📜 ⬆️ ⬇️

LEDs, tapes and their power supply from alternating current

Probably I’m not mistaken if I say that more than 90% of people in Russia who know what LED tapes are, to the question “Can the transformers from halogenok be used to power LED tapes?” Will say “no, not!”. The most common explanation will be the banal "electronic transformer is an alternating current, and LEDs need a constant one." This is exactly what we are told in stores, and the overwhelming majority of “professional” articles on this topic have such a leitmotif, which, in general, has taught people to spend significantly more money.

Is it always justified and how do LEDs actually behave in the most common LED tape when powered by alternating current, we will try to find out in the process of reading this article.

Immediately, I’ll make a reservation that I will continue to use the self-evident and quite natural abbreviation of SD to designate “LED” and will not intentionally use the English technical abbreviation LED (Light Emitting Diode) for this concept. In our present country, the absence of any due technical training of managers and vendors in stores has already led to littering and the emergence of such unnatural technical languages, holy fools for the ears, and terrible spelling “ice”, “led'y”, “ice”, or as I recently saw a creeping line - "LED LEDs." Not only is “oil - oil”, I just do not want to repeat and produce this “slovo word” ...
')
The ideological source for writing the study was a long-standing desire to refute the unfounded and categorical assertions about the inadmissibility of the power supply of alternating current diameters. In general, the controversy of this statement certainly catches the eye of any specialist (as well as a “non-specialist”), who understands that the LED, although it emits light, is first and foremost DIODE. This means that it will still emit under the influence of alternating voltage, but only in its half-period.

In fact, we will need to consistently answer three questions :


1) Will the EC "start" when connecting the load in the form of semiconductor diodes;
2) If the ET starts up, the pulsed “variable” electrical effect will not exceed the permissible parameters of the individual LEDs in the tapes. If it does, how long will the SD last under such conditions;
3) What is the economic efficiency of the use of ET in the design of lighting on LED strips.

So, six months ago, I just turned up a convenient experiment.


I needed to light the space in the drawers and pedestals of the tables in my workshop. After the kitchen equipment, I had 1.2 meters of single-color SD tape with a total power of about 17 W (Aztech 14 W / meter) and one electronic transformer from “halogen” - EAC 12V 20-60W, the most common and cheapest, bought for 74 rubles in July 2014. To begin with, in order to start the ET, I loaded it with an ordinary 20 W halogen lamp and then connected in parallel all 1.2 meters of the tape (Fig. 1). As expected, the tape lit. In this case, the luminescence of the ribbon was uniform, of medium brightness, without any flickering noticeable to the eye, which is not surprising, since The output meander ET is modulated in amplitude to a low frequency of 100 Hz. In the course of the experiment, turning off the lamp in such a scheme immediately led to the cessation of the glow of the SD tape, which indicated the impossibility of starting the ET on one half-wave of voltage. Then I broke the tape into two sections and turned them on counter-parallel (Fig.2), which was intended to ensure the operation of the output stage of the ET on both half-periods. At the same time, in order to eliminate the bias of the currents of the opposite direction and overheating of the output winding of the ET from the appearance of a constant component, I ensured equality (8 W) of the number of LEDs in both arms of the load. Immediately after connecting according to this scheme (Fig. 2), the transformer safely entered the generation mode, and both LED strips lit up evenly and were left for 1 hour, during which neither they nor the EB itself heated up at all, which indicated rather normal electrical modes than not.

image

So, the answer to the first question is whether the ET will start when replacing halogen lamps with an LED - positive. Yes, it will start! If you provide a counter-parallel tape as shown in Figure 2.

And running ahead ...
Looking ahead, I will say that, as shown by a further experiment, an ET with a passport minimum starting power of 20 W, was safely launched even at 10 W of the total LED load (5 W in each arm).

Go ahead. Now we try to find the answer to the second question of our research. But now we have only a few experiences, we need knowledge from TERCIE (Theory of electro-radio chains and elements), which ultimately will allow us to assume: is it possible to feed the SD tape in this mode for a long time without serious damage to their durability, if we talk about the damage?

Let's start with the SD tape device. The tape consists of connected in parallel working areas (Fig. 3) of three emitters (indicated in the diagram - E) representing three separate LEDs under a common phosphor layer. Each diode (in the diagram - D) of the radiator is serially connected in triads with diodes from other radiators and a resistor, which sets the calculated operating point of the diodes (See. Fig. 4).

image

The resistor in the triad is selected in such a way that, when powered from 12 V and the calculated operating point of the diode Upr = 3.3 V, Ipr = 14 mA, an excess voltage of about 2 Volts is extinguished on it.

By the way, interesting ...
This arrangement of the triad is reliable and practical, because in case of failure of a single SD in the triad, none of the emitters completely shut off, but continue to burn, albeit with a third less brightness. You can of course create a triad based on a single emitter (and such tapes are found on sale). In them, a fragment with a single emitter and a resistor will determine its cutting area, but in this case, the failure of a single LED in the triad will lead to a loss of luminescence by the whole emitter, which will be immediately noticeable in any luminaire.

Having rummaged around the manufacturers of SMD LEDs, it is easy to find the electrical parameters of the used LEDs:

image

For completeness of the study, I additionally removed the volt-ampere characteristic (VAC) of the working section of the tape (Fig.5), and by simple recalculation I obtained the IV characteristic for an individual DM (Fig.6).

I hope you...
I hope you have no doubt that this could be done physically, and the results would coincide.

image
Pic.5

image
Pic.6

The figures in the IVC do not require additional explanations. I will only add that at a voltage of less than 2.35 V on a separate LED, its luminescence is completely absent, which corresponds to the supply voltage of the working section of about 7 V., and the supply voltage of 15.5 Volts on the tape is completely safe, since the current through a separate LED does not exceed the normal operating 30 mA.

However, all these numerical expressions of working parameters are relevant only for direct current. We are going to experience a diode when exposed to alternating voltage, i.e. pulse voltage in different directions. However, with this power, the maximum permissible values ​​of currents and voltages on the diode can be many times, or even ten times more than the limits for direct current (this is well-known and doubting managers can read lectures on TERC) - it all depends on the duration and frequency of exposure. But here's the problem: the output voltage of the ET has a rather complicated form, which does not allow it to be mathematically reliably described within this article, and the performance characteristics of the LEDs are not provided with a section of absolute values ​​for pulsed modes of operation. Although there, however, there is one parameter (Ipr imp), but for which pulse duration it is relevant - it is not clear for which duty cycle this is applicable, it is also possible to guess.

The thing is ....
The thing is that the pn junction of a semiconductor when operating from an alternating (pulsed) current operates with a variable load. The current periods causing heating and LED operation due to the emission of light waves are replaced by rest pauses (at which the current does not flow through the junction) and in which the semiconductor cools. And the question here is not so much in the absolute value of the current through the semiconductor, but rather in whether the semiconductor will have time to cool so much in the period of the current-free pause that would compensate for the heating that occurred in the current period. Those. prevent thermal breakdown.
Here, I want to remind the "physics" of a semiconductor failure. This will allow us to understand the essence of the processes. She, the physicist, is generally known, but still in her own words: the durability of any device is determined by its fault tolerance. Failures of diodes during normal operation occur in the case of thermal or electrical breakdown.

Electrical breakdown usually occurs when the allowable reverse voltage (Uobr) is exceeded. In this case, the diode loses the property of single-sided conductance and begins to conduct in both directions. In most cases, the electrical breakdown is reversible and the device’s working capacity is restored.

But thermal breakdown, on the contrary, is irreversible and occurs when the excess current is direct (less often the opposite direction that has arisen already after electrical breakdown) and leads to destructive changes in the semiconductor crystal as a result of strong local overheating of the pn junction, unable to pass through a large number of charged particles .

The essence here is such that the conditions for the occurrence of thermal breakdown have not yet been created - the semiconductor is working. I repeat that in general it does not matter what absolute value the current through it flows. It can be very big! The main thing is that our diode does not have time to overheat. In the passport for any diode, two maximum permissible parameters are indicated: Maximum forward current Ipr mzx and Maximum reverse voltage U arr max for long-term exposure to direct current, which under standard operating conditions will not lead to either electrical or thermal breakdown.


Therefore, to study the degree of the impact of the alternating voltage ET on the LEDs, we will push off the postulate that any long-term pulsed current can lead to such a direct current value where the work done by the LED under the influence of the pulsed current will be identical to that at constant current.

How do we evaluate the work done by the LED? Yes, very simple. The LED under the action of current flowing through it performs work on the release of light energy and heat. And we just very easily can measure and compare these two parameters for both types of current, and therefore determine how much the LED's output voltage loads the LED compared to the standard 12 V stabilizer.

To estimate the light energy emitted by a separate working area of ​​the SD tape, I removed the dependence of the illumination on the supply voltage. Illumination was measured at a distance of 10 cm from the emitters (Figure 7).

image
Fig.7

Thus, at this stage, we are ready to receive an answer to the second and third questions of our research.
Let's get started

To begin, examine the output voltage of our ET:

image
Fig.8

At once I will say that it is impossible to use a household electronic tester-ampervoltmeter for measuring the voltage amplitude of this form. It is designed to measure strictly harmonic oscillations, and in our case it will lie very strongly, because we are dealing with an alternating pulsed voltage modulated in amplitude by a current of twice the industrial frequency. The modulation frequency is 100 Hz, the filling frequency: 10 kHz - a bidirectional square wave, the amplitude of the signal is U = 18 Volt. Separate emissions amplitude of more than 18 V oscilloscope is not recorded. Since the filling is a square wave, the effective value of the voltage will fully obey the law of the modulating signal, and therefore, in our case, Udeist = Ua / √2 = 18 / 1.41 = 12.7V. That is why in the passport for ET it is indicated that the output voltage is ~ 12V.

Looking at the diagrams and comparing them with the performance characteristics and current-voltage characteristics, it becomes clear that under the action of a direct current on the SD, we are hardly going beyond the limits of acceptable parameters. The stated limiting direct pulse current for a single LED of 60 mA is achievable only at Upr> 3.9 V, i.e. when the supply voltage on the tape is more than 20 V (see current-voltage characteristics), but such values ​​we, as we see, still do not reach. On the other hand, it is easy to see that the duration of exposure to voltage above the mentioned and completely safe 15.5 V (at which the current through the LED is not more than 30 mA) is no more than 8% of the total power supply time from the considered ET. I think it was dangerous for diabetes. OK. Remember. Check out later.
Now let us estimate whether we will go beyond the limits of permissible reverse voltage and when exposed to the reverse half-period of voltage. In this case, the resistance R in the triad can be neglected, UA (18V) will be evenly distributed along the LED in the triad, and the amplitude value of the voltage on the diode will be 6 V, which is more than the stated 5V. But, the duration of the excess again does not exceed 8% of the total operating time of the LED, and the second thing that confused me a lot is that the allowable reverse voltage in all datasheets is very suspiciously the same for different LED series. It is always 5V. OK. Let us remember this and begin to sum up the first results.

So, theoretically, in the forward half-period, we should not exceed the forward currents for diabetes, and in the inverse half-period, the excess of the stated allowable reverse voltage is small, both in terms of the duration of exposure and in absolute value.

Well, now it's time to test our findings in practice. Let's practically evaluate the light and heat returns. If the light and heat emitted by the tape do not exceed those that are emitted when powered from a standard power source for SD tapes, then our positive theoretical conclusion will be confirmed.

Having fed the tape from the EB, we measure in parallel the light output of a single working section of the tape of three radiators and compare the values ​​with the characteristic in Fig. 7. The light meter records the values ​​at the level of 970-990 lux, which corresponds to the power supply of the tape from a voltage source just below 10 V !!! The heating of the tape turned out to be insignificant and after 1 hour of operation did not exceed 35 degrees Celsius, at an ambient temperature of 25 ° C. In similar conditions, but with DC power supply Upr = 12V, the tape heated up to 49 ° C, and the generated illumination was about 2000 Lux. These results clearly indicate that, despite all the marketing exhortations, the semiconductor, when powered by the ET, works in an underloaded mode and it is hardly necessary to expect its imminent death. By the way, looking at Fig. 9, and making measurements of the areas of the figures of light blue and brick colors, it is possible to understand why it is the LEDs that shine as if they are powered by 10V. The fact is that the light blue figure characterizes the conditions under which the LED tape performs useful work (remember that this happens when Upit> 7 Volts). The light brown figure minus the light blue is the conditions under which the SD tape stands idle - does not work! The ratio of their areas is just 10 to 8. Everything converges, however, hehe.

image
Fig.9

And yet, against the background of a positive answer, the second question of our study, the thought of an insignificant, but still exceeding the allowable reverse voltage, gave me no peace. In short, I decided on the hard one: I connected the tape to a DC power source and, gradually increasing the reverse voltage, I waited for the milliammeter to fix the electrical breakdown. By bringing the reverse voltage on a separate LED to almost 20 Volts, I did not achieve a breakdown. At the same time, the reverse current did not exceed 15 μA. Leaving the whole thing for almost a day - I was convinced that nothing had happened to the emitters, and apparently from the short impulse effects of 6V versus 5V and, even more so, nothing should happen in the foreseeable future.

Of course, I must admit ....
Of course, I admit that this is perhaps the most controversial point in my research, but the bottom line is, experience is more valuable than mathematical calculations. After all, experience is a reflection of the essence, and theory is just an attempt to calculate this essence in the brain.

Conclusions and answer to the third question

It is possible and impossible to use ET from halogen caps to power LED strips; it does not seem to affect the durability of the LED strips and light sources. Rather, the opposite will affect, but they will serve longer. Probably. So far it turns out that way. Forget about the parallel connection and the equality of the shoulders.

Now the main question is not that - is it possible? The question is - Is it worth it?
The answer is: if you are going to mount a lighting system with Novi, then you probably shouldn't . So, the low cost of the EB will be blocked by buying more or more LED power, because at 10 V the luminous flux created by the LED tape is two times less than what we have at 12V (see. Fig. 7)

Food from the EC is justified in cases where:


Thank.
Vink01

Source: https://habr.com/ru/post/251695/


All Articles