|
Arabic
Bengali
Chinese
English
French
German
Hebrew
Hindi
Italian
Japanese
Korean
Malay
Polish
Portuguese
Spanish
Turkish
Ukrainian
Vietnamese|
ENCYCLOPEDIA OF RADIO ELECTRONICS AND ELECTRICAL ENGINEERING About simple and powerful voltage stabilizers. Encyclopedia of radio electronics and electrical engineering
Encyclopedia of radio electronics and electrical engineering / Surge Protectors Self-manufacturing of powerful (and most importantly, simple circuitry!) Voltage stabilizers (SN) and power supplies (PSU) is very important. Factory powerful PSUs (SN) can be difficult to acquire, and the prices for these products are very high (from tens to hundreds of dollars, depending on the parameters). Since the manufacturer does not make a PSU for himself, he saves on everything possible. Specialists can make you a powerful PSU under the order. Having got acquainted with the filling, the buyer realizes that he paid 70...90% of the PSU price for the design (box). Modern pulse-type power supplies can be very complex circuitry, so even an experienced specialist can find it difficult to restore the power supply to work (and it happens that repairs are impossible). The reliability factor that an amateur can afford in terms of "overspending" of materials (copper iron, etc.) and components speaks about the expediency of manufacturing a powerful PSU. Here, the manufacturer is not a competitor to us, and we will not worry about overheating of any component or assembly. If you need a powerful PSU that can replace a car battery in many situations, then it is often more profitable and easier to use continuous SN. The fact is that the fleet of powerful radio-electronic means (RES) is constantly growing and being updated. Thus, automotive RESs are very diverse and very "gluttonous" in terms of energy consumption (audio complexes, transceivers, security systems, converters). For just one check, not to mention the repair of the RES, it is required to have a very powerful PSU (SN) capable of operating with load currents of 20 ... 30 A or more. By the way, amateurs who repeated BP [1, Fig. 7] were satisfied with his work. About transistors. To put into practice the characteristics of the BP [1], you need to use the recommendations outlined in [2]. Fans were especially interested in the issue of replacing powerful transistors of the pnp structure of the KT8102 type with affordable transistors of the npn structure of the types KT802, KT803, KT808, KT819. Unfortunately, transistors KT8101, KT8102 are still inaccessible to our outback. Moreover, it is the defective KT8101, KT8102 that go to the outback, they can be easily identified with a pointer ohmmeter, because. they ring in all directions. Such defective products can be detected even without a meter [3]. We use any 30 V rectifier and a 30 kΩ resistor (Fig. 1).
For a working transistor, the ammeter will not record anything. But even bad transistors with Uke = 5 ... 10 V I did not throw away. They are able to work in low-voltage key circuits and as analogues of powerful zener diodes. Practice shows that only transistors with low leakage currents work for a long time and without fail. By the way, I believe that earlier transistors were made "in good faith". Three KT803A transistors are more reliable than one KT8101. I happened to check many foreign transistors with the device [3], there is no idea about such leakage figures as in our technical specifications. I also made a portable meter Uke.max [4] for testing in the conditions of the radio market, since transistors must be selected according to the parameters (and the acquisition of marriage is unacceptable). For less scarce transistors KT802, 803,808, 819, a power dissipation margin of about 50% is needed, especially when the number of transistors is 5-10 or more. Each transistor must be tested and matched to operate in parallel. A random set of transistors in a battery leads to a chain reaction of failures, as soon as the CH is loaded well in terms of power. Such a measure as an increase in emitter resistances (by 100%), unfortunately, does not apply to random instances with a number of more than 5. Only a preliminary selection of all transistors according to h21E and Uke.us will significantly reduce the values of emitter resistances and thereby reduce the needlessly dissipated on them power. So, in order to select transistors for parallel operation, it is necessary to measure h21E of each transistor at a current Ik = In.max / N, where In.max is the maximum current for the entire CH battery; N is the number of transistors connected in parallel. By the way, h21E for the entire battery of transistors should not exceed 100 (but also be less than 20). Therefore, transistors KT8101 and KT8102, having h21E> 200, are generally unreliable in powerful linear circuits. But that is not all. It is necessary to check the transistors for power dissipation, i.e. turn them on to a load corresponding to 50 ... 70% of the maximum power, and "torment" them for a long time. More than 10 years of practice shows that this procedure is necessary and sufficient for long-term and trouble-free operation of a transistor battery in high-power SN. At the same time, it must be remembered that overheating of the transistor crystal is its "death". Therefore, you need to check for power carefully, knowing the required heat sink area and preferably the temperature. The fact is that as the temperature rises, the maximum power decreases, which is equivalent to a decrease in the potential capabilities of the PSU. Up to 20 (!) pieces were installed using the specified method. transistors of types KT803, KT808, KT819, etc. By the way, if each transistor of the battery is installed on its own heat sink, then the correct selection of transistors can be checked by the same heating of the heat sinks. It is very important to choose the correct PSU voltage. Transistors heat up and fail most often at a minimum voltage (approaching the short circuit mode). The check is done as follows: an oscilloscope is connected to the SN output, and the primary winding of the power transformer is connected through the LATR and the voltage at the LATR output is reduced until pulsations appear at the SN output. In this case, the current in the MV load should be maximum. It is necessary to determine the margin for fluctuations in the mains voltage. If a network voltage stabilizer is used, then the task is simplified. The author used the parallel mode of switching on old, but very reliable ferroresonant stabilizers of the CH-315 type to power powerful power supplies. By connecting 2-3 such stabilizers in parallel, we get a power of 600...900 W [6]. The trouble is that a sharp increase in the voltage in the network leads to an increase in the voltage at the output of the rectifier, and consequently, to an increase in the voltage drop across the transistors, which can disable them due to thermal overload. If you reduce the resistance of the resistors in the emitters to 0,1 Ohm, then you can partially equalize the spread of transistor parameters by including resistors with a resistance of up to 10 Ohm in the transistor base circuit. The inclusion of these resistors almost always guarantees the elimination of CH self-excitation. Self-excitation is a real scourge for most CH circuits. At the same time, the transistors in the CH burn out instantly, and at the power in the load is much less than the nominal one. Powerful transistors (heat sources) must be spaced around the radiator away from each other. The case itself fits very well. The disadvantage in this case is the long connecting wires between the CH circuit and powerful transistors. Therefore, near the output of the base of each powerful transistor, a choke (20 ... 100 μH) is turned on. Using segments of ferrite rods from the circuits of RF equipment, you can independently manufacture such chokes by winding the wire D0,5 ... 0,6 mm in one layer and then pouring it with epoxy glue. The PSU case for 30 A was made of two U-shaped aluminum plates 2 ... 3 mm thick. 4 (8) transistors were placed on the lower part of the case, and 6 (12) on the upper part. The number of transistors for a more powerful version of 50 A is indicated in brackets. A big plus of the circuit [1, Fig. 7] is that all transistor cases are connected to a common wire of the CH circuit. Therefore, there are no big difficulties in terms of fastening and mounting 10-20 transistors. The situation is even simpler with plastic KT819. They cost literally a penny, but there are defective batches (they can’t withstand even 30 watts in power). Many amateurs are chasing metal KT819AM-GM, considering them better than plastic ones. But, according to reference data, for plastic KT819A-G, the maximum power decreases with temperature by 0,6 W / ° C, i.e. every 10° "eat" 6 W, and for metal this coefficient is 1 W / ° C, i.e. at 10 ° "eat" 10 W! This is where the "old" transistors like 2T803 are beneficial, which keep their 60 W up to 50 ° C. But what about KT8101 and KT8102? The reference literature is silent on thermal factors, and the guaranteed maximum power is only valid for temperatures below 25°C. But the radiator will warm up several tens of degrees higher! So, it is easiest and cheapest to install transistors of the KT819B-G type in a powerful CH at the rate of one transistor for every 2 ... 2,5 A of the output current (KT803 transistors - for one 3 A transistor). Since it is difficult to bend the sheet material of the housing, the housing is made of six parts. Since the lower part warms up more, fewer transistors are installed on it than on the top. CHs made using this method of selecting transistors had to be repaired very rarely, except perhaps due to the negligence of the owner of the PSU (it is better not to rent powerful PSUs to anyone). In addition, it does not hurt to equip the CH with thermal protection: the heat sink has overheated, and the CH is turned off. One of the time-tested thermal relay circuits is shown in Fig. 3.
Thermistor R3 type MMT-4. It is a temperature sensor, therefore it is fixed on the heat sink of powerful transistors in the place where the temperature is maximum. It is necessary to take care of the electrical insulation of the R3 thermistor body from the heat sink, because. one of its pins is its corpus. But if the circuit of Fig. 3 is powered by a separate rectifier, then it is not necessary to isolate the R3 case. The KT829 transistor can be replaced by the KT972 or an analogue of the Darlington transistor on the KT315 and KT815 (817) transistors. The circuit is not critical to the type of thermistor, which at 25°C can have a resistance of 1,5 to 4,7 kOhm. As R1, it is better to use a multi-turn resistor (they set the response threshold: the lower its resistance, the higher the shutdown temperature). This scheme can be installed in any PSU. It is important that the supply voltage exceeds 14...15 V (the relay operation voltage is 12 V). The current generator Fig. 3 can be made according to any known scheme. A field-effect transistor current generator is well suited. If increased stability of the response threshold is required, then D2E is used as VD818, R3 is increased to 10 kOhm, R1 and R2 are selected. The operating current of the current generator is set to 11 mA. The temperature of the thermal protection operation is set within 50 ... 80 ° C, not higher. About diodes. High-power diodes, although expensive, are easier to obtain than high-power transistors. For example, D122-40 must be taken as direct polarity (without the X sign) and reverse (with the X sign). This allows two instead of three heat sinks [5]. The "ancient" B50, B200, etc. will also do. You can get by with two diodes and one heat sink (Fig. 4). This circuit is designed for diodes in which the cathodes are connected to the case.
And if it was not possible to get diodes with a working current of more than 30 A? You can get by with 10-amp ones by turning them on according to the scheme in Fig. 5. Just do not need to "squeeze" the maximum current out of the diodes (no more than 7,5 A). Diodes of types D242(A), D214(A), D215(A), D231(A), KD213A were used. Preference is given to diodes with the letter index A, because. they have less heat loss. Our diodes are more reliable than imported ones, for those the maximum current can be safely reduced by 1,5 times, or even more.
Very convenient diode KD213A. They have a cathode - a case, so a dozen of these diodes can be mounted with one bar. No need for insulating gaskets and ingenious flanges used in industrial mounting systems for diodes KD2997, KD2999. The last diodes have a working current greater than KD213 (KD2999 - 20 A, KD2997 - 30 A), therefore, for them, the resistance of the resistors decreases to 0,02 Ohm. In this rectifier, modern diodes with a Schottky barrier work perfectly. It is only necessary to select specimens with the smallest leakages (this can be done even with an ohmmeter, since the leakages are huge compared to silicon diodes). Diodes of the KD2998 type are more profitable to use in a bridge rectifier. For Schottky diodes, equalizing resistors are not needed, they can be safely placed in parallel (Fig. 6).
About resistors. Their number in the scheme of Fig. 5 can scare away. But making them is easy. These are pieces of enameled wire D0,6 mm, 80 ... 100 cm long, wound on any mandrel. Such a resistor can withstand power much more than 5 ... 10 watts. Don't skimp on radiators. At least 100 cm2 of radiator area is required for each diode, since at temperatures above 75 ° C, the maximum average current must be reduced. About filter capacitors. The 2000uPH50V batteries are a good fit, both in terms of price and reliability. Their number is chosen from the ratio of 1000 microfarads for every 1 A of current. If the MV will often be operated at currents above 20 A, then a capacitance margin should be provided, based on the ratio of 2000 μF per 1 A of current. These capacitors are most afraid of temperature and ripple, so they need to be placed in the coldest place in the PSU. And the magnitude of ripples can be reduced only by increasing the capacitance. About the transformer. Various options have been used. Consider the simplest and cheapest TC-270. The magnetic circuit of this network transformer is capable of delivering 500 W or more to the load. The ceiling depends on several factors: the diameter of the primary wire, the quality of the assembly of the core, and, oddly enough, how "planted" the iron. The latter factor is easily detected by measuring the no-load current (Ixx). If Iхх≤0,25 A, then the transformer is normal. If Iхх≤0,35 A, then such a transformer has been working hard for many years. If Iхх≤0,5 A, then it is better to use the transformer at powers less than 270 W. With powers up to 300 W, the primary winding does not need to be rewound. But since in this case a power of about 600 W is needed, two TS-270 transformers were used. The primary windings were connected in parallel, and the secondary windings were connected in series (on one - winding IIa, on the other - IIb). Usually, for the 30-ampere version, each of the windings is wound with a double wire D1,8 ... 2,2 mm or a triple D1,5 mm. Based on the foregoing, the CH scheme is shown in Fig. 7.
References:
Author: A.G. Zyzyuk
Artificial leather for touch emulation
15.04.2024 Petgugu Global cat litter
15.04.2024 The attractiveness of caring men
14.04.2024
▪ Mars helicopter Ingenuity sets altitude record ▪ MOTOROLA FV500AA Compact Radio
▪ section of the site Interesting facts. Selection of articles ▪ article Zipper. History of invention and production ▪ article What is osmosis? Detailed answer ▪ article Manufacture of transformers. Encyclopedia of radio electronics and electrical engineering
Home page | Library | Articles | Website map | Site Reviews
www.diagram.com.ua |
See other articles Section 
Latest news of science and technology, new electronics:
All languages of this page