Difference between revisions of "Switching regulator"

From OpenCircuits
Jump to navigation Jump to search
(yet another website discussing switching regulators)
(fix broken link)
Line 52: Line 52:
 
* [http://www.national.com/onlineseminar/2001/bpease/switching_controllers.html National: "Designing DC-DC Power Supplies Using High Performance Switching Controllers"]
 
* [http://www.national.com/onlineseminar/2001/bpease/switching_controllers.html National: "Designing DC-DC Power Supplies Using High Performance Switching Controllers"]
 
* "High-voltage, low-noise dc/dc converters" http://www.edn.com/article/CA6582859.html?spacedesc=readersChoice and http://www.edn.com/blog/1700000170/post/440031844.html : has nice photos of dead-bug solid-copper-plane wiring style.
 
* "High-voltage, low-noise dc/dc converters" http://www.edn.com/article/CA6582859.html?spacedesc=readersChoice and http://www.edn.com/blog/1700000170/post/440031844.html : has nice photos of dead-bug solid-copper-plane wiring style.
* [http://www.national.com/nationaledge/mar03/article.html "Comparing Topologies and the (Design) Rules of the Game"] by Sanjaya Maniktala 2002 ... emphasizes "r", the ripple factor, the ratio of the ripple in the inductor to the average current in the inductor, which applies to practically every SMPS topology. It describes "The formal design procedure for any converter design".
+
* [http://www.ti.com/lit/wp/snva575/snva575.pdf "Comparing Topologies and the (Design) Rules of the Game"] by Sanjaya Maniktala 2002 ... emphasizes "r", the ripple factor, the ratio of the ripple in the inductor to the average current in the inductor, which applies to practically every SMPS topology. It describes "The formal design procedure for any converter design".
 
* [http://www.dos4ever.com/battery/battery.html "An Electronic 90V Plate/Anode Battery: A "power"-inverter which emulates the 90V plate/anode battery for vintage battery tube receivers."] (battery powered) a web-log by Ronald Dekker. A very well documented series describing the entire process from the original idea. It seems that everyone who designs a switching regulator hits several unexpected problems, and this description does not shirk from describing the particular unexpected problems seen here.
 
* [http://www.dos4ever.com/battery/battery.html "An Electronic 90V Plate/Anode Battery: A "power"-inverter which emulates the 90V plate/anode battery for vintage battery tube receivers."] (battery powered) a web-log by Ronald Dekker. A very well documented series describing the entire process from the original idea. It seems that everyone who designs a switching regulator hits several unexpected problems, and this description does not shirk from describing the particular unexpected problems seen here.
 
* Flemming Frandsen made a design (based on Bob Blick) that takes noisy car power (8 to 16 volts) and converts to clean, regulated 12 V power. (Its SEPIC topology can convert up and down). A shutdown circuit turns it off when you take the key out of your car.[http://dren.dk/carpower.html]
 
* Flemming Frandsen made a design (based on Bob Blick) that takes noisy car power (8 to 16 volts) and converts to clean, regulated 12 V power. (Its SEPIC topology can convert up and down). A shutdown circuit turns it off when you take the key out of your car.[http://dren.dk/carpower.html]

Revision as of 22:47, 25 November 2012

A "voltage regulator" is designed to hold a constant output voltage.

A "switching voltage regulator", usually called a switching regulator, also called a DC to DC converter, accomplishes this goal by rapidly switching a transistor from full on and full off.

A "linear voltage regulator", usually called a "linear regulator", accomplishes the same goal by holding a transistor in its active region, somewhere between full on or full off. Basic Voltage Regulators are generally linear regulators.

The 723 Voltage Regulator can be used as either a linear or a switching regulator.

The Linuxstamp uses a the LTC3407-3 switching regulator.

The ARMUS Embedded Linux Board uses ...

The MCP1252/3 Flyback Switching Regulators can generate an output voltage greater than the input voltage -- something impossible with linear voltage regulators.

While learning about SMPS, teachers typically try to gradually ease you into understanding them by starting with perhaps the easiest-to-understand switching converter, the buck converter. Then they move on to the other "simple" converters than can be built using 1 inductor, 1 switch, 1 diode, and 1 capacitor (and some stuff that turns the switch on and off "appropriately"): the buck converter, the boost converter, and the inverting converter. Then they throw every equation even remotely related to SMPS at you, making them sound far more complicated than they really are.

In practice, it seems that many switchers are "flyback" converters -- although some people say that Cuk converters are better than flyback converters; some even go so far as to say that Cuk converters are the "optimum topology".

A maximum power point tracker has many similarities to a switching voltage regulator.

dealing with over-current

A chain of components inside the switching regulator transfers power from its input to its output. One component along that chain is the weakest link -- there's only a limited amount of current it can handle before it permanently fails. (Usually that weakest link is the power transistor.)

If you gradually add more and more stuff to the output of the regulator, or if you suddenly short the output wires together, that is going to demand more power than that weak link can handle.

There are at least 4 ways of dealing with this fact:

  • ignore it, and be very, very careful that you never overload the output. (This is usually fine if the regulator and the device it powers are hard-wired to each other in a way that makes it difficult to add more stuff or short the outputs together).
  • design the regulator to measure the output current with some current sense method, and when it gets close to the current setpoint, the regulator automatically reduces the amount of power transferred to the output. When the output looks like a short circuit, allow exactly the setpoint current to flow. (current limit)
  • design the regulator to measures the output current, and if it ever goes even a tiny amount over the current setpoint, the regulator assumes something has gone horribly wrong, and automatically turns off all output power for a second or so.
  • foldback ... add description here
  • Use a "intelligent switch"[1][2] that automatically turns itself off if it gets too close to failure.

further reading