surprised to see how much power they really use when you can actually track it with the 12V battery as the power source. The other item that is pretty hungry is the PWM solar controller, I was surprised about that also. I am only talking about a volt or two, but when you hit a non-charging day, you will run dry very quickly.
I am now in the process of adding another battery to be in parallel with the other battery to see if the solar panel can deal with the added work, I have a feeling that I might have two upgrade the panel to a 100W version, but not sure. I am also trying to setup the Activity Board ADC component to have it give me data about what the two batteries and solar panel are doing. The worse part about this is getting the wiring right, I already let out some magic smoke that was within the wire, actually now it is only a lot of black wrinkled outer coating that is left with the cooper innards of course. More to come.
Ray
I chose the 12V system because the essential parts for this are readily available. If you go with something other than a 12V system you will probably have quite an interesting time trying to get the system to work like it is expected. For me their is no "green" philosophy or "saving the earth philosophy" involved, just trying to see if the sun can do some work for me, other than giving me a sunburn or stroke.
My early observations, I have to start thinking about implementing devices to replace the power hungry Activity Board and RPi. Yes, I was surprised to see how much power they really use when you can actually track it with the 12V battery as the power source. The other item that is pretty hungry is the PWM solar controller, I was surprised about that also. I am only talking about a volt or two, but when you hit a non-charging day, you will run dry very quickly.
I am now in the process of adding another battery to be in parallel with the other battery to see if the solar panel can deal with the added work, I have a feeling that I might have two upgrade the panel to a 100W version, but not sure. I am also trying to setup the Activity Board ADC component to have it give me data about what the two batteries and solar panel are doing. The worse part about this is getting the wiring right, I already let out some magic smoke that was within the wire, actually now it is only a lot of black wrinkled outer coating that is left with the cooper innards of course. More to come.
Ray
Comments
A good one should really switch itself into a standby mode if it detects a power input that's less than required to run the charging cycle.
The Propeller should be quite capable of logging battery status for as many batteries as you'd like.
You could even use it to switch on and off the PWM controller if you add a relay or MOSFET.
(Switching relays are nice. Only uses power when they change state. )
I will be connecting the voltage regulator directly to the battery and set up a program for the Activity Board to monitor the battery voltage level, and may have it shutdown the RPi and possibly go to sleep when the battery level approaches, maybe 10.5V - 11.0V. Then how in the heck would I wake up the Activity Board when the battery levels were back up to 13V or more. Not sure at this point if I could even have the Activity Board restart the RPi since it will have been shutdown. There is a lot of exploration to be done.
Ray
This morning I tried to replicate the problem, and sure enough when the solar controller is not on, the readings were as expected from the solar panel. As soon as I put the solar controller on line, the readings from the solar panel were at 12.49V, about the same as the battery readings. It is almost like the battery is backward feeding the solar panel via the innards of the solar controller, I wonder if the manufactures of this controller forgot to put in a diode to prevent the battery from backward feeding of the solar panel.
I guess now I have to figure out some way verifying if in fact that is what is really occurring. Maybe the solar controller unit has some kind of switching device where at some preset voltage the backward feed is shutdown. Or maybe because this is a PWM functioning device, it has something to do with PWM operation. Still scratching my head on this one, I guess I need some advise on this one.
Ray
The solar controller has a preset for when it starts to charge the battery, which is 13.3V. Yesterday their was a couple of hour period where the charge rate was at about 5V, and the three hours at 10.6V. All of that charge capability was basically wasted! I would say that the battery technology, as it concerns storing solar energy, is not up to doing a very good job.
Not sure where the improvements could be made, do you make 12V batteries that have 5V or 3V volt individual cells, or some other low volt cell combination, that could be charged at the 5V - 13.3V condition? And of course you have to keep in mind the amp hour capabilities. If there is going to be a widespread use of solar power, definitely the battery storage capabilities would have to improve.
At the moment my solar setup has an Activity Board/RPi combination for allowing me to sit at my desktop computer to program the Activity Board, and view the charging data. I am thinking of removing the RPi component and having the Activity Board pick up some of the functionality. I have an XBee WiFi module ordered, which I will be able to plug into the Activity Board, and I shall see where that takes me.
Ray
Ray
The last couple days have been a combination of cloudy or sunshine, so the solar controller is working as it should be, I guess I did not fry the thing, or at least not completely. I am also now thinking that with the battery charging conditions that are available, maybe a smart battery charger should be used. Maybe use the house electricity as a backup source with the Propeller Board keeping an eye on the battery status, not sure if such a battery charger exits where the Propeller Board could turn on and off. Or a battery charger that could work with a solar panel setup, I think the solar controller would get very confused, and would probably start feeding the solar panel.
So, it looks like I have a lot to do...
Ray
Now, I have two questions, how do you figure out how big of a capacitor you would need for what I am thinking of doing? The main thing would be to keep the Propeller board from shutting off when it makes the power switch, if that is possible. And does anybody know where you can find the connectors that are being used with the Power Pack cable. To create the circuit I will have to provide a cable the would connect to the Raw Power OUTPUT connector on the Power Pack. I really do not want to cut off the barrel plug on the one cable that came with the Power Pack. I also looked in the Parallax store, but could not find any replacement cables for the Power Pack, maybe they are hidden somewhere.
Ray
Ray
http://www.amazon.com/s/?ie=UTF8&keywords=watts+up+meter&tag=googhydr-20&index=aps&hvadid=31545634277&hvpos=1t3&hvexid=&hvnetw=g&hvrand=17187206206686013551&hvpone=&hvptwo=&hvqmt=b&hvdev=c&ref=pd_sl_900yd7ygyf_b
As to the ADC, if you were measring 12+ volts, was it through a proper voltage divider?
Jonathan
I am still trying to figure out what the heck happened yesterday. My setup, I am using the Li-ion Power Pack and the Activity Board with an XBee module installed. I have a two 10K resister setup for checking the Power Pack status on adc_volts(3). For the solar system battery I have a 3/1 resister setup, 3 10K resisters in series for the positive side and 1 10K resister for the ref ground side, which uses adc_volts(2). With this setup I was getting the correct reading for the Power Pack, but the reading I was getting for the solar system battery was .0000247V, which of course means something was not right. After looking at the circuit I noticed that the solar system battery ground was not tied into the Activity Board, so at this point I added the solar system battery ground to the Activity Board ground socket, and the Activity Board shut off. After hitting the on/off switch a couple of times, still nothing, no green LED(s).
I have a power pack that I made that uses 4 li-ion batteries, 2 in parallel and setup in series, with the batteries fully charged I was getting 8VDC from the power pack. I used this to do a quick check of the Activity Board, at which point, when plugged in, I started to get smoke coming out of the barrel plug socket on the Activity Board. I am using this power pack with the PropBOE and never saw that occur.
After doing a quick check the Parallax li-on Power Pack had the fuse blown, the Activity Board A/D converter chip no longer functions. It looks like the XBee module did not get scorched, but I am not sure about P17, which is what I was using to check the Power Pack. Not sure what I will be doing with this board, I have to figure out a cheaper configuration if I am prone to releasing smoke from electronic parts.
Merry Christmas to Everybody
Ray
/* ABmobile.c */ #include "simpletools.h" #include "adcDCpropab.h" #include "simpletext.h" #include "fdserial.h" serial *xbee; void Bat_Log(); void vtotal(void); float v3,v2,y1,y2; int main() { // Add startup code here. adc_init(21, 20, 19, 18); /* Rx Tx Mode BAUD */ xbee = fdserial_open(11, 10, 0, 9600); pause(50); char inBuff[40]; while(1) { // Add main loop code here. readStr(xbee, inBuff, 40); if(!strcmp(inBuff,"help")) { writeLine(xbee,"Finally got here!"); } else if(!strcmp(inBuff,"sysbat")) Bat_Log(); else if(!strcmp(inBuff,"solbata")) vtotal(); else { writeLine(xbee,"??"); //retval = -1; } } } void Bat_Log() { v3 = adc_volts(0); v3 = (v3*1.99); writeStr(xbee,"System Battery "); writeFloat(xbee,v3); writeLine(xbee," V"); pause(250); } void vtotal(void) { // adc_init(21, 20, 19, 18); int x; for(x=0;x<1000;x++) { y1 = adc_volts(2); y2 = y2 + y1; } y2 = (y2/1000); y2 = (y2*3.72); v2 = y2; writeStr(xbee,"Solar BatteryA "); writeFloat(xbee, v2); writeLine(xbee," V"); pause(50); }Using SimpleIDE.Code Size 31,756 bytes (31,988 total). WHAT THE HECK!
Ray, sorry to hear you are having problems with your solar setup. When a project gets a bit complicated like your solar setup you need to follow a methodical step by step approach to designing, building, and testing it. Here is what I do. Hope it helps.
1 Disconnect or turn off power before plugging or unplugging any component or section.
2 Draw a block or circuit diagram for each section of the project. A picture really is worth a thousand words in this case. Add expected voltage and current readings at appropriate points. I find paint very handy for this.
3 Test each section as you finish it. If possible measure the voltage in to IC's before you insert the chip. Make sure you have the expected voltages and/or currents at the appropriate points.
4 Do a pre-integration check before connecting the sections together. Connect the grounds only together and measure the voltages and/or currents at the appropriate points.
5 Add the rest of the section interconnections. If possible measure the voltages and currents at the appropriate points before inserting the IC's. Measure the voltages and currents at the appropriate points with the IC's in place.
These are pretty general guidelines that will need to be modified or expanded on to suit specific projects, but they will save time, aggravation, and expense in the long run.
The more I think about what it is that I am working with, the more I think that I need a board where the major components are on sockets, that way, you fry an ADC chip, pull the bad one, plug in a new one. To bad the Propeller chip is not available in a small form factor that you could mount to a socket. Since I pulled out my old Propeller Platform Board, that Propeller chip sure looks really big and takes up a lot of room.
Ray
I am still trying to figure out the ADC thing, I might have to purchase an ADC chip, that has support on the OBEX, breadboard with my QS_HIB, and go from there. I hope my local RS has a chip or two on hand. I still have not figured out what happened with the Activity Board, maybe with my QS_HIB setup, I will have less of chance at doing major damage.
Ray
The problem that I am now having is that I can not seem to get the correct value to show up for the avolts function. It seems like every time I call the avolts function I get a different value that shows up. The value for the mvolts function, on the other hand, is a consistent value, which is the actual battery pack value. So, this has me stumped, why am I getting a consistent value for the battery pack, mvolts, and why am I getting an inconsistent value for the 12V battery, avolts? I am using my voltage divider setup as described in the previous post. Am I missing something very obvious?
Ray
/* SAabsun.c */ #include "simpletools.h" #include "simpletext.h" #include "fdserial.h" #include "adcDCpropab.h" serial *xbee; float v3,v2; void menu(); void mvolts(); void avolts(); int main() { // Add startup code here. char inBuff[40]; adc_init(21, 20, 19, 18); xbee = fdserial_open(11, 10, 0, 9600); pause(50); while(1) { // Add main loop code here. if(fdserial_rxReady(xbee)) { readStr(xbee, inBuff, 40); if(!strcmp(inBuff, "help")) menu(); else if(!strcmp(inBuff, "mbat")) mvolts(); else if(!strcmp(inBuff, "abat")) avolts(); else { writeLine(xbee,"Unknown Command."); } } } return 0; } void menu() { writeStr(xbee,"Menu - help, mbat, abat \n"); } void mvolts() { v3 = adc_volts(3); v3 = (v3*2); writeStr(xbee,"mBattery "); writeFloat(xbee,v3); writeStr(xbee," Volts\n"); } void avolts() { // adc_init(21, 20, 19, 18); v2 = adc_volts(2); v2 =(v2*8.50); writeStr(xbee,"aBattery "); writeFloat(xbee,v2); writeStr(xbee," Volts\n"); }Ray
I decided to use the AB and the XBee, as the sole monitoring device for the solar panel and battery. I have the Li-ion power pack that is powering the AB, and I can do a check to see what the voltage status is. Next thing I will be doing is trying to create the circuit for having the power pack either power the AB from the batteries or the attached solar power. Hopefully I can get right combination of caps to work, although the system would not suffer if the AB rebooted after the power switch, will see how that works out.
Ray
I was considering going back to using Spin, just because of the easy manner in which to use the COGs, but now that SimpleIDE PropGCC and simpletools.h working in XMM mode, I am not looking back to using Spin. These of course are my own opinions. Now if I could only automate the softRTC code to get the real time and date somehow without typing in the data...
Ray
Since I am using PropGCC with SimpleIDE, I was hoping that by the new release of SimpleIDE it would have COG availability in xmmc mode. Right now I want to have a visual of the status of the Power Pack voltage, I can check it with the monitor program, but it would be nice to have that hardware visual, that would be more code efficient if done in its own COG. I would also like to have an auto data logging function running in its own COG, that way I could do some analysis of a specific section of the solar system. Yes, I am waiting for SimpleIDE to get fully implemented.
Ray
The system is still the 12V 50W solar panel (windowsill) and a PWM solar controller, with two 35Ah batteries in parallel. I have added a battery charger to the mix. When I had only the batteries on line, and it was the summer months, the sun was almost directly above, hence no direct sun to the solar panel, the best it could do is get the batteries up to 12.30V.
As soon as you have a cloudy day, the solar charge gets up to only about ~12.00V, and the morning starting rate is about 11.70V. As I started to add some devices to the system, the morning starting rate was 11.60V. So, I now have been experimenting with using the battery charger, a couple of hours in the morning, up to when the solar panel starts doing a charge, and a couple of hours, in the evening, after there is no charge from the solar panel. Now, the sun is lower in the sky, and the solar panel charge rates have gone up. Also, the batteries are staying above the 12.0V level.
Since I have to turn on/off the charger manually, I am now thinking on how I could automate this process. I was looking for a reasonably priced smart charger, but that does not exist, or at least I have not found one.
Right now the system consists of the things mentioned above, and I have added: Propeller Platform Board, Raspberry Pi, Staff Badge, 5V 10Amp regulator, and a 9V 1-2 Amp regulator.
The 5V regulator has a female USB connector, for use with any of my devices that need to be charged, like my Power Bank, as one example.
The 9V regulator powers the Propeller Platform Board (PPB), which powers the RPi. I have the Staff Badge plugged into the RPi USB, and I also have the PPB plugged into the RPi USB. I hope SimpleIDE for the Raspberry Pi gets updated, real soon now.
After making some software updates, to the Staff Badge, it is now working as my automatic clock updater, for my other badge. As a prototype setup, this is working as expected. I intend to add more services to the Staff Badge, and the PPB, probably the sht11 module, as I start to expand this system.
Ray
In my two 12V battery, in parallel setup, with a 50W solar panel I can now state that during the winter season, the batteries are brought up to a 13.7V charge. And during the summer season, the batteries are brought up to a 12.7V charge. Because this is an indoor windowsill project my definition for winter season is, the sun is low in the sky and the solar panel gets direct sun exposure for at least five hours. The summer season is, the sun is high in the sky and the solar panel gets indirect sun exposure for at least eight hours.
To make this system more efficient I am considering adding another 50W solar panel and have some kind of circuit where I could switch the solar panel wiring from, in series, during the summer season too, in parallel, during the winter season. It would be nice if this could be done automatically, but I am finding it very difficult to even come up with a manual circuit.
The other addition that could be of use to this setup is some kind of regular power backup. I have been using a standard battery charger, but the charger is not smart enough to have it on all the time and have it just start to work when the battery voltage gets down too, lets say 12V. I did an Internet search and did not find any chargers that would do the job and come in at less than a $100.
I did add a couple of new gadgets, a 12V DC to 5V DC @ 3Amp regulator that has a female USB connector. I tried this charging session with my Nook tablet, and it works as expected. I can now attach the Raspberry Pi 3, which really needs about 2Amps, when you start to load up the Pi. Now I will be able to add, maybe the QuickStart board, and have it powered and programmed from the Raspberry Pi. Since I also have a 9V with a barrel plug contraption also, I can attach other Propeller boards.
If I can figure out how to resolve the problems I just mentioned, this could be a very low cost solution for powering up and charging things. So far I think if you put this setup together from scratch, you could could keep the price down to less than $500, and that would be with a two solar panel system.
Ray
PS - I would be tempted to connect the two panels in series permanently and use a switching regulator to reduce the output voltage to 13.7V. That is simpler and should provide better overall efficiency.
I am using a Renogy 50W 12V solar panel which is rated at:
open-circuit voltage - 22.7V
optimum operating voltage - 18.5V
My concern is frying the solar controller that I have, even though it is a PWM, 12/24V unit. If I have two panels connected, in series, directly to the solar controller, during the winter season the output could get as high as 37V. I guess I could heat the room with that output. :-)
The switching regulator sounds like a feasible approach, although it would have to be a pretty hefty regulator. When I have looked for switching regulators, on Amazon.com, I do not remember seeing any that were in the 37V range, that are affordable and function as expected, plus now I am not sure what kind of Amps it would have to be. I like this idea, I checked my solar panel output this morning at 8:00AM, with a heavy overcast, with rain moving in, the panel was putting out 7.9V. With two panels in series it would be ~18V, the only problem with this is, what, in terms of battery charging power? With this scenario, the charging day could be as long as twelve hours, I think.
Ray
Is this an MPPT controller? If so, perhaps all you need to add is a cirtcuit to automatically switch between series and parallel connected cells based on the input voltage from the solar cells .
Adding an automatic series/parallel switch would reduce the input voltage to a more reasonable range for a switching regulator as well.
The light level affects the total power from the solar cell, so on heavily overcast days you will not get much battery charging power. A single panel puts out 50W, so on a sunny day it could provide as much as 4 amps to charge the battery, but on an overcast day you will get a small fraction of that. Two panels could double that.
Before you go spending money on added electronics you should take a look at the specs of your current controller or measure the current and voltage it provides to charge your battery under various conditions. If it is an MPPT controller additional regulators may not help, while adding a second solar panel would, as long as the controller can handle the additional current.
Ohm's Law I=P/E
So:
Solar panel optimum operating voltage - 18.5V
Solar panel is a fixed? 50Watts
= 2.77778 amps
Solar panel rating 12V
50Watts
= 4.16667 amps
So, if you reduce the voltage you get more amps, I have to make an assumption that my PWM solar controller is doing that.
This morning at 7:00AM I checked my solar station:
Solar panel 12.78V - no direct sun, sun is fairly high on the east side. Panel is facing generally to the south.
Batteries 12.72V - not sure what this reading is describing, battery charging level or actual battery state.
If I had a two panel, in series setup, the solar controller would be dealing with 25.56V. Since my PWM solar controller is a 12/24V, automatic set, does this mean the 12V batteries would be considered to be 24V now, and treated that way? I get the feeling that you have to be very careful about making equipment assumptions and how they work.
A general observation, yesterdays reading of the solar panel was 7.9V and todays reading, at about the same time, 12.78V. The way my panel is located I am not getting any direct sun, this morning, and because the sun is still on the east side, I can say that it could be called very little indirect sun. This monocell panel is highly sensitive to any wave length change?
Ray
It’s a bit of both. You would need to disconnect the battery from the charger and measure or control the load current to get the battery state.
Absolutely. It sounds like it will work with both 12V and 24V systems, but you really need to get the controller specs to be sure.
There is a surprising amount of indirect light coming from dust scattering, water droplet refraction, and reflection from clouds in the atmosphere. Solar cells do operate from a specific range of wavelengths, but I would not say they are “highly” sensitive to wavelength.