Time to update the stickies, so im thinking of adding a load of general detail about mobos, hopefully including details on power phases, D-PWM, chipset behaviour and details on how the whole lot works.
If theres anything you want covered, PM me OR submit (preferably) indepth info on things yourself. Doesnt have to be me who deals with these stickies.
Feel free to point out mistakes, oversights etc as info is added.
Voltages and PWM
Figure 1 shows a perfect steady-state (stable voltage) of voltage V that switches from 0 (off) to supply voltage under ideal conditions... switching time is zero (impossible). Nothing special, and definately not life-like
Figure 2 shows a more realistic voltage. The voltage is turned on at time 0, but voltage does not rise to supply voltage instantly, takes a finite amount of time, shown by the non-vertical line. Another point to note is the over-shoot, or spike which occurs before the voltage sinks back down to the desired level. This can be controlled to some extent in a real circuit by a capacitor connected to ground, which will be explained later. The "steady-state" voltage is not actually steady, it fluctuates depending on power supply stability, the load it powers and temperature,amongst other factors.
Figure 3 shows a perfect PWM signal of 50% (not really important..yet) The steady-state voltage is the same, but it is only on for 50% of the time, so the voltage to the load is half of what it was in Figure 1. In real life, say a fan is spinning more slowly, shifting less air, making less noise. But thats not real life!
Figure 4: real PWM, more or less. Again, there is the overshoot when the voltage is turned on, and the sinking back to the desired voltage. Notice also that when the voltage is turned off, the drop is not instant, it takes time (the fall time), but the switch is faster than the rise time. I missed the fluctuating "steady-state voltage" just imagine its there
For a given supply voltage, the duty cycle (% of time of the signal is high) can be changed to give more AVERAGE voltage, or less to the load. Apart from this, the frequency of the signal can be changed, but most controllers have a maximum permissible frequency.
Figure 5 shows a low duty cycle, the average voltage to the load is around 25% of the voltage supply.
Figure 6 shows a high duty cycle, around 75% of the voltage supply.
Some maths:
Steady state: voltage is active to the load 100% of the time, voltage to the load is the voltage supply (which fluctuates slightly around this value)
50% duty cycle: voltage to load (VL) is applied for half of the time, voltage to the load is voltage supply/2
25% duty cycle: voltage (VL) is supply voltage /4
75% duty cycle: voltage (VL) is supply voltage/1.33
-> 0.75x1.33 is 1, or 100%. 75/3 is 25, and thats what needs to be added to get 75 to 100. 25 is 1/3 of 75, or 33%
The capacitor to ground:
In a circuit thats completely off, and has been for a while, a capacitor acts as a short circuit (just trust me! lol) When the circuit is powered, think of a capacitor as a gap that gets bigger, until the point where no current (a flow of electrons) can get past the gap. The capacitor, in a steady-state DC circuit is an open-circuit...like something that isnt connected. The first electrons to move along the path to where they want to be (they create the spike) fall into the hole, and no electrons can get past the hole until its full. This deals with the spike. If the hole is too big (capacitor is too big) it takes "too long" to fill the whole and circuit is slow. If hole is too small, the spike is not dealt with properly. The second picture shows a circuit layout to use it in real life. Its good electronic design to always put a grounding capacitor in a circuit this way
picture of analogy, and circuit diagram
-> More to come, as I figure out what goes where to make most sense
If theres anything you want covered, PM me OR submit (preferably) indepth info on things yourself. Doesnt have to be me who deals with these stickies.
Feel free to point out mistakes, oversights etc as info is added.
Voltages and PWM
Figure 1 shows a perfect steady-state (stable voltage) of voltage V that switches from 0 (off) to supply voltage under ideal conditions... switching time is zero (impossible). Nothing special, and definately not life-like
Figure 2 shows a more realistic voltage. The voltage is turned on at time 0, but voltage does not rise to supply voltage instantly, takes a finite amount of time, shown by the non-vertical line. Another point to note is the over-shoot, or spike which occurs before the voltage sinks back down to the desired level. This can be controlled to some extent in a real circuit by a capacitor connected to ground, which will be explained later. The "steady-state" voltage is not actually steady, it fluctuates depending on power supply stability, the load it powers and temperature,amongst other factors.
Figure 3 shows a perfect PWM signal of 50% (not really important..yet) The steady-state voltage is the same, but it is only on for 50% of the time, so the voltage to the load is half of what it was in Figure 1. In real life, say a fan is spinning more slowly, shifting less air, making less noise. But thats not real life!
Figure 4: real PWM, more or less. Again, there is the overshoot when the voltage is turned on, and the sinking back to the desired voltage. Notice also that when the voltage is turned off, the drop is not instant, it takes time (the fall time), but the switch is faster than the rise time. I missed the fluctuating "steady-state voltage" just imagine its there
For a given supply voltage, the duty cycle (% of time of the signal is high) can be changed to give more AVERAGE voltage, or less to the load. Apart from this, the frequency of the signal can be changed, but most controllers have a maximum permissible frequency.
Figure 5 shows a low duty cycle, the average voltage to the load is around 25% of the voltage supply.
Figure 6 shows a high duty cycle, around 75% of the voltage supply.
Some maths:
Steady state: voltage is active to the load 100% of the time, voltage to the load is the voltage supply (which fluctuates slightly around this value)
50% duty cycle: voltage to load (VL) is applied for half of the time, voltage to the load is voltage supply/2
25% duty cycle: voltage (VL) is supply voltage /4
75% duty cycle: voltage (VL) is supply voltage/1.33
-> 0.75x1.33 is 1, or 100%. 75/3 is 25, and thats what needs to be added to get 75 to 100. 25 is 1/3 of 75, or 33%
The capacitor to ground:
In a circuit thats completely off, and has been for a while, a capacitor acts as a short circuit (just trust me! lol) When the circuit is powered, think of a capacitor as a gap that gets bigger, until the point where no current (a flow of electrons) can get past the gap. The capacitor, in a steady-state DC circuit is an open-circuit...like something that isnt connected. The first electrons to move along the path to where they want to be (they create the spike) fall into the hole, and no electrons can get past the hole until its full. This deals with the spike. If the hole is too big (capacitor is too big) it takes "too long" to fill the whole and circuit is slow. If hole is too small, the spike is not dealt with properly. The second picture shows a circuit layout to use it in real life. Its good electronic design to always put a grounding capacitor in a circuit this way
picture of analogy, and circuit diagram
-> More to come, as I figure out what goes where to make most sense
