Line Regulation
If you disable the feedback loop of a SMPS and let it run like that, you will notice that voltage is more than double what it should be. You'll also have all the caps blow up in your face. 
Does that mean you can get double voltage out of the power supply just by using the appropriate capacitors? Well, no. The inductor will most definitely not like it, but that's not what this post is about.
If you put a conventional unregulated power supply side by side with a SMPS, you will notice that the transformer and capacitors are several times smaller in the SMPS. "That's because the conventional transformer runs at 50Hz so it has to be bigger and the caps have to hold up the voltage for longer" you'll say. Correct. But this is only half of the truth. A 50Hz unregulated DC supply relies on a high quality transformer and a large capacitor bank to keep a steady voltage output. We won't talk about the linear regulator here, because everyone knows its mode of operation - it simply burns up excess power.
Now, the output caps in a SMPS are smaller because a SMPS runs in the 10s or 100s of kHz, where a linear supply runs at 50Hz - 0.05kHz! However, we have overlooked one important aspect. The voltage input to the SMPS is still 50Hz!!! And yet, the primary capacitors located after the bridge rectifier are still much smaller than would be required for a linear supply of the same power.
Back to our "double voltage" power supply. Suppose you have installed higher voltage caps and the inductor doesn't get hot. You connect an audio amplifier to this power supply you just made. Result? An impressive amount of line frequency hum. How did this happen you might ask?
Because of the small primary capacitors, the DC voltage entering the SMPS transformer has 10-20 volts of 50Hz ripple riding on top of it. 20 volts of ripple. You read that right. Now, assuming the primary to secondary turns ratio of the transformer is 10:1 (just to keep things simple), the ripple appearing at the output would be 10 times smaller. That is still 2 volts of ripple, which is unacceptable. The output filter of the SMPS will remove very little of that because it is designed for a much higher frequency.
Here's where the control loop comes into play. The fast feedback loop basically "rides" the 50Hz ripple, along with any other fluctuations that might appear on the line, and continuously adjusts the duty cycle of the switching transistors, to maintain a steady voltage at the output. Now, this voltage it rides doesn't come out of nowhere. This is why the peak voltage available at any given moment must be significantly higher than is required at the output - the largest disturbance that the control loop can correct is given by the voltage reserve available at the transformer. This is also why you see diodes with peak voltage rating of 100 volts on a 12 volt rail.
I hope you have now understood another aspect of SMPS operation.
Does that mean you can get double voltage out of the power supply just by using the appropriate capacitors? Well, no. The inductor will most definitely not like it, but that's not what this post is about.
If you put a conventional unregulated power supply side by side with a SMPS, you will notice that the transformer and capacitors are several times smaller in the SMPS. "That's because the conventional transformer runs at 50Hz so it has to be bigger and the caps have to hold up the voltage for longer" you'll say. Correct. But this is only half of the truth. A 50Hz unregulated DC supply relies on a high quality transformer and a large capacitor bank to keep a steady voltage output. We won't talk about the linear regulator here, because everyone knows its mode of operation - it simply burns up excess power.
Now, the output caps in a SMPS are smaller because a SMPS runs in the 10s or 100s of kHz, where a linear supply runs at 50Hz - 0.05kHz! However, we have overlooked one important aspect. The voltage input to the SMPS is still 50Hz!!! And yet, the primary capacitors located after the bridge rectifier are still much smaller than would be required for a linear supply of the same power.
Back to our "double voltage" power supply. Suppose you have installed higher voltage caps and the inductor doesn't get hot. You connect an audio amplifier to this power supply you just made. Result? An impressive amount of line frequency hum. How did this happen you might ask?
Because of the small primary capacitors, the DC voltage entering the SMPS transformer has 10-20 volts of 50Hz ripple riding on top of it. 20 volts of ripple. You read that right. Now, assuming the primary to secondary turns ratio of the transformer is 10:1 (just to keep things simple), the ripple appearing at the output would be 10 times smaller. That is still 2 volts of ripple, which is unacceptable. The output filter of the SMPS will remove very little of that because it is designed for a much higher frequency.
Here's where the control loop comes into play. The fast feedback loop basically "rides" the 50Hz ripple, along with any other fluctuations that might appear on the line, and continuously adjusts the duty cycle of the switching transistors, to maintain a steady voltage at the output. Now, this voltage it rides doesn't come out of nowhere. This is why the peak voltage available at any given moment must be significantly higher than is required at the output - the largest disturbance that the control loop can correct is given by the voltage reserve available at the transformer. This is also why you see diodes with peak voltage rating of 100 volts on a 12 volt rail.
I hope you have now understood another aspect of SMPS operation.
Those resistors are found in a number of places for several reasons. They serve to discharge the primary caps (for the safety of us who service them). Sometimes you will find them across the X caps as well.
But it is true that some manufacturers don't fit them... don't ask me how i know. 
Comment