J. Mike Rollins (Sparky) [rollins@wfu.edu]
  CISSP, GIAC GPEN
Hyperbola New
My Cats New
Kitty New
Mike is on
  LinkedIn
  FaceBook
BackYardGreen.Net
HappyPiDay.com
Green Cycle Design Group
CamoTruck.Net
  Resume  
  My Stuff  
  Art  
  My Truck  
  People  
Electronics
   AC Inverter
   Taz (My EV)
   Ethernet
   PIC Switch
   LM3524 Motor Control
   Microcontroller
   PWM and MOSFET
   SMP with 317
   Power Supply
   Function Gen
   Simulator: Zener
   Charge Controller
   IR Audio
   PWM with 555
   Solar
   Entertainment Center
      Thermostat
      Thermostat Schematic
      Emitter Follower
      Thermistor
      Switch
      Power
      H
   BLDC
   GPS & WD-C2401P
Jacob's Ladder
Scripts
Math
Notes
My House
My Cars
My Cats New
My Jokes
Pi Poetry
pumpkin
Toro Mower
Development
Speed of a Piston
Not a Pipe
Linux
















Disclaimer:The following are my notes. As I am learning electronics, I am making my notes available. I hope they will be of benefit. However, I do not guarantee the accuracy of my work. I recommend the reader exercise critical thinking.
Power

My thermostat circuit is fairly basic. When the temperature passes a threshold, a relay will close providing power to the fans.

I had a periodic problem with relay chatter. When the fan kicked-in, the relay would click a few times before stabilizing. I spent many hours trying to figure out the cause.

My power supply consisted of a 120 volt to 12.6 volt AC transformer, followed by a rectifier, followed by capacitors and a voltage regulator.


In an attempt to explain the relay chatter, I considered the voltage ripple as a cause. I found an equation for voltage ripple. With a full wave bridge rectifier operating on a 60 Hz source, the equation is equation becomes current/(2 * 60 * capacitance).


My initial design included a single capacitor of 470 uF. A single fan pulls about 200 mAmp. This means the voltage ripple would be about (0.2 Amp)/(120*470 uF) = 3.55 volts. This could cause all kinds of problems with my 12 volt circuit. I needed to account for the the significant impact of voltage ripple when powering all the fans pulling 2 to 3 Amps.


The transformer is a 120 volt AC to 12.6 volt AC transformer. When I measured the output, it actually read 14.0 volts AC on my digital meter. Next, I pass this through a fuse and then to a full wave bridge rectifier.

When building the system, I tested the output of the rectifier with a 220 uF, 30v capacitor and no load on the system. I found the DC voltage to be 17 volts! This puzzled me -- I forgot about the Root Mean Square voltage rating system for AC voltage. For example, a 120 volt outlet has an average, or Root Mean Squared, voltage of 120 volts. The peaks of the AC oscillations are actually about 120 * 1.414 volts, about 170 volts.

Let's assume we have 12.6 RMS voltage from the transformer. If I calculate the peak voltage, 12.6*1.414, I get about 17.8. Factor in a diode drop or two for the bridge and I come close to the value I measured at the capacitor.

I decided to use two 4700 uF and three 1000 uF capacitors for a total of 12400 uF!


The voltage ripple for 3 Amps is now estimated at (3.0 Amp)/(120*12400 uF) = 2.0 volts. I measured 17.3 volts DC at the capacitors. A ripple of 2.0 volts could bring my voltage down to 15.3. The 12 volt regulator requires at least 14 volts input. I hope this problem is solved.

I found some datasheets for the 7812 voltage regulator. The sheet from Motorola recommends an input bypass capacitor under certain conditions (page 14). Just in case, I used a 1.0 uF tantalum capacitor beside each voltage regulator. This is the little yellow thing at the base of the voltage regulator.

The three voltage regulators for three separate circuits. An additional 3300 uF capacitor follows each regulator. I recently learned the value of fuses, so each circuit contains a one Amp fuse between the capacitor and the output terminal.