Switching Power Supplies
08/19/2001
A special Note before we start our Technical Discussion:
Regardless of what some folks may tell you, NEVER EVER turn on a Switching Power Supply without some kind of a load. Granted, there are some that you can safely turn on without a load, but when testing Switching Power Supplies, ALWAYS assume that these units need some kind of a load! It is simply much safer that way.
I had preached this in the classroom for years, and one year one of my students wanted to prove me wrong with a Switching Power Supply he had turned on & off a few times. He brought it over, set it up, and while about half of the class was watching and with a big smile on his face he turned it on with no load. Unfortunately, his smile disappeared when smoke and fire exuded from the unit (while those in the class watched). I then said, "I guess you need to repair the damage. You're married to it until it's fixed". He was not a Happy Camper.
Introduction
For the material presented here, it would be helpful to print out the Switching Power Supply Schematic ( 76 Kb pdf file).
Switching Power Supplies can be divided into 3 Physical Parts:
You must also think about:
The Primary Section
Note: The highlighted areas here are not Links, but are references to your printed Switching Power Supply Schematic
- An important initial consideration: The Primary Section is
actually a basic power supply itself, complete with a basic rectifier
section and basic filter (D1/C3),
that supplies DC Voltage for the Switching Circuitry (T2/Q1).
- This DC Voltage is simply applied to one terminal of the Primary Winding of the Switching Transformer T2, with the other terminal of that Primary Winding attached to the Collector of the High-Power High-Voltage Switching Transistor Q1.
- This transistor serves as the Switching Element to Pulse the DC to create "Pulsating DC", which will behave as AC in the Secondary Winding, due to the changing magnetic field.
- By this action, where DC in the normal sense will not work on a transformer - this switching action basically converts the DC into pulsating DC, causing changing magnetic fields, which in turn induces a voltage in the Secondary of T2.
- The frequency that the Switching Transistor is pulsed (by
the Oscillator/Pulser) is an extremely
important aspect of the Induced Voltages in the secondary.
- Think of it this way - one pulse in the primary = 2 induced voltages in the secondary. One induced voltage in the secondary on the rise of the pulse in the primary (when the magnetic field increases), and one induced pulse in the secondary when the pulse shuts off (when the magnetic field collapses).
- Now, obviously, something like one pulse per second would not generate much voltage in the secondary, and equally obvious is the fact that the more often the Switching Transistor is pulsed, the more often voltage will be induced in the secondary.
- It should now become apparent that the frequency of pulses to the Switching Transistor will have a significant impact on the output voltage available in that secondary.
The Secondary Section
The highlighted areas are references to the Switching Power Supply Schematic.
- The most important aspect here, is that there are rectifiers
and filters (D2/C4) in the Secondary
Section to do the normal conversion of AC pulses into DC pulses,
and filter them to create smooth DC.
- With no load on the DC Output, the DC Voltage will climb to the PEAK Value.
- With a load on the DC Output, the DV Output Voltage will fall.
- Effectively, the action of the Switching on the Input will have a "Pumping Action" on the Secondary, where greater loading on the DC Output will require more "Pumping Action" to keep the DC Output Voltage up.
- Obviously, there needs to be some method of coordination
between the Load Demands and the frequency of those "Switching
Pulses", which "Pump up the DC Output Voltage".
- Light loads on the output would require a lower frequency of pulses, and ..
- Heavy loads on the output would require a higher frequency of pulses.
The Control Circuit(s)
The highlighted areas are references to the Switching Power Supply Schematic.
- One method used is to use the level of the DC Output Voltage
to Inversely Control the Frequency of the Pulses
fed to the Power Switching Transistor, which is in the Primary
Winding of the Switching Transformer. I.e.:
- When the DC Output Voltage is low, the Pulsing Frequency needs to be increased.
- When the DC Output Voltage is high, the Pulsing Frequency needs to be decreased.
- This may be accomplished by using a Zener Diode (Z1)
in conjunction with an Opto-Coupler (U1).
- The Opto-Coupler provides very important Isolation of the Output Circuit to the Input Circuit.
- Assuming that the Opto-Coupler is actually an LED, Light-Coupled to a Light-Sensitive Transistor (Photo-Transistor), an increase in voltage above the set limit of the Zener Diode will cause the LED to begin conduction. This optically coupled Photo-Transistor will begin conduction (and thereby lower it's resistance).
- The now conducting Photo-Transistor will affect the frequency of pulses created in the in the Primary Circuit (via the Oscillator/Pulser) as to make them decrease their activity.
- If the load in the DC Output Circuit should cause the DC Output Voltage to fall, then the Zener Diode will either lower the conduction of the LED, or drop out altogether.
- In this instance, where the LED conducts less, then the Photo-Transistor will also conduct less, which will allow the frequency of pulses delivered to the Switching Transistor to increase, and thereby cause the DC Output Voltage to be "Pumped Up" more.
- Another method of control is using what some call the "Inductive Kickback Method"
Other Considerations
- Those of you that have used an Electronic Flash Unit may well have noticed that there is an initial "whine" when the unit is first turned on, and then the "whine" will increase in frequency as it nears it's charge (lighter load) and becomes ready for use.
- Switching Power Supplies, because of their very strong pulses,
do generate a considerable amount of RF Noise, commonly called
"Radio Frequency Interference" (RFI), and/or "Television
Frequency Interference" (TVI).
- RFI or TVI is obviously a pretty serious problem towards not only common Hi-Fi or Television Sets, but is a potentially serious problem in Radio Communication Systems, be it CB, HAM Radio, Police and/or Emergency Communications.
- RFI will show up as a very broad banded noise that sounds like "Hash" in Communications Gear, and will cause lots of "Snow" on your TV set.
- Because of this potential noise interference problem, the FCC requires that certain steps be taken to generally protect the consumer. However, in spite of possibly meeting these "stringent" requirements, the Communications Industry will sometimes find that additional steps are necessary to further reduce this very bothersome interference. More on this later.
- You will find certain "Hash Filters" in the design
of these units, located immediately on the line side input.
- You will discover what appears to be a small transformer (T1) in the line side input, that seems to be rather strange, where one side of the line input goes through one winding, and the other line input goes through the other winding. This device is one portion of the "Hash Filter".
- You will also discover what appears to be a fairly conventional "Balanced Pi-Filter" (C1/L1/C2) just beyond that unusual transformer. This is another portion of the "Hash Filter".
- You may also find a "Toroidal Unit" (an Inductor), that the power line goes through, which tends to prevent "Spikes" from passing from the "Switching Section" in the Primary to the AC Line.
- With the AC Adapter that I use with my Laptop Computer, in conjunction with some Communications Gear, I found that there was still too much of this bothersome "Hash" that showed up in the signals I was receiving. I found it necessary to form an outside "Toroid Filter" and pass the AC Line through it, with about 3 turns as loops through the Toroid. My noise problem virtually disappeared.
- Another rather interesting circuit
found in Switching Power Supplies is an "Overvoltage Protection
Circuit".
- Consider that protecting the Equipment that the Power Supply is driving, is actually more important than the Power Supply itself, and often fuses are not fast enough to protect the equipment. In fact, there have been instances where a design incorporated a series transistor that would blow before a fuse could.
- In the unusual Overvoltage Protection Circuit I mentioned, you may very well find an Zener Diode/SCR combination circuit (Z2/SCR), where if the DC Output Voltage exceeds what is considered safe limits, the Zener Diode will fire the SCR, which presents itself as a DEAD SHORT on the DC Output. While this DEAD SHORT is hanging in there on the DC Output, we hope that either the regular fuse will now blow, or we have some kind of an "Intelligent Circuit" (U2) that will recognize an "Over Current Condition" and will quickly shut things down.
- This circuit to accomplish this, is interestingly enough called a "Crowbar Circuit", as if someone was to place a crowbar across the output.
- The "Over Current Protection
Circuit" may be accomplished with what is known as an "Silicon
Controlled Switch".
- Note that this circuit can be 2 separate transistor components, and it is also available as a 4-Layer Device.
- The way this circuit functions is rather interesting, and also quite simple. It starts with both transistors turned off, and with an incoming voltage at the base of SCS_Q2 to overcome the bias, SCS_Q2 turns on .... which turns on SCS_Q1 ... and then SCS_Q1 turns on SCS_Q2. This means that they simply keep each other turned on. In this manner, this circuit behaves like an electronic switch that has been turned on.
- If a negative signal is applied to the base of SCS_Q2, it
will turn that transistor off, and since the conduction of that
transistor through SCS_R2 is what biases SCS_Q1, then SCS_Q1
will turn off also. Violla! Now both are turned off again.
- There are many potential applications for this novel circuit.
Safety
- The first thing to consider, is that the front end of a Switching Power Supply is usually directly connected to the AC Power Line, with no isolation of any kind. This simply means that where we would think it insane to deliberately grab hold of the HOT side of an AC Power Line, that same individual might very well dive into an open Switching Power Supply, without realizing the immediate danger involved.
- Note that there are two different grounds... they are not the same, and you should never assume that the Gnd #1 is actually Ground... It is actually SUPPOSED TO BE NEUTRAL, but it may not. In any case, treat it as a potentially hot wire.
- I can't tell you how many times various folks have started
to see if they could find out what's wrong with their Switching
Power Supply, and hooked their Oscilloscope Ground to somewhere
in the Primary side of the system. One of these things usually
happens:
- Either there will be a flash of smoke and fire from that ground lead (because their scope ground really was at ground, and the PS AC Plug was backwards),
- Some idiot removed the ground 3-wire from the Scope AC Plug (this is a SERIOUS mistake), which means that now the chassis of their Oscilloscope may be at AC Line Potential! (do you like to be "plugged in" to do bodily harm to yourself, while putting on a show for your friends?),
- The moment you go to attach another piece of test equipment........
- Even with an Isolation Transformer connected as an input, you still need to remind yourself that between those two AC Line Inputs is a potentially dangerous situation, and due care and caution must be taken.
- What the Isolation Transformer does do is to allow you to hook the ground lead of the Oscilloscope to Gnd #1 in the Primary Side, and make observations, .......BUT do not make observations in the Secondary Side of this circuit with your Oscilloscope while you are Grounded to Gnd #1!!!!
- When you want to look at the Secondary Side of this circuit, move the Oscilloscope Ground Lead to Gnd #2.
- If you are taking voltage measurements with an Isolated Meter, then as long as you avoid making any body contact with any portion of the circuit, you can safely take measurements just about anywhere, but again ... don't try to take measurements across the Primary/Secondary boundry, as they need to remain totally isolated from each other.
Troubleshooting Methods and Procedures
- "Isolation" between the AC Power Lines with an Isolation Transformer is a must!
- An Isolation Transformer is simply a 1:1 transformer that effectively disconnects any "direct hot wire" to the AC Power Source. As mentioned above in "Safety", any point in the front end of that circuit may be hot, unless you use an Isolation Transformer. Be sure to follow those Safety Rules described above.
- Note that you should have a high +V at TP-1. If your line voltage is about 115VAC, remember that this is called 115V-RMS, which can be computed to about +165V peak! (Peak = 1.414 x RMS).
- This voltage is available for the Switching Circuit, and you should also find that this high DC Voltage should be available at the Collector of Q1. If not, then you have an open Primary Winding.
- If you do not see small positive pulses on the Base of Q1, then your Pulsing Circuit is not working. No Pulses - No Transistor Conduction. No Conduction Pulses - No Magnetic Changes in the Core of T2. No Magnetic Changes in the Core of T2 - Not Induced Voltages in the Secondary of T2. No Induced Voltages in the Secondary of T2 - No Voltage to Rectify and send to the Filter ... ad infintium ...
- A special consideration is that these Switching Pulses are
at rather high frequencies, in the order of several Kilohertz
or more, and these high frequency pulses allow two major
advantages:
- A physically smaller transformer is needed for the necessary Power Requirements.
- These high frequency inducted voltages in the Secondary are much easier to filter.