Re: Interfacing wind turbines to the grid
Reply #34 –
DESIGN CRITERIA OF VERSION 2:
1] Fully Automatic. No need to check on things or flip switches every hour. I can leave home for a week and return to a happy system.
2] Reliable. Fail Safe. Fuse protection. Known ratings on all components, suitable for their purpose.
3] Self-Contained. Despite having numerous inputs, and dependent on an external Microinverter, otherwise "just wire it up".
I just finished installing and getting it "commissioned". I had just as much fiddling around to get it installed as the first one, but that's the fault of the manufacturer ("me") not thinking through a bunch of the installation process, and forgetting to hook up a few bits inside the box before trying to install it over the battery system. Now that I've finished the fussing around, it works well.
It's controlled by a PIC micro-controller, a compact IC with 8 pins. I gave it a tiny DC buck converter for its power supply. The AC is switched with a micro relay and the relay is controlled by a FET. Just a few other bits to provide input filtering of the PWM signal coming from the charge controller's resistor and a LED to tell me the ON state. All the rest is managing interfaces around the box.
HOW IT WORKS:
The controller passes the DC right through, using some of it as power for the Microcontroller as its brain. I've used a PIC microcontroller in this one, because I wanted to drastically simplify it and also because I was feeling old-school at the time. Although I had to build a power supply for it, rather than buy it, well frankly I wasn't having much luck finding a DC-DC power supply that went from 60V to 5V that didn't cost 100 bucks and cover half a square foot of circuit board. I found a very simple compact DC buck converter in a 5-leg package that easily installed on my prototyping board. I just had to feed it with inductors and capacitors, and for such a low-power purpose this thing works great.
I built the control board on two simple proto-boards (photo above):
Each is powered by the 48-60V supply from the batteries. The PIC operates at 5V so there's a DC Buck control used as a voltage regulator. For those interested:
- Texas Instruments LM2576HV
- Motorola PIC12F683 (has been flashed with an interpreter by PICAXE)
- International Rectified IRL540N power MOSFET, N-channel
- Panasonic DPST relay
Using a PIC micro-controller meant I went back to basics to program it, but my program is very simple. In words only (pseudocode):
INITIALIZE:
- declare pin 4 as output ' controls the gate of the FET
- set pin 4 low ' turn the FET off, to start
- declare pin 2 as input ' detects the state of the charge controller diversion
MAIN:
- if pin 2 is LOW GOTO QUIET ' The diversion is off
- if pin 2 is HIGH GOTO ACTIVE ' The diversion is on
- End MAIN ' it never gets here, just oscillates between the two subroutines
sub QUIET:
- set pin 4 LOW ' turn the FET off, Microinverter stops
- wait for a minute ' The micro-inverter is off
- GOTO MAIN ' Loop back to where we test pin 2
sub ACTIVE:
- set pin 4 HIGH ' turn the FET on, activates the Microinverter
- wait for a 20 minutes ' The micro-inverter is active for this time
- GOTO MAIN ' Loop back to where we test pin 2
That's it, aside from switching an LED on and off, and spitting the On/Off status out to the COM port, it doesn't need to do anything else.
If it's windy, the diversion load will be active, and trigger the controller. If the wind stops, then within about 20 minutes the controller will shut down. If the wind continues, then the controller continues. If the battery gets low, the diversion won't activate, therefore this controller won't activate. The controller doesn't run the battery down because the diversion doesn't run in those conditions either. The controller only operates when it detects voltage at the "dump load" heater resistors.
Topic: Interfacing wind turbines to the grid (Read 2818 times)
previous topic - next topic
