Update on technology

The status of consumer two-axis tracking hardware is unchanged.
Elsewhere though, much has changed.

Solar panels are getting cheaper.
New technologies are making panels much more efficient.

These developments will effect the long term future of solar energy for residential consumers.

Excel translator

Hello,

   Someone has written an Excel formulas translator in the Python language.
Python is one of the languages used in my system. I'm going to explore the use of that software as a substitution for my transcription.
   If the program includes forward and reverse transformations, I may be able to save/convert data in excel format.

Backup everything

This post deals with an mportant topic: backups.

All equipment will eventually breakdown.

The only solution is to keep some sort of backup system in stock.
In my case, the current backup is a lower powered version of the present system. This will do until I get something with more power. The current backup will run my light and stereo, provided I don't play the stereo too loud.

I have spare cables and conectors for the solar station.
I have a 300w inverter and a 60A PWM controller.
Panels are not likely to go out or bad for over twenty years. Uless falling tree branches or speeding automobiles are an issue, there's no need for a full backup of panels, though one or two spare panels would be convenient.

If a component goes bad or is damaged beyond repair, I count on at least a week for a replacement to arrive. As an electronic engineering technician, I keep test and repair equipment on hand. I even have backups for those.

My batteries run about $150 each. My inverter runs $150-$250. The controller runs around $150.
If a cloudy day or two occur, I have a quiet 2000w gasoline generator that runs my system and charges the batteries, using an automobile battery charger. The batteries are marine grade dep discharge type.

Using a backup system as emergwncy aid is a good idea.

Actuator power relay switch

Actuator power relay switch installed.

   I ordered and received a two-relay card. I use a four-channel version to reverse the actuator power for forward and reverse movement.
   The two-channel unit controls the actuator power supply battery connection. This unit is controlled from the same script that moves the actuator, hence the solar array.
   When the script runs, now, the two-channel unit turns on and connects the battery bank to the boost converter. The boost converter boosts the battery voltage to that needed to drive the actuator, approximately 32vdc-36vdc.
   Previously, the unit was connected by manual operation of a mechanical switch.

Re-thinking the system

   The system is running well. There are some changes that are needed but are difficult to implement.
   Roof-top installations are the most popular now, but also the most inefficient. I opted for a single axis, sun tracking, ground mounted system. The most efficient is the two-axis tracking system. The hardware needed is being made, just not for consumers (complaint).
   Due to the lack of adequate hardware support by the industry, I've had to innovate my own single axis tracker. There are limitations in the range of movement in the panels. This can be increased by raising the system off the ground and/or adding a linear actuator with a longer reach.

   My current system uses one inch galvanized steel pipe as a mounting axis. Two rows of three panels each are mounted, one above the other. The pipe comes in ten foot lengths. The pipe is tilted at an optimum forty-five degrees, for this latitude. This creates a classic Pythagorean Isosceles  triangle.
The pipe forms the hypotenuse of the triangle. If I extend the length of the pie by adding a short segment to the bottom, one of the angles will change. To maintain the proportions, I will also have to raise the upper end of the pipe, and possibly add a similar short segment to the top end.

   I added a vertical ten foot section of strut to the left side to stabilise the upper and lower rows of panels. I need to add another to the left side.

   In thinking this over, I realize I could extend those struts and add at least one more row of panels to the top.

Of course, adding all this extra weight means I may need to replace the one inch pipe with a two inch pipe.

   I also consider adding a two foot extension to the bottom of the pipe.

   After all this, I'm reconsidering the use of concrete placed rectangular steel tubing along with fixtures to make a dual-axis tracker. The model for this is a unit on YouTube from a Chinese company. I have to consider whether the model is scalable, i.e. whether a larger model will hold more panels safely and scurely. Needless to say, cost is still a factor.

It's that time again.

Every so often, the Raspberry Pi orgnization issues a new release of the operating system for the Raspberry Pi. Most recently, and coincident with the debut of the Rapspberry Pi 4B+, is the Debian Buster operating system. Before Buster were Jessie and Stretch.

When a major update to the operating system is issued, I have to re-format the SD memory card.
Then install the memory image on the card. I have to add a couple of files in order to communicate with the RPi without using an external monitor and keyboard. I can go 'headless' and use my local wifi router and my laptop to run SSH sessions.

Once communications is established, I set some system parameters. Next, I begin downloading and installing programs and other software to run on the RPi.

This process, with testing, can take half a day or more. For this reason, I keep a backup of the SD card once I'm done with updates and modifications. This way, if a card goes bad or some damage occurs, I'll have a spare ready for installation. Just plug and run.

October - more changes

Hello,

   The array is now more exposed to early sunlight. I've cut down trees in the line of sight of morning sun and now get a couple more hours of sunlight. Good timing since the neighbor's tall evergreens and the old oak, on the other side, have begun casting a shadow on the array in the two o'clock  zone.
   I repositioned the actuator arm's connection to the array, further from the center to get a wider range of azimuths. The array begins at 150 degrees and ends at 210 degrees. This can be increased by raising the array high enough to swing East to 180 degrees and West to 270 degrees, and using an actuator with a thirty inch reach. As is, the array will be stopped by ground contact at current exteemes. Since I've oriented the array South, 180 degrees, and 45 degrees elevation, the optimum for mmy latitude, noon is thirty degrees or 5-6 seconds, and evening is 11-12 seconds of on time.
  The array now moves five degrees for each second of on time. A total of twelve seconds moves the array between extremes and across sixty degrees of azimuth.
   I've scrapped the old actuator script and now am using a new manual adjustment script until I write a new acuator automation script.
   I'm considering using the sensor to align the array instead of calculations or interpolation.
There would need to be coding to eliminate stutters and shakes from minute changes in the environment. A sleep cycle lasting five minutes between moves, would work.
   If the ADS1115 should fail, a safety should be triggered to prevent array movement or switch to calculated or interpolated motion and print an error message.
   One of the ADS channels has had a problem for sometime. I replaced the ADS1115 with a new unit and the bad channel is still not functioning correctly. The new unit problem channel has an intermittent signal, meaning the wiring is probably faulty. I think I merely twisted the wires together and applied heatshrink. The signals are susceptible to line noise and this shows in the graphs of the channels. Shielded cable is needed, though the differential between channels and ground is correct. Since the readings are correct with respect to each other and five wire shielded cable is expensive, I'll forego the shielded cable unless actuator movement is effected if I use the sensor for array positioning.
   I need to solder all joints and connections.
   The controller was shutting down USB ports and reopening others, a sort of "wack-a-mole" with USB ports. Eventually, all ports ceased to function. I had read that fusible links were used to protect USB ports from over current states. I have made some unconventional power terminations over the last few months. I decided to keep the controller off line for a few weeks. When I reconnected, I had no problems and have had none since. I now make conventional power terminations. A power switch would be a helpful addition. An update: The problem reoccurred. On the advice of another RPi user, I'll add an FTDI-USB cable/connector. This is supposed to solve the problem. Updating the operating system may also work since the problem seems operating system related.
   A month or so past, I did another water check for the battery bank. The usual units needed water added. I had attached the desulphator device, but a month has passed and I have removed the device.
   As usual, more batteries are needed and more panels are needed. Two more batteries and six more panels are the near term goal. New or altered mounting of the array is also planned.
   Currently, my backup sytem consists of the C-60 Xantrex PW controller (not an MPPT device) and the old three hundred watt sine inverter. The inverter will drive LED lights and the laptop but not much else. Certainly neither of the refrigerators. I'm not sure about the water pump.
   I need to get system extra parts: a new MPPT contoller and a new inverter of at least a thousand watts, preferably two or three thousand watts and that is a pure sine wave inverter.