Linear Power Supplies and Regulators
At a time when so many people write about SMPS I felt like doing differently…
Once again I will be the underdog’s advocate by writing about some technology which tends to be replaced by better and brighter things. Are linear power supplies and regulators going to disappear? Not so soon and probably never, let’s see why.
Linear Power Supplies have been around for many years. They have certain advantages, as we will see in this article, but one big disadvantage. They are inefficient, in many cases less than 50% efficiency; because of this they tend to be bulky and heavy.
SMPS are more efficient (some have more than 95 % efficiency), consequently small and light. You could not imagine using a linear charger for your mobile phone that would weight half a kilo and have the size of a brick; therefore the tiny thing, smaller than a pack of cigarettes, is much more like what you want. SMPS circuits are more complex than their linear counterpart but, manufactured in larger quantities, they became really cheap.
SMPS, however, have two disadvantages that Linear Power Supplies do not have:
1) They are not always safe (see my earlier article on “Double Insulated SMPS”).
2) They tend to generate Electromagnetic Interferences (EMI)
I experienced the first problem when I was using a 12 VDC SMPS adapter for a microcontroller project. The microcontroller development board includes a programmer that connects to a PC through one of the USB port. In the semi darkness of my office I noticed a small spark when I connected the USB plug to my laptop. Investigating further let me wrote the article mentioned above, warning about the stray voltages in unearthed SMPS! Indeed, while unearthed SMPS are OK for charging your mobile phone they are not 100 % safe. They are banned from hospitals and dairy farms. It is also recommended that you do not use your mobile phone while it is connected to the charger…
I did not want to damage my laptop! So I needed a clean and safe 12 V power supply that I can use for the microcontroller experiments but also for some RF projects I was working on.
The first contender for this application was a small, light and compact SMPS that can deliver up to 3 A; this is much more current than what I needed but this additional current would be a plus. I remembered having tested this PS for stray voltages. The results were not good when the earth wire was not connected but, with the earth connected, my laptop would be safe. So I could live with that. The label indicates that it is CE certified so we can assume that it complies with the European EMC requirements. Electromagnetic compatibility would mean that it cannot generate more than a certain amount of interference. It also means that it is not too sensitive to interferences…
So let’s put it on the testing bench! No too harsh, I loaded it with a 47 Ohm resistor, so it will only deliver a current of 250 mA. As it is connected to the Earth there is no problem of stray voltages. But look at the output with and oscilloscope:
This noise, superposed on the DC 12 V, could also be heard in a shortwave receiver nearby (one of my projects); it was loud and all over the frequency ranges, therefore not acceptable for my RF projects…
It provided me with a good excuse for building a small 12 V power supply the old fashion, using a transformer, rectifier, couple of capacitors and a linear regulator.
Here it is, cheap and easy. It delivers a clean, stable and perfectly safe 12 V, almost as good as a battery. And it looks cute, doesn’t it?
The second problem with SMPS manifested itself by a high pitch noise coming from the speaker that was connected to my Laptop. The speaker has its own amplifier which needs to be powered through a USB port. As I needed to spare the laptop USB ports for other purposes I was using an external USB power supply for the speaker. The noise was caused by Electromagnetic Interferences generated by the USB power adapter. EMI are regulated internationally and SMPS must comply. However, sometimes the minimum allowed is still too much for some applications. Again, the easy solution was building a small linear power supply that I can also use as a USB charger and to power the USB oscilloscope.
Like the previous one, by using a small transformer, a rectifier, a couple of capacitors and a linear regulator. In this case a 7805.
About linear regulators efficiency
A linear power supply can be built using only a few cheap and easily available components as shown below. The most expensive component would probably be the transformer.
Linear voltage regulators have been around for so long that there are many different types available. There are the fixed types such as the 78xx series (xx indicates the output voltage) and the adjustable types such as LM317 or LT1083 where we can set the desired output voltage with two external resistors. Now why are they so inefficient?
A series linear regulator can be seen as a variable resistor controlled by a feedback circuitry. It forms a voltage divider with the load. If the load varies, the feedback circuit will act on the “variable resistor” to keep the output voltage constant.
If we disregard the current taken by the feedback circuit (only a few mA) we can say that the input current is the same as the output current. The efficiency is the ratio between the output power and the input power expressed in %. We can easily calculate the efficiency with the relations below:
Note that the efficiency doesn’t depend on the current, but only on the input and output voltages!
For example if you need an output of 5 V out of a 12 V input the efficiency will be 5/12 = 0.417, i.e 41.7 %. For a 9 V input with a 5 V output the efficiency will be 5/9 = 0.555, i.e. 55.5 %. On another hand, if you need an output of only 3.3 V out of an input of 12 V the efficiency will only be 27.5 %.
The less the difference between VIN and VOUT is, the better the efficiency will be.
You may wonder where the limit is: could we, for example, use a 7805 to build a 5 V power supply out of a 6V input, hence having 5/6 = 83.3 % efficiency? The answer, unfortunately is no! The 7805 needs at least a 7.5 V input to work. On another hand we need to provide for the input voltage fluctuations. In this case a 9 V input would be safe and we would have to settle with only 55.5 % efficiency. For low power applications this might not be an issue.
Linear regulators might still be around for many years, mainly for low power applications, when efficiency and size are not an issue. In the examples quoted above we could have found SMPS solutions, with a better SMPS and proper filtering and shielding to eliminate the EMI problems. But why making it complicated when simple solutions exist? Simplicity is one of the strength of linear power supplies and regulators.
For more information about Gerald Musy and his other superb articles please click on the link below:
Please give a support by clicking on the social buttons below. Your feedback on the post is welcome. Please leave it in the comments. By the way if you have any good repair article that you want me to publish in this blog please do contact me HERE.