DIY low-power transmitter projects: PLL Synthesized AM Transmitter for 530-1710 kHz

Home | Audio Magazine | Stereo Review magazine | Good Sound | Troubleshooting

The AM88 is a low-power (100 mW) AM transmitter for use under Part 15 FCC regulations for low-power unlicensed applications, such as hobby, short-range AM broadcasting, carrier current radio, and experimental use. Old timers may remember the "phono oscillators" popular as construction projects back in the 1950s, using vacuum tubes such as the 12BE6 and 12SA7. These were penta-grid (5 grids) tubes that acted as an oscillator and modulator in this application. They operated in the AM broadcast band and were used as hobby AM broadcast stations and for playing records (remember them?) through the AM radio. FM was not as common then, and stereo had not yet been available in the form of radio broadcasts. These phono oscillators typically ran the tubes at around 120-150 volts at 3-5 ma plate current (360- 750 mW power input), used a length of wire as an antenna, and had free-running LC oscillators. Although they worked well, they were really just toys used mainly as wireless microphones and for experimentation.

This section describes a modern version of the old phono oscillator. This AM88 AM transmitter has features that would have been science fiction in 1950. It is crystal stabilized, covers a 10 to 1 frequency range, runs off a 12-volt supply, and has high level AM modulation. The AM88 is phase locked loop (PLL) synthesized and crystal controlled for frequency stability and easy compatibility with both analog and digitally tuned receivers. The PLL synthesizer has a resolution of 1 kHz. The AM88 has a wide frequency range and may be operated anywhere (1-kHz steps) between 100 and 2000 kHz (150 kHz to 1710 kHz for best results). This range includes standard AM medium-wave broadcast (530-1710 kHz) and the long-wave AM broadcast band (150-285 kHz) used in Europe and Asia. Resolution of the PLL supports 10-kHz or 9-kHz channel spacing used in various areas of the world. Carrier-current operation is possible using the lower frequencies (generally less than 500 kHz). This allows signal distribution over AC power lines. Although the transmitter is basically AM, continuous wave (CW) operation (for Part 15 experimental 160-190 kHz license-free operation) is possible. FM operation is also possible for carrier-current applications by means of a simple PC board switch setting. The transmitter is designed for 100 mW, but RF output can be increased up to 1 watt for these applications.

Although the FM band is popular for applications such as this one, as well as for supporting stereo audio, for some applications the AM band might be better. Range can be better because higher field strengths are allowed. At 600 kHz, up to 400 uv/ meter is allowed at 10 feet. Receiver bandwidths are narrower, and AM is superior to FM under weak-signal conditions (below threshold). The AM band generally has more usable channels than the FM band, especially during daylight hours. Construction is less critical because frequencies are low, and only simple test equipment is needed to set up an AM transmitter. For some applications, such as the "talking house," whereby real estate brokers use small transmitters planted in houses so the prospective buyers can hear the sales pitch on the car radio, AM may be more useful because almost every car has an AM radio, but it may not have FM. (FM reception is still poor or limited in many areas in parts of the world.) A simple AM transmitter is also an excellent learning tool for beginners, who might find the very high frequencies and the added complexities of FM stereo a bit daunting as a first-time learning experience.

FIG. 1 Block Diagram of the AM88 Low Power AM Transmitter

FIG. 2 Schematic of the AM88 Transmitter

The AM88 uses four integrated circuit (IC) devices and nine transistors to implement a complete PLL-synthesized AM transmitter. The transmitter can be broken up into several subsystems: the audio amplifier, AM modulator, PLL frequency synthesizer, and RF output amplifier and filters. See the block diagram ( FIG. 1) and the schematic ( FIG. 2) for reference.

The audio section is made up of IC3a and IC3b, a dual op-amp LM1458N, and AM modulator Q6 and Q7. Line-level audio inputs of approximately 0.2-1.0 volt RMS (high impedance approximately 10K) connected to J1 is fed to a gain-con trolled R1 and a diode-controlled attenuator D1, D2, and R1. The diodes act as vari able resistances to small signals (50 mv or less). This is done to provide a means for automatic gain control of the audio section. Coupling capacitors C3 and C1 pass these signals to an audio amplifier op-amp circuit consisting of R3 and R4, and operational amplifier IC3a, with associated components R5, R6, C3, and C2. C2 limits frequency response to 10 kHz. R5, R6, and C3 provide a bias of half the supply volt age for the op amp, eliminating the need for a negative supply. The audio gain of this stage is nominally 20 times (26 dB), assuming that D1 and D2 are not conducting.

The audio from this stage is coupled to R7 and blocking capacitor C5 and to DIP switch S2A and S2B. This switch is used to route audio to either the AM modulator for normal AM operation or to the PLL circuit for FM. FM may seem an oddity at normal AM frequencies; this subject is discussed later.

The AM modulator consists of Q6 and Q7. The circuit is configured as a shunt feedback pair, and the bias point is set via R14, R15, and pot R16. The emitter of Q7 has a DC plus audio component and is used as a source of modulated DC for the RF output stage. The voltage at the emitter of Q7 sits around +5.0 DC volts. With audio drive from S2B , this voltage swings between less than 1 volt and to within 1 volt of the full supply voltage. Base drive resistors R12 and R13 form a split resistance to allow capacitive coupling of the modulator audio component to the junction of R12 and R13. Because the voltage across a capacitor cannot change instantly, a sufficiently large capacitor acts as a battery. This has the effect of maintaining a constant voltage across R13 and hence a constant drive current. This allows the base of Q6 to swing above the supply voltage by about 0.7 volt, ensuring that the emitter of Q7 can reach almost the full supply voltage. This technique, which is widely used in audio power amplifiers, is called bootstrapping. Because the RF output of the RF stage is proportional to the supply voltage, this technique allows full AM modulation of the RF output voltage to be achieved. The bias point is set to achieve symmetrical modulation with R16.

If the audio peak voltage was not limited, severe distortion would result from clip ping of the RF output peaks and cutoff of RF output on negative peaks. This is called overmodulation. To prevent this problem, a sample of the modulator output is taken through resistive divider R17, potentiometer R18, and R19. This voltage is compared with the op-amp reference voltage in comparator IC3B, and if it exceeds the bias level (half the supply voltage), then IC3 pin 7, which is normally at +2 volts, goes more positive. This is applied to resistors R9, R10, forward-biasing D3, and charging C4.

C4 is an audio bypass capacitor that prevents audio components from feeding back through the automatic gain control (AGC) network; C4 also determines the time constant for the AGC network. The DC bias developed across C4, if more than about 1.2 volts, forward biases D1 and D2, causing their dynamic impedance to drop sharply from nearly infinite down to as low as less than 100 ohms. This reaction causes the audio input present at the junction of R2, C1, D1, and D2 to be attenuated, reducing the modulation level. In practice, R18 is adjusted so that this occurs at 85-90 percent modulation. Although this scheme does not prevent negative clipping or deliberate overmodulation, it works well for normal application with speech or music.

The PLL synthesizer section uses an MC145151-2 LSI chip. This chip contains a reference oscillator, reference divider, charge pump phase detector, and variable divider that can be programmed for divide ratios of 3 to 16383. The reference divider is programmable, by hardwiring pins, at various fixed ratios that are mostly powers of two. In this application, it is set up to divide by 8192 so that a standard 8.192 MHz crystal will provide a reference frequency of 1 kHz. This sets the resolution of the synthesizer. The maximum input frequency that this chip can directly handle is approximately 12 MHz worst case. The AM broadcast band 530-1710 kHz has channels at 10-kHz spacing in the United States and Canada (this includes the newly expanded U.S. 1600-1710 kHz section) and most areas in the Americas. In other parts of the world, channels with 9-kHz spacing are used. Additionally, in Europe and parts of Asia, the long-wave band is used, with frequencies between 150 and 285 kHz, also at 9-kHz channel spacing. The synthesizer covers all of these frequencies, but in the interest of optimum synthesizer performance and considering cost limitations and circuit simplicity, the tuning range has been restricted to 2 MHz. The synthesizer supports all channels between 150 and 1710 kHz in 1-kHz increments.

Frequencies from as low as 50 kHz and as high as 2047 kHz can be generated, but the circuit component values-mainly in the filters and RF chokes-are not optimum at these extremes.

In order to cover these extremes, some changes in values of these components are necessary, which is beyond the scope of this discussion. In order to eliminate a noisy microprocessor and display, the frequency desired is set by using an onboard 12-section DIP switch S1. Eleven sections are used for programming, and the twelfth for some thing else. A binary code is used, the desired transmitter frequency is determined, and its binary equivalent is set using the DIP switches. Normally, once a clear channel is found or determined, the transmitter is set here and left alone. In most populated areas, relatively few clear channels are available, and especially at night when distant stations can be heard, a good, clear channel can be hard to find. In some areas they are rare, so, unlike a receiver, there is little need to reset the frequency often.

Additionally, most low-power AM units such as this one operate between 1600-1710 kHz or 525-535 kHz. Antenna efficiency is best at higher frequencies because any legal antenna used for U.S. FCC Part 15 applications is a very small fraction of a full-quarter or half-wave radiator. Therefore, a microprocessor and display would be overkill in this application.

Although direct generation of 150-1710 kHz frequencies can be achieved with the synthesizer, this is more than an 11:1 frequency ratio. A single-loop synthesizer will necessarily need a voltage-controlled oscillator (VCO) with this wide range, and the divide ratio needed in the variable divider will vary from 150 to 1710. It is difficult to control synthesizer loop performance to get reasonable behavior (settling time and damping ratio) over such a wide range; however, there is an easier way. The synthesizer chip IC1, a Motorola MC145151-2, has a programmable reference divider. If a standard 8.192-kHz microprocessor crystal is used, the reference divider may be programmed to divide by 8192. This will yield the desired 1-kHz reference frequency needed for 1-kHz frequency resolution. Because the chip can be programmed to divide by up to 16383, in binary, the variable divider section of the chip can be set up to divide by 8192 by permanently tying its most significant digit high. Then, by grounding the next two significant bits low, and using the remaining 11 bits in conjunction with a DIP switch, the divider can be made to divide by 8192 to 10239. This allows a frequency range of 8.192 to 10.239 MHz to be generated by the PLL synthesizer.

If we take this frequency range and mix it with the 8192-kHz reference oscillator signal using a mixer circuit, and a low-pass filter on its output, we end up with an output of theoretically DC to 2047 kHz. The DIP switch has to produce a binary code that is the binary equivalent of the output frequency. This task is simple, cheap, and allows a synthesizer loop design that has only a 1.25:1 range, which is easily done. The chip will directly handle these frequencies with no prescaler needed. Of course, a mixer and filter is needed, but this device is simple and straightforward and does not require any loop design compromises. Practically, because of the need for filters, coupling capacitors, and RF chokes in the transmitter, we use only the 150- 1710 kHz portion of the possible output frequency range, but this is not absolutely necessary, and with suitable components, the full range could be used if desired.

The circuit of the PLL synthesizer operates as follows: Q5, an MPF102 FET, is configured as an oscillator whose frequency is determined by L1 and the capacity of varactor diode D6, trimmer C28, and the circuit stray and FET input capacitance.

This oscillator operates in the 8.2-10.2 MHz range. R46 provides DC bias for Q5.

C53 RF grounds the anode of D6, which is fed a variable DC bias from R44 and R45. Depending on this DC bias and the setting of C28, this will be anywhere from 8.2-10.2 MHz. Oscillator signal is taken from the source of Q5. Part of this signal is passed to amplifier stages Q2 and Q4. R27 feeds signal to Q2, which is connected as a wideband feedback stage with R24 and R26 for feedback and bias. A signal large enough to drive the input of the variable divider section of IC1 (pin 11) appears at the collector of Q2. The divider is programmable via S1 to divide between 8192 and 10239, which is equal to the desired output frequency plus 8192.

For example, if a frequency of 1600 kHz is desired, then the divide ratio will be programmed as 1600 via the setting of S1. The internal variable divider will produce a signal to be fed to the phase detector at the transmitter frequency divided by 1600 because we have a division of 1600 times from the programmable divider. Mean while, the phase detector is fed a 1-kHz reference signal derived from an internal reference oscillator and divider, which uses external components R21, C12, X1 (8192 kHz crystal), and trimmer C11. These parts determine the oscillator frequency. C11 is used to set the frequency exactly to 8192 kHz. An internal divider divides this by 8192 and produces the 1-kHz reference. The output frequency accuracy depends on having an exact 1.000 kHz, which in turn needs an exact 8192-kHz crystal oscillator frequency. The phase detector generates a voltage, which depends on the relative phase difference between the reference and variable divider output waveforms.

For example, suppose the divider output starts to lag the reference. This implies that the divider, and hence the VCO frequency, is tending to go lower. In this case, the phase detector produces positive-going pulses and feeds these to the sample and hold network R38, C26, R39, R40, and C25. C25 is charged to a higher DC voltage.

IC2, a CA3420 CMOS op-amp, acts as a buffer amplifier for the PLL phase detector and provides a very high impedance for the sample and hold circuit, minimizing 1-kHz reference frequency sidebands and allowing smaller capacitors to be used in the compensation network. It also provides an easy method for injection of audio signals into the VCO for directly modulating the carrier frequency.

The high-impedance CMOS amplifier consisting of IC2, R41, R42, and R43 produces a positive-going output, which is fed to D6 via R44 and R45, causing the oscillator frequency to increase. The opposite happens if the VCO drifts higher, causing the divider output to lead the reference. Then D6 is biased with a negative-going change in DC bias and causes the VCO to lower its frequency. In this way, the VCO frequency is locked to the reference frequency and will not drift. It will be exactly equal, in kHz, to the programmed divide ratio plus 8192. In our case, we have programmed 1600 so that the output frequency will be 1600 + 8192, or 9792 kHz. Next, this frequency must have 8192 kHz subtracted from it to yield the final desired 1600 kHz output.

The final output frequency is obtained by mixing the PLL output 8192 to 10239 kHz (nominal) frequency with the 8192-kHz reference oscillator in a mixer circuit.

A sample of the 8192-kHz oscillator is tapped off IC1 via emitter-follower Q2 and divider R22 and R23. This signal is fed to the emitter of mixer stage Q3. Q3 is biased by R28, R25, and R28. L4 and C4 provide a low-impedance path for the desired difference frequency output. A sample of the VCO frequency is fed to buffer Q4 through R31 and is coupled to the base of mixer-transistor Q3 via C15. The input frequencies and their sum and difference frequencies appear at the collector of Q3.

The collector feeds low-pass filter (LPF) C16, L56, C17, L6, and C18. Only the difference frequency passes, the rest being rejected. DC bias is fed to Q3 via RF choke L3. R50 and C19 couple the difference frequency component to output amplifier stages Q8 and Q9.

The transmitter RF output signal that is to be modulated is produced by amplifying the mixer output from the LPF in Q8. Q8 is a conventional common emitter stage, with DC feedback biasing via R32, R33, and R34. C22 prevents AC feedback and preserves stage gain. R35 is a load resistor for Q8, and C24 couples signal-to-output amplifier stage Q9. R36 is a bias resistor, and Q9 is fed modulated collector supply voltage of around 5 volts from the emitter of Q7. The emitter of Q8 is connected to ground via a jumper, which can be opened in order to insert a key or keying circuit for CW (Morse code), commonly used in the 160-190 kHz Part 15 experimenters' band (also known as the 1750-meter band). C23 provides RF bypassing in this application.

Q9 was chosen to withstand operation into an open circuited load and to deliver up to 1 watt CW power at the lower frequencies (Part 15 160-190 kHz operation) or for carrier-current applications. It will deliver a 100-mW carrier at the higher frequencies (1600-1710 kHz), which requires 400 mW PEP output at full AM (100 percent) modulation.

In this stage, the signal is amplified to the final output level and then fed to a set of harmonic filters L7-L14 and C32-C47. The filters are low-pass, five-element Tschebychev types and should be designed to attenuate the second harmonic of the signal by at least 20-30 dB or better. A filter is useful only over about 65-90 percent of its cutoff frequency, so that the second harmonic is well into the stop band of the filter.

Hence, four filters are provided on the PC board, to cover the AM broadcast band and the 150-280 kHz range. S3 and S4 are DIP switches used for filter selection. Only one filter is used at a time. R20, bypass C9, and LED D5 are used as an output indicator, and LED D5 will not light if the switches S3 and S4 are inadvertently set to different filters. The LED will flicker slightly when modulation is present, therefore serving as a rudimentary modulation indicator. The filters shown are for 1200-1750 kHz, 800- 1200 kHz, 530-800 kHz, and 160-280 kHz operation. They may be changed by scaling the inductors and capacitor values inversely in proportion to the frequency ranges desired. RF output from the transmitter should be fed into a load of 50 ohms.

IC4, a 5-volt regulator, supplies 5-volt DC to IC1, the VCO, and the mixer circuits. Bypasses C20 and C21 ensure regulator stability. Capacitor C6 and diode D4 provide filtering and reverse-polarity protection of the 12-volt DC input, which may be from 11-16 volts in actual use. Excess of this voltage may cause damage, and less than 10 volts may produce poor results. Optimum power is 12-13.2 volts. Excessive noise on the DC supply line may cause this noise to be heard on the transmitted signal as interference and hum. It is normal for Q7 to get warm in operation, and if you prefer (not needed), a small clip on heatsink can be placed on Q7 to cool it.

The FM mode is useful for the low-frequency carrier-current operation that this transmitter is capable of. With a PLL, very little circuitry is necessary for FM, and this mode is obtained practically for free. Deviation up to 75 kHz is easily obtained, allowing much quieter carrier-current operation than AM can provide at the low frequencies used in carrier-current work. Typically, frequencies in the 100-300 kHz are used, but power line noise can be severe at these frequencies. Frequency modulation is accomplished by injecting audio from the audio amplifier IC3a into IC2. Instead of being returned to ground, R41 is fed from a preemphasis network R42, R49, and C30. This gives a preemphasis compatible with standard FM broadcast practice.

Potentiometer R48 sets the deviation. The audio across R42 is fed to IC2, and IC2 has unity gain for this audio. Therefore, an audio voltage is superimposed on the voltage to varactor D6.

Because the bandwidth of the synthesizer loop is less than 20 Hz, the relatively high audio frequencies are not "corrected out," and as long as no DC component is injected (assuming symmetrical FM, which is the usual case), the variations in frequency under-modulation are averaged out. The resultant modulation is clean and low in distortion because the VCO has a dynamic range of several volts, and a 1-volt change produces about 300-kHz frequency variation. Therefore, only about 250 mV peak audio (about 176 mV RMS) is needed for full modulation. This means about the same audio needed for 100 percent AM at the input to the transmitter will also be sufficient for the FM mode, assuming that R49 is set to about 80 percent of maximum.

The VCO is highly linear over such a small range, ensuring good-quality FM audio.

Note that section 12 of the programming switch S1 should be closed to disable the AM modulator and obtain full RF power in this mode; S2A must be closed; and S2B must be open to route the audio to the FM modulator circuitry.

As with any device coupled to the AC power line, RF should be fed into a suitable isolation network with components rated for the job. This means capacitors AC volt age rated at least twice the peak line voltage, and any coupling transformers used should withstand 1500 volts and preferably more. The transmitter wants to see a 50 ohm load, and suitable matching circuits and padding resistors are needed because most power lines have RF impedances less than this amount (see FIG. 3).

The AM88 needs an antenna. For many applications, a 56-ohm resistor shunted with a simple whip antenna as a radiator is adequate. The whip antenna should be only as long as needed, no more than 10 feet (3 meters), to avoid violating Part 15 FCC rules. (This issue is discussed later in the Appendix.) A PC layout is shown in Figures 5-4 and 5-5. Note that grounded leads of resistors are to be soldered on both sides of the board. This step is essential for good grounding. All parts are mounted tight and close to the board, except chokes. This is important for reducing audio noise pickup and for proper operation of the synthesizer and RF circuits. It also gives a professional appearance to the finished board. Begin construction by inserting all resistors in the PC board (see FIG. 4). Next solder all top ground connections. Install all diodes, carefully observing polarity. Next, install all capacitors. Make sure to observe the polarity of all electrolytic capacitors. Next, install the transistors. Be careful with Q7 and Q9 because these have an E-C-B pinout when viewed from the front side. Install trimmer C11, C28, and potentiometers R1, R16, R18, and R48. Preset C11 to 75 percent capacitance (plates 3/4 meshed). Preset C28 to halfway. The AM88 will operate well enough for setup with these initial settings. Install crystal X1, DIP switches S1, S2, S3, and S4, and the ICs, being extremely careful about correct orientation. If you wish to use low-profile DIP sockets, this is permissible.

Carefully check all work done so far for accuracy and orientation. Solder all bottom connections made so far. Carefully fabricate coil L1 and install it in the PC board. The toroidal core is wound with #24 enamelled wire (see FIG. 6). Make sure to connect the leads as shown, or the VCO will not operate. Use the toroid coil winding diagram as a guide. Leave an extra 1-2 inches of lead length on the lead of the 14-turn winding connected to C28, C29, and D6. An extra turn has been deliberately added on the primary winding to allow adjustment of inductance during setup.

FIG. 3 Power Line Coupler for Carrier Current Applications of the AM88

FIG. 4 Parts Layout of the AM88 Transmitter PC board-Component Side

FIG. 5 Component and Solder Side View of the AM88 PC Board.

FIG. 6 Winding and Mounting Details for L1 and L2 Through L14 of AM88 Coils

FIG. 7 The AM88 PLL Synthesized Low Power AM Transmitter in a Metal Case

FIG. 8 Top View of the AM88 in a Metal Case with the Cover Removed

Install RF choke L2, being careful not to bend the leads sharply too close to the choke body because this may damage the choke. Install L3, L4, L5, and L6 in the same manner. Next, install the remaining inductors as shown, standing them on end and bending the top lead down to fit the PC board (see FIG. 6). Again, carefully inspect all work so far. Look for solder shorts, poor joints, missing parts, incorrect parts placement, and so forth. You are ready to check out the board once everything is satisfactory.

To set up the AM88, you will need the following items:

  • · DVM or analog VOM, 20 K/volt or better
  • · Power supply 12-13.2 volts DC (9 AA or C or D cells in series is acceptable)
  • · AM broadcast or communications receiver
  • · CD player, tape deck, or other line-level audio source
  • · Patch cable for audio connections
  • · Test leads as necessary

The setup procedure is as follows:

1. Carefully inspect the PC board for shorts, missing or wrong parts, IC and transistor orientation, polarity of diodes and electrolytics, and any assembly mistakes, such as missing or poor solder connections. Make sure that the top traces supplying +12 volts to the audio section and +5 volts to the synthesizer circuitry are soldered to the component leads, passing through them, and that jumpers (J) between top and bottom traces are installed where necessary as indicated in FIG. 4.

2. Preset the following switches and controls:

S1: Positions 1 through 12 all "off "

S2: Position 1 (S2A) "off "

S2: Position 2 (S2B) "on"

S3, S4: Position 1 "on"

S3, S4: Positions 2, 3, 4 "off "

R1: 25% of full clockwise rotation

R16: 25%

R18: 75%

R49: 75%

C11: 50% full plates half meshed

C28: 100% plates completely meshed

3. Connect a 56-ohm 1/4-watt resistor between RF output and ground.

4. Connect the 12-volt power supply to D1 and ground; the negative lead of sup ply to ground; and positive to D1. Observe the current drawn; it should be about 50-200 ma. If it is appreciably less (less than 50 ma) or more (more than 250 ma), repeat step 1 because something may be wrong. Nothing should be getting hot, although Q7 will normally run quite warm after a few minutes. If still no errors are found and the current drain is lower than specified, there may be nothing wrong that will cause damage. In this case, proceed with setup and, eventually, any errors will be located. Excessive current and something overheating is a definite warning sign, and in this case the cause should be found before setup is completed.

5. Connect the negative lead of VOM or DVM to ground. Check the following voltages, using the positive lead of meter (12-volt supply is assumed):

Jct D4, C6, IC4: 11.4 volts

Pin 3 IC: 15.0 volts

Pin 7 IC2: 9.0 volts +/- 0.6 volts

Emitter of Q7 (TP3): 5.0 volts, varies with R16

Collector of Q9: Same as emitter of Q7

Collector of Q3: 4-5 volts

Collector of Q4: 0.5-1 volts

Drain Q5: 8.8 volts +/- 0.6 volts

Jct R5, R6, C3: 5.8 volts +/- 0.8 volts

Pins 1, 2, 3 IC3: 5.8 volts +/- 0.8 volts all the same

Pin 6 IC: 5.8 volts +/- 0.8 volts

Pin 7 IC1: varies with R16 and R18 from less than 2.5 to more than 7.5

Jct R8, R9, D3; varies with R16 and R18 from less than 1 to more than 1.5

A variation of 10 percent is normal. Remember to allow for meter accuracy and component and supply voltage variations. If any major variations are noted, repeat step 1. Reset any pots moved during testing to their original pre set positions, except set R16 for +4.5 to +5.0 volts at TP3.

6. Set S1 (dipswitch) for a frequency of 1700 kHz, within 20 kHz if 1700 kHz is busy in your area. (See the following chart for switch settings.) Leave section 12 of S1 OFF. Section 12 is used only for FM carrier-current work and is never used for frequency programming. Make sure that jumper J1 in the emitter of Q8 is installed. Read and understand the programming procedure.

7. Tune a nearby AM receiver to 1700 kHz, or as in step 4, if 1700 kHz is busy in your area. This should be the same frequency as in step 4. Monitor this channel with your AM receiver as you proceed.

8. Connect a DVM or VOM to TP1 (pin 6 of IC2). You should read almost 9 volts. If you see less than 9 but more than 2 volts, this may still be acceptable.

Listen on the AM receiver. Now start rotating C28 to disengage the plates. At some point, the voltage at TP1 should drop. If not, try removing a turn from the end of the 13-turn winding on L1 connected to C2If you initially saw less than 9 volts at TP1, this voltage should drop immediately upon rotating C2If the voltage is "stuck" low or will not reach as high as +7.5 (but will change with C28), then you should add a turn to L1. Set C28 for +7.5 volts at TP1.

This should occur with C28 set at 10-60 percent mesh. If C28 has to be set to more than 75 percent, add a turn to L1. You should hear a dead carrier (a signal without audio) in the AM receiver at this point. As a further test, disconnect or shut off the DC power. The carrier should simultaneously disappear; it should reappear when power is restored. This checks out the PLL synthesizer and mixer sections.

9. Remove DC power and program a frequency of 128 kHz on the DIP switches.

Set S3 and S4 four positions closed, all others open. Next, measure and record the voltage at TP1. This should be 2-4 volts. Verify that rotating C28 will affect this voltage. Reset C28 to get the voltage you just recorded. This checks out the synthesizer range. If the voltage seen at TP1 is too low and C28 has no effect, then add a turn to L1 and return to step If everything checks out so far, remove excess lead length from L1 and resolder it to the PC board. It is advisable to coat L1 and fasten it to the PC board with a clear lacquer base cement such as Duco cement or Q dope, or clear fingernail polish.

Do not use anything with pigment because it may degrade the coil. Hot melt glue is also acceptable for this purpose. After the coating dries and cools, recheck C28 as in step 6 and reset C28 for +7.5 volts at 1700 kHz if necessary.

10. Connect an audio source to the input and adjust R1 for the loudest signal in the receiver before any distortion is noted. Adjust R16 for 4.5-5 volts at TP3 if not done before. This sets the carrier. Then slightly increase the audio drive until distortion is evident. Adjust R18 to just eliminate this distortion. It should now be possible to increase the setting of R1 a little without experiencing much of a change in received audio level, although some compression may be noticed. This checks out the audio limiter circuit. If a scope is avail able, R16 and R18 can be adjusted for best modulation by observing the modulated carrier across a 56-ohm resistor connected across the RF output terminals.

11. If a scope or RF voltmeter is available, check the voltage across the 56-ohm resistor connected across the RF output terminals in step 3 to verify that the transmitter is producing RF output. About 2 volts rms across 56 ohm or 5.64 volts p-p will be present. This is best observed with an oscilloscope, but a detector probe on your DVM will suffice as a relative indicator. If you cannot perform this task because of lack of equipment, skip this step but confirm that LED D5 lights properly.

12. This completes testing and setup of the AM88. Remove power and audio connections and install the AM88 board in a case, making sure all adjustments are accessible. Set up for the final desired output frequency. See the programming instructions and following sections for more detailed information.

Frequencies and modulation modes are programmed into the AM88 by setting DIP switches either closed or open, in a pattern depending on the desired frequency and mode. Each time the frequency or mode is to be changed, these switches must be reset. At first, this process may seem inconvenient, but in practice, there are often only a few available clear channels in the AM broadcast band, and once set, the frequency will probably not be changed often. The mode settings will probably be rarely changed unless you are doing a lot of experimental work because more than 2000 frequencies can be programmed.

Four DIP switch assemblies and one soldered jumper are used for programming.

The soldered jumper is used to permit insertion of a key or a keying circuit for Morse code (CW) operation. If this will never be done, the jumper is installed permanently and left in place. DIP switch S2 and section 12 of DIP switch S1 is used for mode setting (AM, FM, or CW), and if only one of these modes is to be used, S2 can be permanently set or replaced with wire jumpers. S3 and S4 are used to select harmonic filters and are only changed when a large change in programmed frequency is made (see DIP Switch Settings).

Frequency is programmed by entering the desired frequency in binary on sections 1 through 11 of S1. The most significant digit is entered on section 1, the next on section 2, and the least significant digit on section 11. An easy way to get the binary number equivalent of a decimal number is either to look it up in a chart (books on computer science or math references provide these) or, if you have an IBM-compatible PC running Win 3.X or Win 95, use the Windows calculator that is provided in the accessories group. Use the scientific mode and the online help menus if this ability is not obvious. A chart of commonly used channels is included in these instructions, but it is impractical to list settings for all 2000 or more possible channels. You can also directly calculate the binary equivalent of any number by using the successive division by two method.

For routine Part 15 use, we recommend using as high a frequency as possible. It is a good idea to confine the signal to only the area needed. A 4-foot whip as a radiator will easily allow the signal to cover an average house and is mechanically easy to construct or salvage from a junked TV set. A 4- to 10-foot whip antenna or length of wire in parallel with a 56-ohm resistor may be connected to J2.

Parts List:

Resistors 1/4 W 5% Capacitors Potentiometers PT10YH2.5 Diodes Switches

Transistors Coils and Chokes

Integrated Circuits

DIP Switch Settings

MODE S1 posn 12 S2-1 S2-2 S3 and S4 posn 1-4

Filter Settings

1. S3 and S4 must be set to the same configurations.

2. Only one section each of S3 and S4 switch on at any time.

3. The LED (D5) must be lit, or S3, S4 are incorrectly set.

S3, S4: 1 ON; 2, 3, 4 OFF FLTR # 1 FREQ 1200-1800 kHz

S3, S4: 2 ON; 1, 3, 4 OFF FLTR # 2 FREQ 800-1200 kHz

S3, S4: 3 ON; 1, 2, 4 OFF FLTR # 3 FREQ 530-800 kHz

S3, S4: 4 ON; 1, 2, 3 OFF FLTR # 4 FREQ 150-280 kHz

For 280-530 kHz, either use filter #3 setting plus an external low-pass filter to cut off the second harmonic, or the values of L and C in one of the other filters may be changed as needed to suit the application.


DIP Switch S1 Frequency Settings

Important: Note that a zero (0) signifies that a switch section is ON, and a one (1) signifies that a switch section is OFF.

Suggested and test frequencies are listed. It is impractical to list all possible frequencies. See next section for a method of deriving settings for other unlisted frequencies.

Calculation of the Binary Code for Frequency

Any decimal number can be converted to binary by successively dividing the number by 2 and then separating the remainders from the result. This is best illustrated by an example rather than a rigorous mathematical description.

A number is either odd or even. An even number can be divided by 2 and has no remainder. Adding 1 to an even number results in an odd number and vice versa. For example:

8, 346, and 1500 are examples of even numbers.

9, 347, and 1501 are examples of odd numbers.

8 ÷ 2 = 4, 346 ÷ 2 = 173, and 1500 ÷ 2 = 750

An odd number will have a remainder:

9 ÷ 2 = 4 1/2, 347 ÷ 2 = 173 1/2, and 1501 ÷ 2 = 750 1/2

You divide the number to be converted by 2. If you have a remainder, place a 1 to the right of the result; if not, place a 0 to the right of the result. This signifies that you have (1) or do not have (0) a remainder. This is the least significant digit (LSD) of the needed binary number.

Discard the remainder (1/2) and repeat, continuing until you wind up with a 0. The ones and zeros in the right are the binary equivalent that you need. The final 1 or 0 that you place on the right is the most significant digit (MSD).

As an example, we want to convert 1585 to binary (1585 is odd):

1585 ÷ 2 = 792 + 1/2 remainder; therefore, place a 1 to the right 1

792 ÷ 2 = 396 no remainder; therefore, place a 0 to the right 0

396 ÷ 2 = 198 no remainder; therefore, place a 0 to the right 0

198 ÷ 2 = 99 no remainder; therefore, place a 0 to the right 0

99 ÷ 2 = 49 + 1/2 remainder; therefore, place a 1 to the right 1

49 ÷ 2 = 24 + 1/2 remainder; therefore, place a 1 to the right 1

24 ÷ 2 = 12 no remainder; therefore, place a 0 to the right 0

12 ÷ 2 = 6 no remainder; therefore, place a 0 to the right 0

6 ÷ 2 = 3 no remainder; therefore, place a 0 to the right 0

3 ÷ 2 = 1 + 1/2 remainder; therefore, place a 1 to the right 1

1 ÷ 2 = 0 + 1/2 remainder; therefore, place a 1 to the right 1

We have successively divided by 2 and have reached 0. The last digit is the most significant digit. The binary number is read out from the bottom of the column to the top and is:

1585 binary = 11000110001 (MSD to LSD)

If a number has less than 11 binary digits, place enough zeros to the left of the MSD to result in an 11-digit number. The LSD must be programmed into position 11 of S1.

Example: we want 188-kHz output. Convert 188 to binary:

188 ÷ 2 = 94 no remainder; therefore, place a 0 to the right 0

94 ÷ 2 = 47 no remainder; therefore, place a 0 to the right 0

47 ÷ 2 = 23 + 1/2 remainder; therefore, place a 1 to the right 1

23 ÷ 2 = 11 + 1/2 remainder; therefore, place a 1 to the right 1

11 ÷ 2 = 5 + 1/2 remainder; therefore, place a 1 to the right 1

5 ÷ 2 = 2 + 1/2 remainder; therefore, place a 1 to the right 1

2 ÷ 2 = 1 no remainder; therefore, place a 0 to the right 0

1 ÷ 2 = 0 + 1/2 remainder; therefore, place a 1 to the right 1

The result in binary is 10111100; however, this is 8 digits and we need 11 digits, so the number becomes 00010111100, with three extra zeros inserted on the left.

These zeros are programmed at S1 switch positions 1, 2, and 3 as "on." Position 4 will be "off," 5 will be "on," positions 6 through 9 will be "off," and 10 and 11 will be "on." This sequence is because the chip IC1 has internal pullup resistors on its logic inputs, and any given input line will be at a logic high (1) if the switch connected to that input is "off " (open) and a logic low (0) if the switch is "on" (closed).

A kit of parts for the AM88 transmitter consisting of a drilled and etched PC board, complete documentation, and all parts that mount on the board, is available from eBay:


Top of Page PREV.   NEXT   Guide Index HOME