Methods of combating and ensuring the operation of electronic devices with increased voltage in the network
I think I won’t be much mistaken if I say that a fairly large number of people have experienced the failure of electronic devices as a result of high voltage. This happens due to various reasons, but the result is almost always the same, the power source burns out.
Just as there are different reasons for the occurrence of this, there are different methods of dealing with it, and that’s what we’ll talk about today.
By overvoltage, in this case, I mean a voltage higher than 242-252 volts at the input of the device for a long time.
There are many reasons for the occurrence of overvoltages in the electrical network, of the most common ones – overlapping of wires due to wind or snow, wire breakage, zero burnout, and recently at least one more has been added, the destruction of substations.
Of course, there are also impulse surges, for example, due to lightning discharges, but this is another topic and other protection methods.
What is usually protected. Well, at least you can put a surge protection relay, the method is simple, effective, relatively reliable, although it has certain disadvantages, for example, if the relay fails, all devices connected to it will remain without protection.
As another solution, protect devices locally, and some home appliance manufacturers offer this. I will not say whether it works badly or well, the very fact that it is.
And that’s exactly what I would like to tell and show about such a protection option today with a real example.
In general, there is nothing innovative in such a method of protection, why is it not built in everywhere? Yes, then it is at least expensive in the short term and unprofitable in the long run. Actually, the first is limited by the buyers themselves, such devices are corny more expensive, and the second is limited by manufacturers, since it is easier to refuse a guarantee than to make a reliable device.
What is usually done for protection. As a maximum, they will put a varistor, which in some situations can really protect due to self-breakdown and, as a result, burning the fuse. Alas, this protection option does not always work, not all devices have it, and besides, sometimes the varistor rating is so high that it can only help with high-voltage impulse noise.
Most often, the varistor has a rating of 470 volts and in some situations it really helps, but protection is one-time.
Another solution, the function OVP in power supplies, again, it is far from always implemented, even more often it is not there than it is. As an example, a power supply circuit based on a chip from Power Integrations and a protection circuit UVP/OVP. The manufacturer himself states that
When the MOSFET is off, the rectified DC high voltage surge capability is increased to the voltage rating of the MOSFET (700 V), due to the absence of the reflected voltage and leakage spikes on the drain.
In a loose translation – when the input voltage is higher than a certain one, then we block the operation of the high-voltage transistor, and since it is 700 volts, and there are no voltage surges due to operation, the power supply can withstand high voltage.
The third solution is the case when the power supply is simply initially designed to work at this voltage. In fact, everything is simple, in the network we can get a maximum of 380/400 volts (excluding extraordinary cases), which means that by making a power supply unit for an input voltage of 400 volts, we don’t care what it has at the input.
This option is convenient in its own way, but it comes out more expensive, and besides, at a voltage of 220/230 volts, and even more so at 180-190, it works in a less optimal mode, which reduces its efficiency, and the efficiency also decreases due to higher voltage power transistors.
Below in the photo are examples of power supplies for which the operating range is declared 100-520 and 80-580 VAC.
True, there is a separate category with automatic switching of 115/230 volts, but such power supplies are not widely used, being limited only to a mechanical switch.
There are quite a few reviews of voltage relays, high-voltage power supplies are rarely used, so I would like to dwell on the protection option at the power supply level ..
As an example of the implementation of surge protection at the consumer level, we can consider the drivers of the LRC-60 LED lamp, especially since lighting is exactly the area where the failure of the power supply is more noticeable. And since such drivers are also used in street lighting, group protection can be inconvenient, because if it fails, we lose the entire lighting branch, and not just one lamp.
I had two drivers at once, although in fact it is the same driver, but one has IP66 performance, the other has IP20.
The operating voltage range refers to the range in which the driver supplies the load, but it is indicated that the driver is able to withstand an input voltage of up to 380 volts for a relatively long time.
Since the degree of protection is different, then, accordingly, one driver is filled with compound, the second is “naked”.
As an example, I will consider the stuffing of a driver with IP20
No, of course, I’m a simple person and, in principle, I could make out a flooded driver, as I already did in one of the reviews, but in my opinion, having a driver without filling in my hands is already overkill.
By the way, the photo shows the previous model, it also withstood 380 at the input, but an important difference of the new one is the presence of an active corrector.
Something brought me, let’s get back to the topic of the article.
What elements determine whether the power supply can withstand high input voltage and at the same time look at the example of this driver.
The varistor at the input of the network, naturally after the fuse (although I somehow met before …), 680 volts, respectively, conventionally for 475 AC, which means that it is here to protect against impulse noise. By the way, I remind you that here the DC voltage is indicated on the varistor, for AC it is equivalent to the amplitude, but for example, for Epcos varistors, they indicate the current voltage.
The X-capacitor had to be soldered to see the value, or rather, the voltage. It costs 400 volts here and this is quite important, because usually they put 250-300 volts in power supplies. There are two such capacitors, before and after the common mode choke.
Another varistor, 470 volts, judging by the cunning switching circuit, it is here to dampen single pulses, but at the same time it does not affect the continuously operating voltage.
Interwinding Y-capacitor, as well as X-capacitors, are different, or more precisely, two types, Y1 and Y2, the first is more resistant to high-voltage pulses, but in this case, the capacitor is not just Y1, but also with voltage up to 400 volts.
1, 2. In fact, in some situations, safety is also increased by series connection, especially for Y-capacitors, since a person’s life may depend on them. The photo shows a pair of Y-capacitors for 250 volts connected in series.
3. They also do with X-capacitors, although much less often.
4. In some situations, to increase the safety of the SMPS, two fuses are also installed at the input, respectively, for zero and phase, although sometimes two fuses are installed in series, for example, before and after the varistor, and with different ratings, but this is rather an exception.
In any case, safety is never superfluous, sometimes people do not think that the phase and zero at the input of the SMPS are marked for a reason, namely for safety, since the fuse is placed precisely on the phase wire.
The control board, for the sake of compactness, is in the form of a submodule, but in general the point here is what a high-voltage transistor costs, because it also determines the resistance of the SMPS to high input voltage. WML08N80M3, 800V, 7A, very good, and here someone may ask, why don’t they put such high-voltage transistors in all SMPS? The answer is extremely simple, the higher the voltage transistor, the usually it has a higher open channel resistance, respectively, a higher drop, heating and efficiency reduction. More often, with such circuitry, transistors of 500-650 volts are used.
And of course, a few tests and a visual demonstration, for which, in addition to the driver, I also needed a load, which will be two LED panels.
Each panel consists of 54 LEDs connected according to the 6P9S scheme, i.e. the total voltage is about 26-27 volts, the driver is up to 60 volts, because there are two panels.
On two panels, the driver produces 52.8 volts, this is determined by the panels themselves, the current is about 900mA, but this already determines the driver, the actual current and voltage are indicated in the name 60-900.
The panels were chosen for a reason, I wanted to get a load close to the maximum 54W, I got about 47W. Of course, it was possible to use an electronic load operating in CV mode, but this is much less obvious.
Since the driver and the load are on the table, I immediately estimated the efficiency, and at the same time checked the statement about the presence of a power corrector.
Well, it’s hard to say about the efficiency, I got something about 90%, 92 is declared, but about the power factor, no questions asked, something in the range of 0.97-1.
And of course, a test of resistance to high voltage, the manufacturer writes about 380-400 volts, but I did not waste time on trifles and increased the voltage by almost 10% (although when I took it off, I increased it even higher), and besides, I checked the resistance to a short circuit at the output .
00:00-00:45 – Smooth change of input voltage.
00:45-1:00 – Abrupt change in input voltage.
1:00-1:35 – Checking for a short circuit on the output, first short, then long.
Along the way, I checked the range of ripples in the output, up to 2% is declared, in reality it turned out a little less. One cell on the oscilloscope screen is about 5% (measured on a 1 ohm resistor), the main ripple (excluding RF noise) fits well into 2%.
In the process, it turned out that after a short short circuit, the driver restores the output immediately, after a long one about 10 seconds after the short circuit is eliminated.
If you don’t want to watch the video, then the main meaning can be conveyed in one photo: the driver turns off at about 303-306 volts, it normally tolerates higher than 430 and turns on when it drops to 290-295 volts.
Three options for solving the problem of overvoltage in the network are considered above:
Power off at group level with voltage relay
Local protection in the device itself does not fail, but is disabled for the time of overvoltage.
The voltage tolerance of the device is up to 380/400 volts, the device continues to work.
Personally, I like the third option more, I use the first one at home, but the second one is no less viable, it all depends on the use case. If we talk about protecting the drivers for lighting, then I would also prefer the second option, in which case the lighting will continue to work during an emergency, sometimes this is very important. I would like to know what you think about this.
