SOLAR SOURCE INSTITUTE Raises the bar…

I just finished up Friday (Yesterday) teaching my new TWO WEEK Photovoltaic and Solar Electric Hands-on Installation Course.  What a great time, everyone worked together to build our “Class 1500W Solar Electric System on Wheels” using nine Solar World SW175 PV panels, one SunnyBoy 3000 Gridtie Inverter and a stacked pair of SunnyIsland 4248 4200 watt 120/240 Battery Inverters installed on the new SOLAR SOURCE Training Trailer/Mock roof. The new addition to our classroom is a real shingle roof 8ft x 18ft on a dual axle Emerson Trailer, custom built by me using welded steel and plywood.

 

 

The whole class also took a Field Trip to visit the One Megawatt Photovoltaic System Installed by Solar Source on the Orange County Convention Center in Orlando, FL.

 

Rick Gilbert and Jim Parish, Solar Source Institute. (The camera caught me saying “shoo it’s bright “

On the last day the class presented me with a gift in appreciation for my astute teaching style and massive information sharing technique. It’s call a “POWER COST MONITOR” and it’s made by Blueline Innovations.

 

All day as my 9000 watts of PV was hard at work it read 00KWH but now that the sun has gone down it is reading 4200 watts being used.  A way-cool device that has it’s transmitter strapped to the meter reading data from the optical port sending data to the desktop readout.

Solar Tracker #4 is now complete…

I installed the last six KC-130’s on Tracker #4 this past Saturday in 40+ mph winds, boy was that hairy but I got’er done.

I setup a temporary ground rack to hold the now orphaned Siemans SR90’s that just feed some power into an old Advanced Energy GC-1000.  I’ll be moving the GC-1000 to a tracking trailer mount I’m putting together with eight old Photowatt 90 watt panels I have.

 

I have installed Nema-3 rain tite disconnects on post near all my trackers.  The tracker assemblies are connected to the system through Multicontact MC-4 latching quick connectors. I’m calling the posts with disconnects “Renewable Energy Access Points”.  I had to separate the tracking mounts from the “Inspect-able Electric System” due to the fact if the tracking assemblies were “Permanent Installations” my building department would require very expensive “Engineered, Signed, Stamped, Sealed Blueprints” with wind-loading of over 120 MPH.  Not really do-able for home-built mounts. I don’t plan on leaving a single PV panel out in the Hurricane if one does hit anyhow. I did pick them up and spread them farther apart as I discovered this past winter they were a bit too close with some shadow overlapping.

   

You’re Hired! I now have a new job.

Friday the 27th of February I was ask to help with a site survey on the roof of a home in Sarasota, FL. A total 15KW Photovoltaic setup on a 5 million dollar home on the water.

After the job I returned to the Solar Source building to meet with the company owner, Wayne Wallace to discuss the possibility of me working for Solar Source as an Instructor for the courses offered by Solar Source University. I met Wayne around 2002 when I asked him to inspect my original Photovoltaic System, Progress Energy’s First Grid Interconnected system. I accepted and I’m proud to be working in the renewable energy field helping train everyone from Inspectors to Contractors.

I was asked by Wayne to attend their 3 day training Seminar which I enjoyed a lot.  I really think I have found my “Solar Nitch”, teaching.

 

Tracker #4 almost finished

I’m waiting on six more KC-130’s to finish off Tracker #4

Then I will move the four SR-90’s to another temporary tracker with the GC-1000 Grid-tie Inverter.

 

UPDATE:  The six brand new KC-130’s have arrived and will be mounted on Tracker #4 when I get back from a brief trip to New York next week.

Solar Pathfinder Report…

I received my Solar Pathfinder site analyses tool last week.  I opted for the cool software for an additional $129 that has a simple to read report.

Here is my Tracking Array #3, which is my NE most tracker.  It’s in the back.

     

         #1                    #2                    #3                    #4

The first is a simple summary, the second is where in relation to the Array was the Solar Pathfinder setup.  A top-down picture is taken over the Solar Pathfinder then imported into the software where you trace the reflections of all obstacles.  The software then creates the digital shading data picture. You can see at the 12N position on report #3 the tracker that is directly to the south of this one.

The system was entered as a two-axis system, which means in this case East to West tracking and seasonal tilting.

Kinda cool.  Most fixed roof arrays give 4-6 “solar hours” per day, trackers kick that up to 7 to 8+ solar hours per day.  The solar hours are listed on the bottom of report #4 as “Sun Hrs.”

TRACE SWCA, an Expensive device to break.

I had one of my two $175 TRACE/XANTREX SWCA Inverter RS232 interfaces quit working on me a few weeks back.  I had opened these units up once to install power through an isolation low voltage drop diode to power a Bluetooth serial adapter.  They are mostly SMT inside with the exception of the MPU which is an Atmel AT89C2051.  The other SMT chips include a 74HC14, 232ACBN, 74HC86, 24LC04B, DG417, and on the back side a PCF8574AT.

     

The interface uses a standard 4 conductor flat telephone cable.  How the RJ11 terminal is crimped on the wire is important. Here is the pin-out as it ends up at the RJ11 to DB9F (provided with the SWCA): RED = PIN2, BLACK = PIN3, YELLOW = PIN5.  The GREEN in all I have seen is not connected.  I use this to provide 5V on PIN9 of the DB9 to power my Bluetooth to Serial converter which accepts 4-9 volts on PIN9 as POWER.  I found +5V on the PCB off a SMT Capacitor on the power rail.

The SMT part that had failed was the DG417 Analog switch.  From what I can see it is what turns on the DATA OUT line which is the RED wire when the SWCA is selected using its preset address.  The address is set using the antique DOS software that comes with it.  You can actually command up to 8 SW series inverters (SW5548 for instance) which requires you to connect the SWCA as the ONLY device on the serial port to “SET” the ID from 1 to 8 in the flash memory.  If you Parallel SWCA’s with the same ID there is no communication because of data collisions.  When I put my defective SWCA on with my good one the good one quit working so it was simple enough to trace back the data wires from the DB9 to the PCB.  I found the RED wire going directly to PIN-1 of the DG417.  PIN-8 is the data in so I just measured across the pins and got like 84 ohms.  The good SWCA measured infinite ohms so I knew the chip was shorted.  I ordered some from Digi-Key for $2.28 each, they should be here tomorrow.  I’m hoping the suspected ESD didn’t get past the DG417.  I could still send commands to the inverter so I know that part is working fine.  Just no data back at me.  The DG417 is an 8 pin SMT device, not real hard to remove. 

I have installed ESD protection on both the DATA LINES.  I use P6KE27CA Solid State devices in all my RS232 projects for ESD protection.  This is the very same device used in the Radio Shack RS232 inline surge protector from many years past.

PARTS DESCRIPTION:

24LC4B (4K I2C Serial EEPROM),  PCF8574AT (Remote 8-Bit I/O Expander for I2C BUS),  AT89C2051 (2Kbyte Flash Microcontroller),  DG417 (Precision CMOS Analog Switch),  232ACBN (High Speed +5V Powered RS-232 Transmitter/Receiver,  74HC14 (Hex inverting Schmitt trigger),  74HC86 (Quad EXCLUSIVE-OR Gate).

UPDATE 08OCT2008:

The part came from Digi-Key right on time.  I soldered in a new DG417 and I’m happy to report SWCA #2 is back in service and fully functional.  The only quirk I have is my two are different versions and the UP/DN arrows on one are reversed which is a pain.

Clay Archer reverse engineered the REMOTE PORT on the SWxxxx TRACE inverters and was kind enough to provide the information on the port.  Here is his letter and the details. 

Here are some of my notes on the SWCA.   I’m in the process of putting together my own interface, initially with a parallel port interface to a PC, and eventually a stand-alone PIC processor with an ethernet interface, when I can find time to work on it.  I don’t intend on making a commercial product with this, but if you develop anything with it I would appreciate some feedback. 

 

SWCA interface pinout: (DB25)

01) V+  5V

02) NC or Gnd

03) Gnd

04) Gnd

05) Gnd

06) NC

07) DB7 (14)

08] DB4 (11)

09) DB2 (9)

10) DB0 (7)

11) NC

12) NC

13) NC

14) LCD Enable (6)

15) Reset

16) LCD Read/Write (5)

17) LCD RS Data/Instruction (4)

18] Read Switch/Enable

19) Write LEDs enable (latch)

20) DB6  [13]

21) DB5  [12]

22) DB3  [10]

23) DB1  [8]

24) NC

25) NC 

 

LED and switch bits:

      LED                     Switch

DB0   over current              set+

DB1   float                     set-

DB2   AC2                       menu up

DB3   inverting                 menu down

DB4   error                     menu ->

DB5   line tie                  menu <-

DB6   AC1                       on/off

DB7   bulk                      Gen

 

Notes:

LCD is an HD44780 based LCD display (I think) with 2 lines &40 chrs.  Numbers in () are LCD module pins. 

 

 You can find documentation on the HD44780 based LCD on the internet. The SW inverter interface (DB25) port was initially designed as an external keypad and display, basically a  duplicate of the internal keypad/display.

 

 The SWCA is a PIC processor designed to emulate the keypad/display and interface to a serial port as a display terminal emulation. To talk to the SW inverter you will need to emulate the HD44780 registers to read data and emulate the pushbutton switches to request data from the SW inverter. You could also just build an external display. Parts used in the internal display are two 74HC573  (8 bit latches to latch the LEDs and pushbuttons) one 74HC14 (to debounce & buffer the enable for the LCD display) and an LCD display with assorted resistors & cap’s etc. I have a rough schematic of the internal display, if you want I can scan it and send you a copy. 

 

Good luck,

Clay Archer

MORE is BETTER! Solar Power That is!

Solar Power Tracker #4 went ON-LINE this past Weekend on temporary wiring.  I still have to dig the conduits back to the main junction box where the home runs are.  Since I have the construction down to a semi-fine science I timed the construction of number four, working at a daily pace with 90% of the materials needed on-site.  The time required, 10 days from start to power-up.

Base construction.

   

For now I only have six Kyocera KC130’s to put on it.  I have 24 Siemans SR90 Photovoltaic panels from my orginial roof setup, 4 are now on tracker #4 feeding my Advanced Energy GC-1000 grid-tie inverter still mounted on tracker #2.  I’m thinking I need to move it.  Since I don’t own enough PhotoWatt PV panels to make six panel arrays, just eight still on the roof feeding the old four panel array system, I’m going to put all eight on a smaller tracker just for the GC-1000 to use.

   

Solar Project Update

The past 10 days have been exhausting, a lot of work in 95 degree heat out in the garage and out at the Solar Tracker Farm.  The “Solar Power Wall” in the garage was built, painted and outfitted with various solar electric equipment.  It’s now 90% finished, I still have to relocate the “whole house” 5.3KVA Best Power UPS to the far end of the wall.  I do some R&D work for Apollo Solar, here are my two Apollo T80 MPPT charge controllers, the one on the left is for testing various solar related things I’m working on.  My current project is an intelligent interface between the T80 and Trace/Xantrex SW series of inverters.  It’s a bit difficult getting the inverters and the T80 to SELL power efficiently without them interacting.

   

I have a master combiner box where all the six panel arrays come in to a KLKD 10amp DC fuse.  All the arrays then get combined and sent to a disconnect switch, from there they go to a Ground Fault Protection Device (GFPD)  The GFPD was formally known as a Ground Fault Interrupter (GFI), So as not to confuse the issue as all homes have GFI’s in the AC panel box the GFPD was named for the DC protection device.

 

I have modified all three tracking arrays to hold 12 panels.  This required a lot of steel supports to be added to the underside.  I also changed the counter balance spring mounts using a porch swing “buckle”.  This works very well, it swivels East to West and also during seasonal tilting.

   

19 SEPT 2008:  One of the last of two major projects has been completed.  The move of the 5.3Kva Ferrups UPS that powers all the 15A wall outlets and lighting circuits.  It’s new home is on the far end of the new Solar Wall.

I’m Frazzeled and I will complete this report later.

My favorite place to buy Solar Equipment is from Maverick Solar.

 

My twelve Kyocera KC-120 panels were purchased from AriStar Solar, Give Lee Walker a call if you need anything.

                              

Mis-Tracking Woes…

Now that we here in Florida are into our 120% humidity and afternoon thunder storm season I have noticed some peculiar behaviour by the Solar Trackers.  With some dew or big raindrops on the clear lens the sensors would get hit with some reflected light and move erratically. I used some new white Krylon paint labeled safe to use on plastic.  They seem to perform a lot better now.  I used some 1″ painters tape around the middle of the clear jar.

 

   

This also protects the lid from UV break-down.  The clear jar is PETA but the yellow lid is made out of some plastic that is not UV resistant.

Solar Power System update

It has been a while since I updated my solar power system progress so here is a brief update.  I’m still making room in the garage to build a “Solar Wall” to hold all of the system and upgrade components.  This project includes organizing my new 60 foot steel arch building that my sons and I built last summer.

The “Tracker Farm” now has three almost complete trackers.  Tracker #1 has 12 KC120 panels, tracker #2 has 10 KC130 panels and tracker #3 has just 8 for now.  My Apollo T-80 charge controller is being “fed” with 6 panel arrays.  The mount on tracker #1 is much heavier duty than the others so I was able to fit 12 panels on it.  I settled on 10 for the rest which makes for a slight 6 panel array problem that I solved by combining panels on different trackers to make six panel arrays.  Tracker #2 has on 6 panel array and 4 left over so I will have 10 on tracker #3 eventually but only 8 now.  Tracker #3 then has one 6 panel array and 2 left over that I have temporally combined with tracker #2’s left over 4 to make a six panel array.  When I install tracker #4 it will have 10 panels and I will put two more on tracker #3.  What this boils down to is three 10 panel trackers or 30 panels broken down into 5 six panel arrays.

Here are the present arrays, left to right, tracker #3 is waiting for two more KC-130’s.  My report of present power production is as follows. 10AM on a full sun day the trackers are making in July 48 amps and the roof tops are making 28 amps.  There are 3780 watts on the trackers right now and 4080 watts on the roof.  At 5PM the trackers are making 50 amps and the roof tops are at 12A.  After 5PM the trackers continue to make diminishing power while the sun is directly west leaving the roof tops in the shade at less than 3 amps from reflected daylight.

My conclusion is as yours should be, keeping the PV Panels as close to 90 degrees to the sun makes for a LOT of extra power.  The mainstream nay-sayers tell you it’s more economical to just install more PV Panels on a fixed mount than to track.  To this I say phfsssssssst.

TRACKING PV PRO’s:

1) The first & foremost, MORE POWER!

2) Easier to discover dirt, bird droppings and clean the panels.

3) Easier to take down for hurricanes. (Florida)

4) Easier to maintain.

5) They provide great shade, a few lawn chairs underneath, a cold drink…

6)  You get to watch them, look at them, talk to people about them and enjoy them!

TRACKING PV CON’s:

1) Easier to run into with things like head, lawn tractor, car.

2) Easier for thieves and Vandals to get to them. (makes video cameras necessary in my case)

3) Cost.  They cost more for the hardware unless you build them yourself like I did. I used old C-Band satellite dish mounts.

5PM UPDATE for 07/09/2008:

At 5PM in full sun, blue sky and 88 degrees the T-80 MPPT charge controller on the trackers was showing 52 amps solid from 3780 watts.  The Trace C40 non-MPPT controller on the rooftop arrays was showing just 28 amps solid from 4080 watts. I rest my case.