I ask the English version site readers’ pardon for possible grammatical and other mistakes, because my native language is Russian and I have no possibility to make a full translation of the whole Russian site www.cqham.ru/ut2fw , because much time and forces will be required, which, as a rule, are not enough. That is why I offer a brief translation…

HF transceiver blocks.

Before giving the description of blocks themselves, I want to tell the readers some conclusions, which were followed after manufacturing of several tens of transceivers with various self-made frequency synthesizers and power (up to 1000 W) transistor outlet cascades. Inventing of something original new with respect to this subject-matter has no sense today. Something which is possible to repeat in domestic conditions with more or less worth quality, the circuits are represented in various variants in periodicals, and many things can be borrowed from analogous apparatus, developed by the industry. There is no hope that in domestic “laboratory” it is possible to produce a high-quality universal transceiver with “super” high dynamic range. The cost of the utter professional receiver or transceiver amounts to thousands dollars; not one person, but whole collectives of high quality specialists develop and produce it. It is just ridiculous to attempt to create something similar or to compare our homemade products with such a technique. The nowadays reality shows that for self-made making, there are possible experiments with simply circuit technique, from which by persevering adjustment it’s possible to obtain quite high receiver parameters, and here it’s possible to compete with cheapest commercial transceivers. But for the adjustment and manufacturing it needs at least a kit of devices, consists of measuring instrument of frequency characteristics, the high-frequency voltmeter, the frequency meter, the measuring instrument of inductance, Q-factor and capacity, the high-frequency oscillograph, the control receiver, the high-frequency generator, the qualitative multimeter. And it is not necessary to hope that if there is no experience in design of such a technique, for transceiver building it will be enough to read the description of transceiver and another books on such a theme.

That is why before starting of manufacturing of such a complex radio engineering device as HF transceiver, it needs to estimate own strength and possibilities. The optimal variant is the transceiver building from prepared units, although it doesn’t give the maximum qualitative result without a careful adjustment of each unit separately, and then additional transceiver adjustment in whole. I hope that my warning will keep you from in vain spent time and rash decision to build a transceiver for yourself without the operational experience and devices.

And for that honourable and dear public, which doesn’t afraid of such obstacles and they made such a technique before, with a positive final result, I continue the description.

Frequent acquaintance radiofans complains about I offer too complex constructions of transceivers for repeating, give the prerequisite for developing and producing of some versions, in my opinion, quite simple and accessible for repeating. The ideas of TRX “Rosa” and “Ural 84M” >>>>> serve as a basis, as the most simply, and developed with quite high parameters. There was made an attempt as much as possible to decrease the circuit complexity without worsening of reception section characteristics and service comforts. As a result of creative tortures – during some years the transceiver construction was developed, that by technical characteristics doesn’t yield to many of the commercial transceivers. The appearance of one of such exemplars (some variants are made) it’s possible to see there >>>>>.

Radio-frequency section of the transceiver has no pecularities, because the HF receivers block diagram for last 20-30 years practically hasn’t changed, there is nothing “new” in this construction. The main perfection, for example in commercial TRX, goes by blocks of processing of signal, transformed up to tens of kilohertz by digital way, i.e. by various DSP systems. IF block is not differ even in competitive firms transceivers. New developments’ difference is mainly by appearance, color and button quantity, instead of qualitative receiver characteristics. For example, in the five-page description of new FT920 – magazine “HF and VHF" №2/99 year, about improvement of reception quality, the author laconically said – “it is pleasure to listen to ether on this transceiver”. Any of the dynamic receiver characteristics were not measured by the producer. More attention is given to the description of back panel sockets, than to the question, why “it is pleasure to listen to ether on this transceiver”, but not on, for example, FT1000. Analyzing all this rather cheap variety, you come to a conclusion, that now there is no universal decision of qualitative reception in the whole HF range L. At last, the author haven’t met a quite cheap receiver, which works well both on 160 m and on 10 m, with taking into view, that in overwhelming majority we use the same “low-frequency cords” as aerials and for high-frequency ranges.

The measurements of the various reception equipment characteristics of both industrial domestic and foreign production, and the self-made transceivers, carried out by Jury UR4EF, are very interesting. All measurements are shown in one table -

Table No1

CALL TRX		RFRX mkV	DB1 dB	DB3 dB	Bandwidth at different levels in kHz
					6bal	9bal	9+20dB	9+40dB	9+60dB	9+80dB
UT7QF TS2000	AMPoff AMPon	0,95 0,19	100 101	96 97	2,04 2,04	3,08 3,25	3,5 5,01	6,82 14,05	21,17 49
UT7QF TS950S	AMPoff AMPon	0,38 0,15	104	95 89	2,7	2,84	2,99	4,95	19
US1QV TS850S	AMPoff AMPon	1,58 0,5	>110 104	96 98	2,24 2,24	2,72 2,84	3,28 3,4	3,55 3,93	6,56 12,31	19,23 40,5
UT6EC	TS50	0,15	97	91	1,6	3,0	3,3	10,7	11***
UU4JG	TS570D	0,32	91	97	2,06	2,51	3,43	17,57	79
UU2JJ IC746 without alterations	AMPoff AMP1on AMP2on	0,25 0,14 0,11	106 103 103	98 97 86	2,33 2,33 2,33	3,0 3,31	3,4 3,62	4,41 12,34	21,2 112
	AMPoff band 1,9 KHz				1,9	2,63	2,84	4,6	45,5
UR7QR IC746	AMPoff AMP1on AMP2on	0,39 0,2 0,2	107 113 111	93 97 86	2,14 2,14	3,1 3,15	3,47 3,48	3,93 4,2	11,57 6,66	183 132
A filter per 1-st IF 70MHz is added
RB5GK IC746	AMPoff AMP1on AMP2on	0,32 0,14 0,1	103 103 95	100 99 91	2,3	2,96	3,32	4,34	19,05
UT7QF	FT100	0,15	104	90	1,96	4	4,5	15,8	76,1
UT7QF	FT100D	0,14	101	95/88	2,1	3,55	4,43	12,31	91
UT5EC	FT847	0,3	93	86	2,06	3,16	3,4	8,97	55
UR9QV	FT847*	0,36	102	92	2,59	3,15	4,35	5,31	7,86	200
UR5IOK	FT847	0,27	107	93	2,27	2,96	3,49	7,64	29
UT4EX	FT990	0,25	94	92	1,62	2,75	3,0	6,49	50
UT3GX	RA3AO	1,7	101	91	0,9	2,6	3,6	4,1	16
US5EQN	AMPoff AMPon	0,6 0,5	113 103	93 90	2,1	2,8	3,1	6,6	28
UR5EL	R-680	0,12	96	94	3,1
UY5RM	R-680	0,16	94	88	1,6	4,29	4,42	10,0	14
UR4EF	AMPon	0,16 0,5	115 105	93 88	2,85	3,4	3,8	6,9	33
UR5EL		0,5	105	88	2,6	3,6	3,9	4,7	10,4
UR5EL	R-399	0,67	89	92	2,74	3,41	5,58	12	59
UR5EQQ	R-399 1983 year.	0,2	94	105	1,8	3,4	6,8	19,2 In the 1-st mixer KP304
UR6EJ		0,85	93	89	-	2,7	3,3	7,9	52
US4EL	AMPoff AMPon	1,18 0,24	99 97	83 90	1,8	3,6	3,8	4,5	25
UR6EH	UA1FA	1,34	89***	70	There is no digital scale
UR6EH	Contour 116	0,6	77***	71	2,1	2,6	2,7	8,5	140
UR4EWT	Ether-M	1,9	84***	59	There is no digital scale
UT4EI	UW3DI	0,34	92***	85	2,7	2,8	4,0	8,3	50

Measurement technique:

RFRX (sensitivity) - the signal from the noise generator (the maximal amplitude is 6 mV) was given via the attenuator till the output signal had increased up to 10Db.

DB1 - the signal 7034kHz which value is more than receiver noise on 10dB was given, and then an interference second signal, which frequency is 7012kHz, was given via the attenuator, while the signal, which frequency is 7034kHz, was increasing (decreasing) on 1dB.

*** It means, that AMP or mixer are poor-quality units in the transceiver …

DB3 - two signals of 7012kHz and 7056kHz were given, and the combinational interference on a frequency about 7099kHz had not reached the value, which is more than a level of the receiver noise on 10dB.

Band - during band measurement at the big levels in devices with ***, the measurements are not correct, because at the big levels their RX noise decrease … because the AMP or the mixer are overloaded.

Measuring generators (7012, 7034, 7056kHz) - quartz with a resonant contour, further the resonant amplifier, then the quartz filter with a band of 80 - 90Hz, attenuator 6dB and the adder on a ferrite ring, output amplitude is 0,5V for 7012 and 7056kHz (+114dB), everything is in the shielded boxes.

For 7034kHz on an output, the amplitude 0,12 - 4mV is adjustable. More detailed information and comments can be looked on a site http://www.cqham.ru/ut2fw Circuit technique decisions of IF channel construction, even in commercial TRX, repeat from model to model, although the transceiver prices can differ more than in 1,5-2 times. Especially it concerns the entrance circuits. Three units, which determine the quality of receiver work, are the first mixer, the basic selection filter and smooth oscillator. The economy on these units always deteriorates the work of the whole device. Therefore I try to apply here circuits and elements with maximum high parameters.

The question about a choice of oscillator must be considered separately, because a lot of "lances are broken" over it. Originally there were tests of the various types of usual heterodynes with reorganization of the variable capacity by the air capacitor, when the mass of radioamateur public could not think about synthesizers. Scanty information on a question- "what generator is the best for using in the transceiver?"- compelled to check all practically- to make these units on various element base- from lamps, nuvistors to "lambda-diodes". Usual coils, pieces of a coaxial silver cable, coaxial resonators, quartz wafers with frequency withdrawal, etc., were applied as oscillatory system. Occured signal generators of industrial manufacturing also were tested as heterodyne. At the same time, RFRX, DB1, DB3 were measured at the transceiver.

In some (now I don't remember precisely its title) thick English professional book about reception devices, I have read, that there is a sense to struggle for heterodyne noise up to value of 100dB. The further "titanic efforts" for noise decibels don't result to desirable (or if you want, assumed) proportional improvement of quality of work of the mixer and accordingly the whole receiver. After conducting the "laboratory works" with various types of heterodynes, it was necessary to agree with the clever English book- "Struggle for decibels" is meaningful up to any certain threshold. It is impossible to catch a difference in receiver work, at least, on measured parameters, whether the generator on nuvistor with the coaxial resonator as oscillatory system or the usual field-effect transistor oscillator with the coil, reeled up by a silvered wire, is established as the first heterodyne. The circuit engineering RA3AO, with some changes and carefully selected and adjusted element base, was used in the receiver. RX parameters are rather high - for example, DB3 was not less than 102dB at signal injection with separation of 8kHz.

The only moment which can't be measured by devices, but it is marked by main "device" - hearer ears, is "transparent sounding" of ether. Maximal "transparency" was achieved, when generators worked on VHF frequencies with the subsequent frequency division. Most probably, it is connected not with frequency division, and with the fact that on higher frequencies it is possible to get very high Q-factor of oscillatory system. As the theory says- "width of the basic lobe of spectral density of the generator noise, is inversely proportional to Q-factor of oscillatory system ". And if overall dimensions are identical, the greater Q-factor is easier for getting on the higher frequencies. Respectively, the collateral noise level of the active oscillator is less, for example, on 100MHz frequency, rather than at the same overall dimensions on 10MHz frequency. In the receiver, high-quality heterodyne with small lateral noise proves, that it is possible to listen to two - three stations, which work on one frequency, due to hearing selectivity to different sounding timbres, choosing required. There is no modulation of one station by another. The 40m band looks as 15m at the maximal transmission years. In the low class receivers with "noisy" heterodynes, the loudest station "stuffs" the channel, and on its background weak signals cannot be understood. These features at using of various generators were possible to observe only in the receiver, "aimed" at getting of the maximal parameters by whole circuit engineering and element base. At connection of the same generators to the transceiver "Ural 84M" with DB3 in 95dB, it was not possible to find out any differences. It means, that at the receiver designing it is necessary to be guided by a principle of "rational sufficiency"- that is evidently traced in circuit engineering of commercial transceivers.

Because of the author does not imagine the modern transceiver without a frequency synthesizer with microprocessor control, the qualitative receiver characteristics will be determined by this unit. It is because of the synthesizer is the most complex and expensive unit in comparison with the first mixer and the basic selection filter.

For those who doesn't like synthesizers, it is possible to recommend using of ready blocks of radio stations generators and receivers of commercial manufacturing. To receive required frequency stability, the block design should be rigid, with a minimum of switches in the oscillator. For example, the generator block of Soviet radio station R-107m is very successful for these purposes, it will be necessary to attach the frequencies divider to it only.

Below I bring the table of measurements of several receiver parameters, recently made by the device, kindly given by Jury UR4EF, by which he made metering, given in Table No1. Specially, measurements are carried out by this device and by a similar technique, to receive comparable observed data.

Table No2

CALL TRX		RFRX mkV	DB1 dB	DB3 dB	Bandwidth at different levels in kHz
					6bal	9bal	9+20dB	9+40dB	9+60dB	9+80dB
UT2FW TS-870S	AMPoff AMPon	0,38 0,12	110	103 93	2,7	3,08	3,5	6,82	21,17	37,2
UT2IV IF unit No3	AMPoff AMPon	0,38 0,31	104	107	2,8	3,6	4,0	5,7	20,4	124,1
UT2FW IF unit No2	AMPoff AMPon	0,27 0,21	107	100	2,7	3,4	4,1	4,9	11,4	107,6
UT2FW TRX No1 IF unit No3	AMPoff AMPon	0,38 0,28	107	100	2,9	3,8	4,3	5,0	9,7	59
UT2FW TRX No2 IF unit No3	AMPoff AMPon	0,71 0,66	112	110	2,7	3,6	4,1	4,6	6,3	30,5

There are some comments to the table: 1. TS870S - all changes found on the Internet about it, are entered. 2. UT2IV - the transceiver is examined in various tests and competitions. It was taken on bicycle race simultaneously with IC735- the self-made transceiver has shown comparable results and in some moments was more preferable. 3. Other transceivers are assembled on circuit engineering described below. Transceiver No2 specially has been made for using on 160-20m bands - therefore RFRX parameter is understated. In my opinion, the small "fork" of figures between DB1 and DB3 is connected with small separation between interference frequencies and a useful signal - 7,012MHz and 7,034MHz. When interference levels approximate to 1V- here not only receiver parameters, but also noise parameters of the interference at small channel selectivity from a useful signal influence figures of measurements! But the whole measurement technique, which Jury UR4EF suggests, is kept to get measurements comparable with its measurement table. Only by this thing I can explain a fact, that figure DB3 in transceiver UT2IV exceeds figure DB1- frequency on which there is an intermodulation interference, is distant on 44kHz from the closest interference, and at DB1 measurement, the interference is distant on 22kHz. Accordingly, interference signal noise at DB3 measurement are twice further from accepted frequency and less mask the measurement reliability. There was the same situation at measurement of parameters of transceivers UR6EJ, UR5EL, UR5EQQ - look at Table No1.

The detailed description of various variants of IF unit No2,3,4 I will not quote in the English version of the site - there are descriptions in the Russian version of the site www.cqham.ru/ut2fw here I will quote the description of last IF unit No5.

There are direct references to circuits and view of ready boards:

IF unit No2. The circuit >>>>> the board picture >>>>> >>>>> >>>>> >>>>>

IF unit No3. The circuit >>>>> the board picture >>>>> >>>>>

IF unit No3. The circuit >>>>>

IF unit No4. The circuit >>>>> the board picture >>>>> >>>>>

Description of IF unit No5.

Firstly about board units, which repeat in all versions without changes. These are: 1. The reference crystal-stabilized generator on the field-effect transistor, Fig. № 1; 2. The preliminary amplifier of low frequency of the receiver, Fig. № 2 (with the amplifier of S-meter and AGC); 3. The final LFA of the receiver, Fig. № 3; 4. The microphone amplifier of the transmitter, Fig. № 4; 5. "Automatics" - the unit which forms the control voltage +13V RX, +13V TX, VOX. Fig. № 5; 6. The telegraph generator, Fig. № 6.

Marking of elements on these circuits is through, i.e. which will be in the description of last version of the basic board, both on the circuit and on the printed-circuit-board. Near outputs of microcircuits and transistors, there are voltage constants on circuits, specified with bold type. Voltages with a sign ~ are the high frequencies voltages measured by the lamp voltmeter. Inscriptions on circuits in the frame are interconnections on the board.

Reference quartz generator.

The circuit >>>>> The generator is realized according the circuit Fig. №1 on field-effect transistor VT9. USB/LSB modes are provided due to shift of generator frequency from bottom to upper slope of gain-frequency characteristic of the quartz filter. As a rule, the reference generator frequency is chosen on 200-400Hz below than a filter transmission band, measured at the level -6dB of the filter bottom slope. For the exact adjustment of required frequency on the bottom slope of the filter, coil L12 serves, and for the adjustment of frequency on the upper slope, the capacitors S79, C77 serve. Mode of transistor VT9 and accordingly, required amplitude of the output signal, it is possible to select with help of R50, R61. Depending on the quartz type and its frequency, selection of capacitors C72, C81 will be required (their value will influence the output signal amplitude!). It will be better to use the most active quartz with maximum Q-factor, from quartzes, usually remained after crystal filter manufacturing. Don't forget, that the frequency stability of the reference generator will depend on a quartz quality, and accordingly general stability of the transceiver frequency holding. On transistor VT10, the buffer cascade which adjusts with the low detector resistance, is realized. Loaded amplitude of the output signal of the reference generator on output Fоp is 0,7-0,8V. On the board the reference generator is located in «the most distant angle» from input circuits, to minimize from it the pickup on the receiver input. Therefore its interconnection with the detector goes not by means of the printed-circuit-board track, and by means of thin coaxial cable. Cable quality should satisfy our problem.

Preliminary low frequency amplifier.

The circuit >>>>> Because of fact that in IF block reversible cascades with a small amplification factor are used - for obtaining high sensitivity of the transceiver, in the first cascade the LFA transistor VT17 KT3102E (C547C) with a small noise coefficient and big amplification is used. For additional individual adjustment of this cascade resistor R103 is used. At the input the filter (R83, C102, R84), which cut-off high-frequency components, is placed. In the emitter circuit, resistor R92 and capacitor S110 are placed, by means of which it is possible to form desirable gain-frequency characteristic of the cascade. At adjustment, for obtaining the maximum possible characteristics, it is necessary to select the most qualitative VT17 KT3102E. The transceiver sensitivity depends on quality of operation of this cascade, which can be obtained - therefore it is necessary to be in earnest about choice of the transistor and its subsequent adjustment. As a criterion of qualitative operation of LFA it is possible to consider the parameter- at signal feed to the connection point of the detector input with a connection coil of a circuit of the last IF cascade of a level 0,5 mkV with IF frequency - the signal should be distinctly audible in phones. Transistor VT23 serves for elimination of clicks at RХ-TX switch-over. The chain of diodes VD30, VD33 provides automatic gain control (AGC). Units R66, C86 select its time behaviour. Such a way of regulation of cascades' amplification, due to shunting diode chains, does not change characteristics of cascades, does not break their operation. Depth of regulation, depending on the characteristic of shunted circuit, usually is not higher than 30Db per one chain. Therefore to support the deep AGC there is required not less than three such chains in a section of the transceiver amplification. For those who don’t love loud sounding of the transceiver, it's expediently to install germanium VD33- it will softly limit the amplitude of LF signal in this circuit. Filter R76, R98, C108 in addition cut off the high-frequency noise which has arisen by operation of VT17. Further the signal goes to the second LFA cascade - operational amplifier DA3. The most qualitative for this cascade is chip K574UD1. It is high-speed, with a high dynamic range and small noise. The amplification factor is regulated by resistor R67, cut in high-frequency part of gain-frequency characteristic by the capacitor C82. At connection of units R56, R57, R58, R24, C73, C74, C75, R49 by contact electrodes K5:1 of relay К5, the pass band of gain-frequency characteristic is sharply narrowed down - for CW mode. Gain-frequency characteristic slopes are flat- therefore with such filter it is possible to listen even to SSB signals- it frequently helps in "dumps" on LF bands. Top of gain-frequency characteristic is narrow - about 250Hz on a 10Db level, average frequency of a filter is 1000Hz. From output DA3 the signal via R78 goes on terminal LFA via the adjuster of volume located on the transceiver front panel. The signal via R72 goes to the amplifier of AGC signal fulfilled on VT8. Diodes VD27, VD25 are included according to the voltage-doubling circuit. The chain from R43, C70, R47 determines characteristics of the charge - discharge AGC loop. At nominals indicated on the circuit, usually AGC characteristics satisfy both to SSB reception and CW signals. By resistor R42 we calibrate indications of S-meter, as which it is possible to apply the milliammeter with a current of arrow full deflexion up to 200-300mkA. At feeding of 50mkV on antenna input of the transceiver, the arrow of S-meter must show 9 points.

Terminal low frequency amplifier.

The circuit >>>>> The amplifier has no any features. On input of LFA "VOLUME1" the signal from the volume regulator on the front panel goes via contacts K6:1 of relay К6. The relay is applied to provide telegraphic listening without any noise and pickups in the transmission mode. As practice of the further using has shown, relay contacts in this circuit are sometimes useful at interconnection of the transceiver with a sound card of a computer. On the chip DD2 the self-listening generator CW and the shaper of telegraphic VOX are assembled. On units DD2:2, DD2:3 the generator is assembled, unit DD2:1 is the buffer, and unit DD2:4 is used in telegraphic VOX-C. Desirable generator frequency can be adjusted by units C60, R33, with the mentioned nominals frequency is about 800Hz. Capacitor S60 should have stable capacity temperature coefficient, otherwise, at warming-up frequency will vary. By resistor R41 we select a desirable level of a self-listening signal. Within small limits by units C109, R87 it is possible to change in addition gain-frequency characteristic and gain factor. LFA. The voltage filter R45, C80 is for interference elimination which arises at usage of powerful transistor wide-band amplifier. For elimination of "step" (the awry reproduction of weak signals) it will be required to increase a rest current of output transistors up to 30mA. For this purpose diode VD29 is substituted for silicon. This LFA badly works on load with a resistance less than 50 Ohm. Therefore it is desirable to apply phones with a resistance more than 200 Ohm. Very well LFA works on Soviet phones such as ТA56 or similar with a resistance of 1,6 kOhm.

Microphonic amplifier.

The circuit >>>>> The signal from a microphone is given on the amplifier fulfilled on operational amplifier DA2 via filtering chain C54, C56, R25. To quality C56 higher demands are made - reactors (the Chinese production), which gave noise in this circuit, came across. Any gain-frequency characteristic adjustments in the amplifier are not used - because it is difficult to recommend any concrete circuit of forming of gain-frequency characteristic of the microphonic amplifier, not being conformed to features of a voice and a speech manner of the operator. The unique regulation of this cascade is obligatory - it is necessary to set a required amplification factor by resistor R19. It is necessary to be guided by a ratio of the modulator signal levels - usually the LF signal should be no more than 1/3 of a level of a signal of the reference quartz generator. If microphone sound is loud, the low-frequency voltage on input VT6 can achieve 1V. Through chain R31, C65 the signal goes to circuit VOX SSB and through repeater VT6, to the balancing modulator. Contact K3:1 of relay К3 serves so that in the reception mode the signal from operating amplifier DA2 did not get on input of LFA receiver. As LFA sensitivity is very high - then even voltage removing from the cascade on VT6 and additional blocking of the base circuit by a transistor key did not give desirable result - therefore it was necessary to apply relay contacts. The screened cable from the microphonic connector to board on no account cannot be connected to the transceiver case in the connector. The cable screen is grounded directly on the board of the microphonic amplifier, and nowhere else has contact with the transceiver case. Usually, the dynamic microphone is used.

“Automatics”

The circuit >>>>> This unit made a good showing of almost failsafe operation and simple circuitry. On transistor VT12 amplifier VOX SSB is assembled. Diodes VD32 and VD31 straighten and double an amplified voltage. Units R94, C90 determine time of VOX holding and its sensitivity. Turning on of VOX system occurs by interconnection of emitter VT20 with case. Timing units for CW are turned on at output VT20, changing the value of C113, R88 it is possible to regulate time delay - holding of VOX in CW mode. The automation forming a voltage +13,8В ТХ and +13,8В RX is collected on transistor keys VT15, VT21, VT16, VT22. Diodes VD34, VD35 protect transistors of keys from the reverse current throws, which appear at relay switching. Transistors of keys strongly heat in case of their wide scatter of parameters. Therefore it is desirable to select transistor pairs KT814-KT815 for keys with approximately identical parameters.

Telegraph CW generator.

The circuit >>>>> The generator can be fulfilled both on field-effect, and on the bipolar transistor. But at usage in the generator of quartzes with low Q-factor and the field-effect transistor with a low steepness, manipulation is "soft"- the generator starts more "smoothly", but because of it, at high speeds (more than 150 signals per minute) short sendings merge. Therefore, if operation on the big speeds is supposed, it is necessary to check experimentally what variant will approach most of all. As a rule, inveterate telegraph operators very much like, when their CW signal differs from other signals, not in the worse side, hi!- the designer must choice, my task is only to acquaint you with the approved possible variants of the unit construction. In addition, the generator voltage is stabilized by stabilitron VD22. The generator switches on by shorting to output case CW-KEY. The sending shape is regulated with the help of R37, C66, R40. The generator frequency should be exposed on 1000-1200 Hz above the reference generator frequencies by changing of C58, C64 nominals. Required amplitude of an output signal we expose by capacitor C59 after obtaining desirable tone of CW signal. Levels of SSB and CW signals are indicated on circuits of basic boards.

IF unit No5.

The circuit >>>>>

The first mixer unit is fulfilled on DA1 - the assembly from four field-effect transistors KR590KN8. DD1 74AC74 forms two antiphase signals for operating of DA1. The amplitude of a high-frequency voltage can be regulated by change of supply voltage of chip DD1 by stabilitron VD21, a current via the stabilitron can be adjusted by resistor R18. 74AC74 at a supply voltage of 8,5V produces a meander with a level of 4,6-4,8V depending on an operating frequency, at a voltage of 6,8V - an output level is 3,6-3,8V, a voltage is 5V - Uout=2,6-2,8V. >>>>> The minimum working high-frequency voltage on DA1 is not less than 1,8V. With such mixer the transceiver has sensitivity not less than 0,3 mkV (AMP off), DB3 not less than 95dB, at feeding of two signals with separation of 8kHz. It is necessary to notice, that all meterings of the receiver parameters were carried out at a wide band of a crystal filter.

Transformator TV1 is fulfilled on ferrite ring К7-10, penetrability is 600-1000. The ratio of coils is adjusted according to an input impedance of 50 Ohm, winding I - 5 coils, II - 12 coils, if the input impedance of 75 Ohm is required, winding I should have 6 coils, it is coiled between coils of winding II which is evenly distributed on the whole core. The winding II is fulfilled by two twisted conductors, their diameter is 0,15-0,22mm. >>>>>

From the mixer the signal comes in convertible cascade VT1 via balancing transformer TV4. The balancing transformer is applied because the cascade on KP903 has low (about 12 Ohm) input impedance.

Transformers TV4, L3, L10 are realized on ferrite ring К7-10, penetrability 600-1000, in two wires of 7-10 coils, diameter of a wire is 0,15-0,22 mm.

On the scheme the first filter consists of six quartzes with nominals of capacitors, which more often turned out at usage of quartzes on 8,867MHz frequency, is shown. For a pass band narrowing it is possible to connect up additional capacitors C16, C42 by relay contacts. At connection of these capacitors, the upper slope of a filter is shifted to the left. Though the author does not consider this variant of CW filter qualitative, the board has connection layout with a possibility of the relay installation and introduction of such a pass band narrowing. After the first quartz filter the convertible cascade on VT4 follows. Chain of C26, VD4, VD5, R3, C11 regulates the cascade amplification at reception, and chain of C50, VT5, VT24 at transmission. Resistance of open channel KP103 is rather high, therefore depth of regulation is small, to provide the big depth of regulation, it is possible to include in parallel some KP103 (VT24, VT25) or to use other transistors with a smaller channel resistance. On VT3 amplifier - delimiter SSB of a signal is realized, which resonant load is loop L11, C48. As L11 the usual choke DM - 0,1 is applied. The choke inductivity and nominal of C48 will depend on selected intermediate frequency (IF). The amplification factor - accordingly a level of limitation is regulated by resistor R15. In case of need it is possible to output the regulation of a limitation level on a front board of the transceiver. For this purpose VT25 is applied, the resistance of a drain - source switch-over varies at feeding of regulative voltage on its gate. Regulative voltage of "Speec10-0V" moves from a slider of the variable resistor installed on a front board of the transceiver. Thus, we have two chains which change a level of limitation - changing an amplification of the cascade on VT3 by resistor R15 and by resistance of switch-over of VT25. By resistor R24 we select a limit threshold and amplitude of SSB signal on an input of a quartz filter in limits of 0,8-1,0V (a peak signal at a loud sound in a microphone). From L10 the received signal moves on the second quartz filter. The quartz filters coordination with cascades is achieved by selection of capacitors C10, C44, C30, C31 and resistors R16, R8. Nominals of capacitors are approximate, the exact value will depend on types of used quartzes, their frequency and a pass band of filters.

The second quartz filter is realized on 4 quartzes. The possibility of slide control of a pass band of this filter is introduced. For this purpose three varicaps VC1-VC3 are installed. Types of varicaps depend on used quartzes. If quartzes are applied to filters for which for receiving a pass band more than 2kHz, capacities of capacitors are not less than 47-68 picofarad, then varicaps should be applied to smaller maximum capacities. Or to adjust smaller capacities C37, C38, C39. On the scheme there are shown nominals of capacitors for filters, which capacities C21, C22, C23 usually are 100-130 picofarad at a filter pass band of 2,4-3kHz. The pass band is adjusted by feeding of voltage on an output “VBT 0-12V” from a slider of the variable resistor installed on a front board of the transceiver. Voltage +13V ТХ goes via diode VD19 and the pass band of a filter in the transmission mode extends up to maximum. The cascade on VT26 is introduced for getting of deeper AGC. As a result we have 4 AGC chains with a possibility of individual tuning of everyone and the big amplification in IF tract, that enables to reduce requirements made to the first LFA cascade. For the detector balancing resistor R18 is used.

Final selection of AGC chains elements should be conducted in the already working transceiver. The author works mainly SSB and therefore all designs tunes in this mode. The optimal AGC characteristics turn out if germanium diodes VD4, VD8, VD42 are used.

On figures there is amplitude-frequency characteristic of IF unit No2 >>>>>

6-th XTAL ZQ SK4-59 6-th XTAL ZQ X1-38

Press on pictures to get the full scene.

There is gain-frequency characteristic of IF unit No3 >>>>> on figures.

Two filters 6+4 XTAL ZQ SK4-59 6+4 XTAL ZQ X1-38

SK4-59 is a logarithmic scale , X1-38 is a linear scale.

Figure of gain-frequency characteristic of IF unit No5 >>>>> on SK4-59.

Figure view is converted in Photoshop. The limit of a smaller ray swinging is determined - accordingly gain-frequency characteristic on the axis X became wider, than on two upper figures. One can see as far as the lower filter slope is really tightened.

Board of range band-pass filters.

The circuit >>>>> >>>>> >>>>> >>>>> >>>>>

On this board are situated the wide-band amplifier of high frequency (VT1), the first cascade of the transmitter amplifier (VT2), attenuator, nine band-pass filters, which are switched on by decipherer DD1, controlled by binary decimal code on bus-bar D, which arrives from the frequency synthesizer. During the reception mode, the signal from the antenna socket, which is commuted with the help of the relay contacts, installed on the board of the power amplifier, goes on contacts K4, К3 of relay P4, P3. In case of need, these contacts connect up the attenuator from resistors R7, R8, R9 with attenuation of 15-20Db. Further the signal goes on contacts K5 of relay P5 controlled by TХ voltage and from them on band-pass filters. Three-circuit band-pass filters with an inductive connection with load, and capacity connection between spools, are applied. Application of such a variant is caused by several reasons - 1. There is not required a plenty of aligning capacitors as it would be in case of a capacity connection with load, 2. The antenna via tie spools is galvanically connected to the case that facilitates a situation with pickups of static electricity in the antenna. The static electricity directly does not get on an input of the receiver. Blocking of antenna slot by dischargers, chokes, resistors does not give a complete guarantee of safety. Real tests of transceivers with such version of band-pass filters construction have shown, that it is possible to work without additional security measures at the time of static electricity pickups, certainly, if there is no danger of direct hit of a lightning in the antenna.

Switching of filters occurs by means of relay RES49, REK23. Usage for switching of diode filters worsens dynamic characteristics of the receiver - therefore relays are applied. Handling on relay switching goes from the synthesizer on busbar D, by the code which is decoded by chip DD1. On the chip input RC filters are installed (R16-R19, C29-C32) which serves for pickup elimination from a digital part of the synthesizer.

The filtered signal goes on the disconnectable wide-band amplifier of high frequency VT1 (AMP). AMP using sometimes is required on high-frequency bands, it can be connected by contacts K1, К2 of relay P1, P2. As VT1, according to the required task, it is possible to use anyone high-frequency, it is desirable to use low-noise transistors. If it is necessary to save maximum digits of a dynamic band, it is necessary to apply powerful transistors KT610, KT606, KT939, KT911 and similar to them. If it is necessary to provide maximum sensitivity - it will be better to use low-noise KT368, KT399. Negative feedback elements determine parameters of the cascade. Common amplification factor is determined by R6, R3. Amplitude-frequency characteristic R4, C7 and not so much R2, C2. Best values for transistor KT368A, with maximum rise of amplification in area of 29 MHz are cited. Amplification can be lifted up to 22dB, by decreasing of R3 value and increasing of R2.

In the transmission mode the signal from the basic board via contacts K1, К2 goes on filters. Depending on a band, the filtered signal of level of 80-120mV goes via contacts K5 to the first cascade of transmitter VT2. The transistor works in a class A with a rest current being 20mA. The circuit of this cascade is similar to AMP. Transformers T1, Т2 are coiled simultaneously by two wires without cabling on ferrite rings with penetrability being 600-1000. The rings' diameter is 7-10 mm, the wire diameter is 0,15-0,18 mm. There is enough to have 7-9 coils. The beginning of one winding is jointed to the end of the second. On output VT2, the amplitude of high frequency signal is within the limits of 0,8-1,5V (effective value in off-load state). Rise of amplitude-frequency characteristic on 29MHz, we select by capacitor C15, an overall amplification of the cascade by resistor R13. The amplified signal from cascade VT2 goes to wide-band amplifier line which is located on a backboard - the transceiver radiator.

Wide-band amplifiers of power being 5W and 10W.

If there is no need to receive in the transceiver 100 W at once - it is possible to use these two variants of amplifiers - they are enough cheap and reliable. In them there are used the transistors of the Soviet manufacture specially developed for linear amplification of signals of a 1,5-30 MHz band at a supply voltage of 13,8 V.

Variant with an output power being up to 5 W. The circuit >>>>> Its cost price is not high, therefore it will be accessible to the majority of radio fans. The output power is practically identical on all bands. If you wish, it is possible to make on high-frequency zones an output power more than on LF. It sometimes is required, when the external powerful amplifier with a blockage on high-frequency bands is used. The first cascade is realized on transistor KT610. The best substitute to it is KT939A, such a transistor is specially developed for linear amplification in a class A. There are more modern transistors with best characteristics, but they are very difficult for finding. For example 2T996B, which coefficient of combinational components on 60 MHz frequency on second harmonic (M2) does not exceed 65 dB, and on third harmonic (M3) does not exceed 95 dB, not each lamp can provide such parameters. Transistor VT1 is used in a class A at a rest current of 120-150 mA. Transformer T1 is realized on a ferrite ring in diameter of 7-10 mm, penetrability is 1000. Winding in two wires without cabling, a wire in diameter of 0,24-0,30 mm, eight coils, interconnection of the beginning of one winding with the end another, form an average output. Upsurge of amplification on high frequency is provided by a negative feedback in the emitter circuit, it is selected with the help of C1. The overall amplification and gain-frequency characteristic slope can be selected by changing of R5, C2 nominals. The amplified signal via the coupling capacitor C6 goes on terminal cascade VT2. At using of KT965A in this cascade and if voltage is 13,8-14 V, it is possible to get not less than 5W power, all harmonics are suppressed at worst not less than 25dB. With such signal it is possible to work on ether and to excite any power amplifier without any complementary filters. Base current VT2 is stabilized by chain of VD1, VD2, VT3. Elements C4, R8 determine an amplitude-frequency characteristic of the cascade. Resistors of negative feedback R10, R11 improve linearity. The rest current in limits of 300-350mA, is determined by resistor R9. Transformer T2 can be realized on a ferrite ring in diameter of 16-20 mm with penetrability of 300-600, or it's possible to apply two columns from rings К10 with penetrability of 600-1000, there is enough 4 rings per column. Depending on resistance of prospective load, number of coils in the transformer is selected accordingly. The chain of resistors R12, R13 serves as the divider at measurement of the output power. Elements R14, C15 compensate non-uniformity of the power meter in the whole frequency interval from 1,5 up to 30 MHz. Chokes DR1 and DR3 should withstand the operating current - DR1 up to 150 mA, DR3 up to 1A.

On fig. 8 >>>>> the variant of 10W amplifier is shown. At adjustment of first cascade VT1 the primary attention must be given to linearity of the cascade work and the maximal efficiency on 29 MHz. It is not necessary to be fond of increasing of the cascade amplification factor, by reducing of R3, R4 and increasing of R5 - it will result to deterioration of linearity and stability of work of the whole wide-band amplifier. Depending on what power we want to get, high frequency voltage on collector VT1, loaded on VT2, is 2-4V. If it is not possible to find KT965A then it is expedient to realized this cascade duple on transistors KT921A, fig. 8. It is necessary to humble with some blockage on frequencies above 21 MHz, the output power with such a cascade achieves 10 W. It is possible to get very pure spectral signal with linear gain-frequency characteristic with power up to 5 W, by increasing negative feedbacks with elements R5-R8, R10, C9, R11, C10. On the circuit separate shift circuits are shown for each transistor - it is the version for "the poor radio fan" which does not have opportunity to adjust the pair VT2, VT3 with identical characteristics. If selection of transistors is supposed, then voltage circuits can be united, as it is made on the circuit of Fig. 9. Preliminary by resistors R14, R15 in chains base current stabilizers it is necessary to determine a rest current within the limits of 150-200 mA on each transistor, and then more precisely to adjust on the nearest even harmonic suppression, which can be heard on the additional receiver. Limits of adjustment of the rest current depend on a steepness of used transistors and quantity of consecutively included diodes VD1, VD2 and VD3, VD4. Chains C7, R1 and C8, R2 provide rise of the gain-frequency characteristic on high-frequency bands. The choke DR3 should provide a current required to the cascade (up to 2A) without the voltage slump on it. It is possible to reel it on a small ferrite ring which penetrability is 600 and more, wire diameter is not less than 0,6-0,7 mm, 10-20 coils are enough. Transformer T1 is realized as "binoculars" from ferrite rings in diameter of 7 mm, which penetrability is 1000-2000. Columns of "binoculars" are glued together of 3-4 rings depending on their thickness, column height is 9-11 mm. The primary winding is 2-3 coils of a conductive wire in fluoroplastic isolation, the secondary is 1 coil of varnished enameled wire of 0,7-0,8 mm. Transformer T2 is realized as "binoculars" too. Two columns are glued together from ferrite rings with penetrability being 1000, in diameter of 10 mm, columns height is 13-16 mm. Depending on a mode of transistors and a load resistance, it is necessary to adjust an optimum ratio of number of coils in windings. The primary winding is 1 coil from a braiding of thin coaxial cable with tap from the middle, or one coil from the combined two conductive wires in fluoroplastic isolation, the beginning of one is connected to the end of the second and forms an intermediate output. The secondary winding, in case of application of a coaxial cable braiding for winding I, passes inside this braiding. The quantity of coils of the winding II can change from 2 to 5 depending on execution of winding I, and they should be adjusted experimentally according to the best efficiency and optimum gain-frequency characteristic of the output cascade on a required load resistance. "Binoculars" cannot be glued without isolation on the printed-circuit-board because some of ferrite marks pass a direct current.

The low frequency filter installed at the amplifier output, cut the rests of harmonics which are higher than 32 MHz (L6, L7, C20, C21, C22). It is necessary to note, that low frequency filter on elements C34, L1, C35, L2, C36 is designed for resistance of 50 Ohm. If load considerably differs from this value, the filter needs to be re-counted or excluded, because in this case it will bring irregularity in gain-frequency characteristic of the amplifier.

The power amplifier of 50-100 W. Fig. № 9

The circuit >>>>> >>>>> >>>>>

The circuit of the amplifier used in the transceiver is cited on fig. 9. This unit is realized on the printed-circuit-board screwed to the back wall - radiator of the case. Solder-up of details is on the one side of board on sites. Such a way of installation allows to fix easily the board on the radiator and provides access to substitution of elements without board turning, thus process of wide-band amplifier readjustment becomes simpler. If the separate stabilized powerful power supply for the transceiver is used, then voltage for this unit can be increased up to 14,5 V, and for other TRX cascades the additional stabilizer on 12-13 V must be introduced. Such measure allows to increase the general amplification factor and accordingly will facilitate a problem of getting of the steady gain-frequency characteristic. The same power at the increased voltage can be got at a smaller current, and due to it slump of feeding voltage on supply lines can be reduced. Don't forget, that at a low-voltage feed of the transceiver and rather big output power, the consumption current can achieve significant values.

Cascades VT1, VT2 are similar to 5-10W amplifiers. Rest current of VT2 KT965A is up to 350-400mA. Capacitor C6 determines the gain-frequency characteristic and in case of a blockage on 160 m, its nominal can be increased up to 22-33Н. Application of KT965A for VT2 transistor at first sight is not absolutely logical decision, because the transistor is "very powerful" for such cascade and is used on 15-20 % from that "is put" in it. Attempts to apply "weaker" transistor in this cascade did not give desirable results. It was not possible to find among Soviet HF transistors qualitative for getting of 5-8W output power with high linearity at supply voltage of 13,8V. Resistor R7 serves for prevention of a breakdown of emitter junction at a return half-wave of a control voltage, and is calculated by formula R=S/2pFgrCe Base current of VT2 is stabilized by chain of VD1, VD2, VT3, R9, C9. Resistor R9 exposes a rest current. By means of elements of negative feedback R8, C4, R10, R11 it is possible to expose required gain-frequency characteristic and an amplification factor of the cascade. It is not required to install VT3 on the heat-dissipating device. The choke DR3 should tolerate a current up to 1,5A. Transformer T2 of "binocular" type consists of rings in diameter of 7 mm, penetrability is 1000, height of columns is 10-12 mm. The primary winding contains 2-3 coils of a conductive wire in fluoroplastic isolation, a secondary winding contains 1 coil of varnished enameled wire of 0,7-0,8 mm or the same conductive wire. Windings are located inside ferrite tubules, their outputs are in the opposite sides of "binocular" holes. Readjustments of the cascade lies in selection of rest current by resistor R9, correction of gain-frequency characteristic and amplification factor by resistor R8 and to a lesser degree by capacitor C4. Preliminary, it is necessary to reel up 3 coils on winding I of transformer T2. Final selection will be carried out at adjustment of the whole wide-band amplifier. From transformer T2 two antiphase signals go on duple cascade VT6, VT5 for the further amplification. The type of used transistors depends on a prospective output power. The most powerful and accordingly expensive are KT967A. From them it is possible to receive an output power being more than 100 W with very high reliability. With transistors KT966A it is possible to get not less than 80W of output power, but they have bigger blockage on high frequency bands. As experience of long using of these transistors has shown, at SWR being up to 1,5-2 they tolerate a double power overloading.

With transistors KT965A it is possible to get 50-60W on the output cascade with comprehensible dependability. At output power of 40-45W the amplifier endures practically any КCВ in a continuous duty.

Antiphase signals from transformer T2 via chains C16, R15, C17, R16 which form the required gain-frequency characteristic, go on output transistors VT6, VT5. Resistors R8, R17 serve for the same purpose, as R7. With the help of C15 a winding 2 of transformer T2 are adjusted on a maximum of output power on the highest operating frequency (29,7 MHz). The base shift stabilizer is realized on elements VD4, VD5, VT4. Transistor VT4 via a mica separator is fastened to the radiator. Choke DR4 is reeled up on ferrite slug in diameter of 4mm, length of 30mm or on ferrite ring with penetrability being 600-1000, in diameter of 14-16mm for winding convenience, wire in diameter not less than 0,8mm on slug up to filling, there are enough 7-10 coils on the ring. Chokes DR5, DR6 of DПM0,6 type can be applied or reeled up on ferrite rings in diameter of 7mm, with penetrability being 600-1000, there are enough 5 coils of varnished enameled wire of 0,35-0,47mm. Transformer T3 is a "binocular" from rings in diameter of 10-12mm, penetrability is 600-1000, length of columns is 28-24mm. Winding 1 is one braiding coil of a coaxial cable, winding 2 is two - three coils of a conductive wire in the fluoroplastic isolation, laid inside the primary winding. The exact quantity of coils of a secondary winding is selected at adjustment in accordance with nominal output power on uniform gain-frequency characteristic and with the best efficiency of the cascade. The rest current of 200-250 mA for transistor, is selected by resistor R24. More precisely the rest current can be adjusted according to the greatest suppression of even harmonics, which can be checked by the spectrum analyzer or the additional receiver. Output transistors demand obligatory pair selection.

To the output winding of transformer T3 "protection against the fool" is soldered, which consists of resistors R21, R22. If wide-band amplifier line lost its load or the unknown construction will be connected up instead of the antenna, then the whole power will dissipate on these resistors. Sooner or later there will be a smell of a burnt paint from these resistors - signal for careless "exploiter" - look, "something is wrong, we burn!!!" This elementary, but effective protection allows, in case of need, without especial fears to turn on the transceiver on transmission on unknown load. Certainly, use of ALC system is not excluded, but long-term experience has shown that the chain of two such resistors will secure emergency conditions when ALC system will not work. Most often ALC system falls out after the first thunder-storm if the antenna has not been switched - off.

The filter of the bottom frequencies with cut frequency of 32 MHz from C34, L1, C35, L2, C36 must be necessary adjusted according to the best efficiency on 28 MHz, by joining - separating wraps of L1, L2 coils. In case of using of the additional matching device between the transceiver and the antenna or at work with the external power amplifier, it is enough to use it for suppression of out-of-band radiation. In correctly made and adjusted amplifier the second harmonic level is no more than 30Db, the third is no more than 18Db, the number of combinational oscillations of the third order, in a pique bending around two-tone signal, is no more than 32Db. Contacts K1 of relay P1 connect antenna socket to wide-band amplifier in the transmission mode. Relay P1 is controlled via transistor key VT4 by ТХ voltage. Diode VD3 serves for protection of transistor VT4 against throws of a reverse current at the relay switching. P1 of RES10, RES34 types with winding resistance being up to 400 Ohm, needs to be preliminarily tested of operation reliability from 12-13V.

The circuit of interblock connections.

The circuit >>>>>

The supply voltage stabilizer of D1 digital part (the roll 5A) is fastened in any convenient place, for example, as it is shown on the draft of the case >>>>> on backboard the radiator. Damping resistor R1 has power not less than 2W. Capacity of capacitor C1 depends on wide-band amplifier output power (accordingly on consumption current throws). Certainly, TRX will be work quite good without this capacitor, but at voltage decreasing, the discharge of the accumulator, long jointing wires between the transceiver and a power unit, slump of supply voltage is possible, for which compensation this capacitor serves. The more capacity of capacitor, the higher its buffer properties. But there is a problem with the stabilizer actuation - very big charging current of the capacitor causes response of protection against short-circuit. Therefore it is necessary to increase an operate current of protection against short-circuit, or it is possible to use switch PWR with additional position, in which C1 is recharged by current-limiting resistor. Stabilitron D1 protects the transceiver from wrong polarity and if the supply voltage will be higher then stabilizing voltage of D1 (18 V). If D815E is applied - this is elementary, but reliable protection, it will operate already at excess of the supply voltage above 15V - FUS safety lock will burn down. The safety lock should be designed for consumption current. Variable resistors AF (LFA adjustment), RF (IF amplifier adjustment), PWR (adjustment of the output power) are situated on the face panel. Sometimes capacitor C3 is necessary for excitation elimination on maximum levels of loudness. Boards in the transceiver case are installed on metal racks. Wide-band amplifier board is fastened directly to the radiator without racks.

Power unit.

The circuit >>>>> >>>>>

This unit is realized in the separate case to not increase overall dimensions and weight of the transceiver, which is frequently used for work in field conditions and feeds from the accumulator. The circuit of the stabilizer has been tested during several years, and at this period it wasn’t possible neither to find out more simple and reliable circuitry, nor to operate. As it was showed before - the stabilizer was operated by the method of "tests and mistakes". The problem has been set - to make maximum simple and a reliable source. After numerous checks of various ideas and versions of such devices, the elementary circuit of the stabilizer "has ripened", which do for a feed of transistor technics being from 12 up to 27V. The stabilizer is with protection against short-circuit in load. And the most important feature of this circuit is fast-acting protection, the stabilizer discharges the voltage faster, than transistors of the transmitter output cascade fall out. It is possible to set a desirable current at which protection works. It is the version of the stabilizer with the maximal current being up to 12-15A. On one KT827A it is possible to get the maximal current being up to 18-20A, but in this case it is impossible to provide reliable protection against short-circuit. To get the reliable source with a current of 20-25A, it is better to apply parallel inclusion of two preliminary adjusted transistors KT927A. In the circuit of emitters, inclusion of resistors of 0,1-0,2 Ohm in series is required. As VT2 it is possible to use any low-frequency low-power transistor, for example KT502. It is possible to use arbitrary diodes in the rectifier, suitable to a current. It is not necessary to economize on capacity of capacitor C1.

Stabilization factor of the stabilizer low, therefore capacity of capacitor should be sufficient. In order to not hear a background, taking into account the big amplification factor at low frequency on IF unit, capacity of capacitor should be not less than 22000 mF, the more, the better. This circuit is very convenient, because the adjusting transistor does not need to be isolated from the case. The transformer should provide a required current at the voltage of 16-16,5V. The straightened voltage on capacitor S1 is 21-22,5V, therefore rather big power at the maximal current falls on VT1, for which dispersion the radiator is necessary. In a power unit which appearance can be seen here >>>>> the radiator is used, which is a back wall of the case, the size is 120х100mm with ribs in height of 20mm. At such sizes and the maximal consumption current, it heats up to 60-80'C, depending on the ambient temperature. If work in "a hot climate" or digital kinds of radiation are supposed, the area of the radiator needs to be increased. The current at which protection works, is set by means of resistor R2, its nominal should not be less than 200 Ohm. The resistor nominal is higher, the current is less. The maximal current is determined by VT1 quality - it is the composite transistor with amplification being not less than 600. In the transceiver for current consumption throws' smoothing at TX the capacitor of 10000 mF is installed, therefore at its turning on there is a throw of a charging current of this capacitor, and protection against a short-circuit works, the stabilizer is not started. It needs to adjust reliable activation of the stabilizer by means of resistor R3. The nominal can be in limits of 100-30 kOhm, it depends on VT1 parameters. Diodes VD1, VD2 must be anyone silicon. Limits of adjustment of an output voltage depend on stabilizing voltage of VD4. The stabilizing voltage of VD4 is less, the smaller voltage can be got on the stabilizer output. If there is supposed to make the powerful output cascade in the transceiver and at significant consumption currents, then the installation of power buses in the stabilizer and a wire which joint the power unit with the transceiver, needs to be realized by a wire corresponding to the consumption current.

Description of the synthesizer with AD9832 http://www.qsl.net/ut2fw/sintez