PCB design and construction

 The SSR mentioned in the last post requires a printed circuit board (PCB).  There are several options.

1. outsource the task and let a company take my data and print the needed boards..

2. make the boards myself. The options expand profusely at this point.

    a. presensitized boards and photo etching

    b. standard  boards and 'iron on' process

        1. with or without lamination

    c. choice of laserjet or inkjet printer

There is much to consider and the choices will not be made in this post. 

Time will tell.

SSR

 

   I'm replacing my custom wired relays with SSRs (solid state relays).

  Traditional relays are electro-mechanical in nature. A coil is energized causing contacts to open or close. The equivalent of  such a coil is referred to as a 'solenoid.'

   All such devices make noise and are susceptible to mechanical wear and corrosion and welding of contacts.

   The SSR is composed of  semiconductor devices, typically, mosfets (metal oxide semiconductor field effect transistors).

   In my design, opto-couplers turn the mosfets on or off. A low power signal is isolated from the mosfet circiutry using these devices. The design and construction is more complex than the relays, but is more reliable, quieter, and can handle more power per dollar cost.

The design I'm using is here:

The first is the linear actuator motor control. The second includes the actuator power supply on/off switch.

Linear Actuator reprise.

    For the last couple of years, the solar array has been stationary and the motor not used. A relay card, rewired by me, burnt up. The actuator motor drew too much current for the copper etch traces when the actuator reached the limit of movement and increased current flow.

   I've installed a limit switch to help prevent future failures. I an also installing solid state relays with a rating almost three times the maximum  required by the actuator motor.

   For the last few days, I've been researching SSR's ( solid state relay). I ran across a suitable design that uses mosfet semiconductors and opto-couplers.

   All day, today, I've been breadboarding and testing circuits. I have prototyped and tested the design I want to use with positive results.

   The next step will be to assemble the final breadboard circuit and install on the Raspberry Pi 3B. The new circuit replaces the custom-wired quad-relay card. The mechanical aspect will be no longer an issue. No welded contacts. No custom wiring. No burnt etch traces.

   The parts come from my stock parts stores with some coming from a Samlex Cotek 1500w inverter that stopped working years ago.

   The extra solar energy will come in handy this winter.

The importance of backup systems.

 Last week, my charge controller pooped out. I decided to replace my inverter too, while I was doing maintenance.

I replaced the charge controller with a Xantrex C60 PWM charge controller. The PWM charge controller is not to be confused with the Xantrex PWM inverter.

I received the pure sine inverter and replaced the PWM inverter.

When the new charge controller arrives, the Xantrex PWM charge controller will go back into storage. The Xantrex inverter, paired with the Xantrex charge controller, will make up a backup system. I also have a backup Raspberry Pi SoC (System on a Card) system data logger and array position controller.

I will have had a system failure and maintained household power availability. This illustrates the importance of backup systems and components.

Inverter poops out.

 Woke up one morning and the Epever Tracer 3210 had quit working.

I got on eBay and started searching. I discovered a newer model, from the same company, with higher current rating.

The Epever Triron 4210 was advertised with a repairable defect. The price was better than I expected. The unit is in the mail and should arrive Wednesday. 

The newer model should be a plugin replacement that requires no modification to the existing software/hardware setup.

Inverter - Modified sine and Pure/true sine

 I decided, recently, to replace the existing Xantrex PWM inverter with a pure sine inverter. The output of the latter is closest to that produced by municipal utilities.

The replacement is a Samlex 2000. That unit is rated for 2000 watts where the Xantrex is rated for 1500 watts.

According to my research, the new unit will be more efficient and produce a better performance from connected devices.

The one thing I dislike is the noise. A fan runs continuously and is noisy. Power is used up unnecessarily. For this reason, I have to turn the power off every night.

I need to rework the fan design or setup automated power boot control.

Alternatively, I could create two different house wiring systems: one for pure sine devices and one for devices using PWM.

Hardware modification

 

The existing mount for the panels could be improved by extending the bottom of the steel pipe support.

The addition of two feet to the bottom of the steel pipe would improve maneuverability. The panels are positioned near the bottom and run into the ground when the panels are rotated to the extremes of position.

This limits the amount of afternoon sun the panels can receive.

Another possibility: add mirrors to the sides and/or top to increase the amount of energy absorbed.  This would increase the energy suplly in Winter when the sun is low in the sky at it's zenith. 

See the YouTube video for details.