The purchase of expensive imported transceiver, usually associated with significant material costs. Often money to buy the PSU quite NS remains. And then to happy radio there is a problem of self-production and supply device. What requirements must it meet?
In the first place, along with the necessary homemade power supply should have good reliability, so that the likelihood of damage to the connected transceiver. Reliability, as is known, depends on the aggregate reliability of all components and their functional importance. In network power supply is crucial node of the voltage regulator. In this paper we describe a self-made network device power, the main "highlight" of which is the regulator circuit. The unit works without any problems for about six months in conjunction with transceiver KENWOOD TS-570D. Recently during the summer heat it passed additional tests, working about a day on an equivalent load at rated current.
The parameters of the power supply:
- The output voltage is 13.8 V (adjustable)
- Rated current I - 25 And
- Current short circuit protection - 27 And
- Drawdown of the output voltage at the rated load current is not more than 0.5 In
- Dimensions - 130 x 140 x 350 mm
No less important than the choice of the circuit of the stabilizer, is an engineering and manufacturing of the power transformer. This task is almost always associated with a lot of difficulties - it is necessary to get the right size iron, wire required cross-section and, more importantly, to produce a laborious winding. All these factors cause the deaf hams aversion to self-production of the transformer and the desire to get ready. That, in turn, delays the moment airing on the new radio into the "back burner".
Actually, homemade transformer - not such a difficult thing. The eyes are afraid, make up! From my experience as a core I prefer to use a W-shaped plate. Despite the fact that the size of the transformer in this case is somewhat larger than with a toroidal core, technological convenience take precedence.
You first need to assess the suitability of existing core or to figure out what to look for. Then calculate the diameters of wire and number of turns of the windings and, finally, to evaluate the obtained results. Looking into the old directory, you can find there the following approximate formula:
It should be borne in mind that the number of turns of the primary winding turns out to be slightly smaller, and the secondary is greater than design. However, you first need to wind the primary winding with a margin of 20 to 30 percent. The stock will come in handy when further adjusting the number of turns for the optimal operation of the transformer. During winding, it is desirable to count the number of turns for subsequent correction of the calculation of the parameter "N".
After a rough winding winding network should be secured CE coils, collect the magnetic circuit and measure the current of the primary winding at idle. This measurement will give a rather complete information about the quality of the work performed At this stage. The measured current depends on the overall capacity of the transformer or, more simply, from the size of its core. For transformers with a capacity of 200 - 1000 watts no load current may have a magnitude of about 100 - 150 mA. If the measured current is smaller than this value, this means that the efficiency of the transformer will be below normal and HC will be able to get the expected output. In this case, the windings need to rewind a portion of the turns and again to repeat the measurement current.
To avoid unexpected hassles associated with random turn-to-turn closures, the first measurement is desirably done by including in series with the winding of a mains lamp power not less than 100 watts. If you plot the dependence of the no-load current of the number of turns, then this graph can be seen pretty dramatic reversal, which shows that when a certain number of turns even a slight reduction leads to a sharp increase in current. So, the best you can count the number of turns, when a little graph of the current NS reaches the fracture site up. The General criterion of quality is made of the primary winding is the lack of appreciable heating of the transformer core when running without load for several hours.
I want to mention that to try to wind the transformer with the method of "round" is very time-consuming job. The primary winding is quite possible to do "in bulk". Modern magnet wire with reliable varnish insulation allowed the technique of coiling. It is only necessary to monitor the uniformity of the distribution of turns on the winding surface, so as not to create plots with high interturn the potential difference. So, the primary winding is finished. The coils fixed, made of flexible findings and turns over padded insulation malagaplan material, which you can use tape made of PTFE, taken from the capacitors FT-3.
Now we have to perform the shielding winding network. This is best done with a thin copper foil by wrapping her in a single layer once the surface is made of a network of winding. The shielding winding has only one conclusion. which then joins to a common (ground) power bus. The shielding coil in any case should not be closed, otherwise it would lead to the death of your transformer. Between and overlapping the ends of the foil should definitely pave reliable insulation. After isolation of the shielding winding is possible to start no less important is the secondary winding, high current windings. Its construction depends on the choice of the rectifier circuit. If you plan to use a bridge rectifier, that dangle simple bezothodna winding. If the transformer has enough free space, it is desirable to use vapour phase full-wave rectifier circuit with two diodes and respectively double secondary winding with the secondary output. Losses in the winding and the rectifier in this case will be smaller than in the first case.
For powerful secondary winding is typically used thick copper wire with a diameter of several millimeters or copper tavern. This makes it difficult to manufacture manual-winding and can cause damage to the underlying insulation of coils. In my design I used a kind of "Litz wire"- the harness of several stacked together, wires with a diameter of about 0.8 mm. With this method of winding is important to monitor the parallel arrangement of the individual wires of the harness so as not to cause electric current to the mismatch between the individual wires of the winding.
An important question is at what voltage to expect the secondary winding? The answer depends on many factors. Such as the properties of the magnetic circuit, the capacitance of the filter capacitor of the rectifier, the limits of possible fluctuations of the mains voltage, the properties of the voltage regulator. On many of these issues easier to get an answer, placing the corresponding experiment than to try to calculate theoretically. In any case, it is necessary to focus on the magnitude of the rectified voltage of about 20 Volts. Increasing this figure is useful for increasing the stability of the output voltage due to the larger supply voltage to stabilize. However, this in turn leads to an increased thermal mode of operation of the transformer and stabilizer, to the necessity of using electrolytic filter capacitors at higher voltage, that is more expensive and overall.
In short, it should follow the rule of "Golden mean" and not allow to achieve unreasonably high load parameters the interpretation of the modes of the nodes of the power supply. After the test winding of the secondary winding is necessary not to forget to re-check the idle current power windings. It should not increase by more than 5 to 10 mA. Further, the quality of execution of each stage of the Assembly of the device power it is desirable to check, loading it to the equivalent, which can serve as a garland appropriately connected incandescent lamps. I used an old 12 volt automotive lamps from lamps high beam, connecting the two helices. One lamp in such inclusion hogs about 6A.
Bringing the scheme rectifier with filter capacitor, measure the load capacity, medium-voltage and voltage ripple at the rated load current. Of greatest interest is the voltage value minimum period of pulsation. Measured by the oscilloscope, it should be NS less than three volts (min stabilization reserve) higher than the output voltage of the stabilizer and, in our case, is 13.8+3=16,8 V.
It is important to choose correctly the capacity of the filter capacitor. Usually it is chosen of the order of 100000 ICF. I had difficulties with the acquisition of a capacitor and scored the required capacity, connecting in parallel the existing capacitors. I managed to place them in all the nooks and crannies of the housing unit, sticking capacitors glue "melt". Conclusions poles need to connect wires at one point, in the immediate vicinity of the outlet. You can use a condenser of smaller capacity, but it is necessary to slightly increase the voltage of the secondary winding by controlling the voltage ripple under load, as described above.
When the Assembly of the transformer and the rectifier was finalized, I had to lie to a difficult question selection circuit of the voltage regulator. On the one hand, there are lots of schemes with transistors as a regulating element, on the other hand, it would be tempting to use the stabilizer is fully integrated execution. The latter option would be preferable and its manufacturability, and quality are guaranteed by the chip, if not for the price.
Earlier and I now widely used in their chip designs KREN. All they are good - price, availability, and its parameters, not afraid of a short circuit. Only here the current is too small. Only about two amps. Imported analogues of our LM317T chip is cheaper, more stable and more powerful, hold it for three amps, but it's still far from what is needed. Even earlier, to increase the power stabilizers I connected the findings of two such circuits in parallel. The maximum current was increased the same way twice.
In this case I made an experiment and connected in parallel as many as nine chips, placing them evenly on a common heat sink. According to the standard scheme annexed two resistors to a common control output and included a simple diagram. The results of load tests fully met my expectations - excellent stabilizing properties of the scheme remained the same as in the individual chips, and the maximum current is increased in proportion to their number.
(click to enlarge)
Used in the stabilizer circuits before installation should be tested separately. The output voltages of each chip can differ by a small amount. But I deliberately did not seek to select instances with the same parameters, reasoning as follows - let, at a current of, say, two amps is just one of nine chips. But when the current is increased to a value more than three amps, loaded the chip will feel overloaded. It will begin to operate internal protection circuitry against short-circuit, that is, gradually increasing its internal resistance, and current will be redistributed to the next chip. This will continue until all the chips will not be included in the process voltage.
With further increase of the current above the nominal will be observed a rapid decrease in output voltage - finally work function overload protection. This scheme, except for the extreme simplicity and low for used items, has another advantage - the best heat distributed through radiator circuits.
In my design, there were three needle radiator from line scan TVs electronics 401", mounted on a common aluminum base. Under the radiator just in case mounted cooling fan, however, enable it to occur - the heat sink temperature, even after intensive work on the transmission is low. Adjusting the output voltage of such a scheme can be implemented in a very wide range from two to several tens of volts. In table.1 shows the average value of the resistance of the adjusting resistor (variable resistor 3.3 K ohms), depending on the required output voltage.
Table 1
Voltage, V
Resistance, Ohm
Voltage, V
Resistance, Ohm
2
115
8
1057
3
276
9
1215
4
433
10
1368
5
586
11
1530
6
745
12
1686
7
903
13
1835
Note that the radiator circuits must be isolated from the casing of the power supply. The case itself is better not to connect electrically with the regulator circuit, and to connect to protective ground. On the input line voltage, it is desirable to install a simple LC filter. It will protect the transceiver from the mains interference. Indication of the power supply operation is performed with two lamps HL1 - any neon, HL2 - filament lamp. This also acts as a discharge resistor. For the duration of its glow after turning the unit off from the network, you can judge the quality of the capacitor C5, and the brightness is about the stability of the output voltage.
In conclusion, I will say that the cost of one chip LM317 in Moscow is slightly more than 3 rubles - almost two times cheaper than our domestic KREN, but on the reliability surpassing her.
Author: S. Makarkin, RX3AKT; Publication: N. Bolshakov, rf.atnn.ru