Bidirectional Low Frequency Up-converter (Bi-LIF)
Introduction:
The concept of a Bi-LIF was developed during the building of the RX only DADP . During the testing process, the following question came up: is it possible to use a HAM radio strictly as a transponder and to modulate – demodulate entirely with software? Software driven RF test equipment is not new; this document describes how amateur radio operators can now build such a device.
Note: Before reading this document, it is important to first read the MDSR “white paper”. This document builds on the ‘RX only’ interface. All the technical specifications outlined in the MDSR document will be included in this design.
Objectives for the project:
· Low Cost
· Demodulation of AM,USB, LSB, CW, FM (NFM)
· Modulation USB, LSB
· The software runs on Windows based platforms
· The soundcard is used as AD - DA converter (duplex)
· All additional hardware has to be simple and repeatable by HAMs
· Requires a minimal amount of measurement equipment for testing and alignment
1. Technical Specifications
Hardware (RX):
Input Frequency 455kHz (+/- 5kHz)
Input Level (max) 1mV – 20mVpp (for S9 meter reading)
Conversion Gain (max) 55dB
Output Level for S-9 0.2Vpp
LIF Bandwidth: +/- 5kHz
LIF Frequency: 12kHz
Hardware (TX):
Input Frequency 12kHz (+/- 5kHz)
Input Level (max) 0.2Vpp
Conversion Gain (max) -3dB
Output Level 100mVpp
LIF Bandwidth: +/- 5kHz
IF Frequency: 455kHz
Software(RX):
AM Demodulation Bandwidth: +/- 5kHz
USB, LSB Demodulation Bandwidth: 300 – 2400Hz
CW Demodulation Bandwidth: 300 – 800Hz
FM Demodulation Bandwidth: +/- 3.5kHz
NFM Demodulation Bandwidth: +/- 2.4kHz
Software(TX):
USB, LSB Modulation Bandwidth: 300 – 2400Hz
CW: future option
Note: The explanation of the RX part of this design is omitted from this document since it is described in the MDSR (RX only) “white paper”.
2. Using the Transceiver as a Transponder
In order for the computer to connect to the radio, a Bi-LIF (Bidirectional Low Internal Frequency) converter is needed. It changes the radio’s IF into a 12kHz incoming and outgoing audio signal which is low enough in frequency to be able to be processed by the audio card. The described method will work with any 455kHz IF radio. The radio used during the testing and development was the YAESU FT-817.
The optional filter port is used as the transponder input and output. By setting the radio to option filter port during SSB mode and plugging in the Bi-LIF into the port, the radio is put into transponder mode. During this operation, the radio has to be set to USB modulation. This ensures that the frequency displayed on the radio is the actual RX-TX frequency. All operations, except the frequency stetting, the input attenuator, the pre-amplifier and the output power setting will be done by the computer.
Caution: It is important that the TX output power is monitored during setup to prevent overloading of the finals. The nominal power rating of the radio should not be exceeded.
(Disclaimer: Author will take no responsibility for damaged hardware as a result of implementation of this device.)
Option filter port extended to the rear of the radio (FT-817)
A small circuit board serves as a platform to solder the 0.075”
spacing strip connector as well as a 470pF capacitor. A small rubber cube glued on top of the PCB insulates and secures
the assembly when the lid is closed. The shielded cable is RG174 and the
connector on the rear wall of the radio is a SMA male, chosen for its small
size. This mod does not impact the portability of the unit nor does it
affect performance in any way. From here, it connects to the bidirectional
input of the Bi-LIF converter.
3. Conversion of 12kHz LIF (Low Intermediate Frequency)
Flow Diagram of the Bi-LIF
4. Circuit Description
TX Signal Pass
The microphone audio is amplified by +40dB by a modified HiFi AF preamp (Q1, Q2). P1 sets the amplification. The preamp is configured to only amplify 300 – 3500Hz. The signal is sent on the left line input channel to the audio card where it is digitized. The software in the computer combines the pseudo oscillator and the mic audio and creates the 12kHz digital IF signal. The data stream is sent to the sound card, converted into an analog signal and delivered to the right channel. From there, it is sent to the Bi-LIF where it is up-converted (U1) and filtered to 455kHz. P3 sets the level of the input and therefore also affects the overall gain of about -3dB (adjusted to below clipping with a 0.2Vss input). A driver circuit (Q3) amplifies the IF to a level the option filter input can process. The trim-pot (P2) on the base of Q3 is adjusted to center the IF signal. The transmitter circuit will accept this signal as a valid IF and process it the same way as if it were generated internally.
RX Signal Pass
The 455kHz signal is down-converted to 12kHz (U2), amplified (U3) and fed to the right line input channel of the audio card. The software in the computer demodulates the signal and sends it to the left output channel of the audio card. This channel feeds both sides of the headset (described in more detail in the ‘RX only’ MDSR document).
RX-TX Combiner
The option filter operates in RX mode as an output and in TX mode as an input. To accommodate that, a diode combiner circuit unites the RX and TX signals. L1 biases D1 and D2 to +5V which opens up an unidirectional signal pass when the cathode is grounded with a 10kOhm resistor, or closes when it is reverse biased with 12V. In order to switch the RX-TX pass, Q4 operates as an inverter enabling one or the other. In normal operation the transceiver provides the signal to set the Bi-LIF to receive (T7 pin 1 high) or transmit (T7 pin 1 low). The switch attenuation is about -50dB in off mode and -1dB in pass mode.
PTT and RS232 level converter
To switch the Bi-LIF from receive to transmit the MDSR sends a serial command to the radio telling it to change from RX to TX or vice versa. The option port of the radio provides a TLL level serial interface as well as the “GND during TX” signal. To be able to connect the serial port to the computer, the level has to be changed to match the RS232 voltages. This is done with U4. The connection to the computer enables a set of commands to be sent to the radio to remotely control it. For simplicity reasons and because the focus is on the software radio, further details on CAT have been omitted from this document, but can be found in the operations manual.
In TX mode, pin 1 of T7 is pulled to ground by the radio switching the Bi-LIF from down-converting into up-converting mode.
Local Oscillator 467kHz
To mix the 455kHz to 12kHz in RX mode a 467kHz oscillator signal is used. The mixer circuit U2 also provides a built-in oscillator. Cf2 is used to stabilize the frequency and the trim capacitor (C33) is used to adjust the frequency to +/- 5Hz. The same oscillator signal is also used to provide the mixing frequency for U1. Since the signal impedance on pin 6 on U2 is high, an additional amplifier (Q5) lowers the impedance before the signal is coupled tightly to pin 6 of U1.
Power Supply
U5 provides 5V for the up and down converter as well as U4. Additional filtering is provided by the PI filters for the converter section and U4. This prevents RF and spurious emissions to enter the rest of the circuit. The microphone preamp is fed directly with 12V. R10 decouples the preamp from the supply.
4. Bi-LIF Schematics
5. The Bi-LIF PCB Layout
This picture shows the prototype PCB without the metal bracket. The
layout has been slightly changed to accommodate small design changes and
some additional components. Added screws provide better grounding and radiation
proofing.
6. Building the Bi-LIF
The schematics of the converter may look intimidating and complicated but all components are available ‘off the shelf’ and there are no coils or transformers that need to be custom made. Furthermore, all the different functions can be built block by block and tested, relying on already confirmed blocks (i.e.: the power supply is assembled first and functionality is tested. Next, the RS232 converter is assembled and tested and so on until the unit is complete). The layout has to be followed very closely. The blue lines are insulated wire connections. All the grounding points (green dots) must be exactly as described in the layout assembly drawing.
The PCB is mounted on an L bracket
that shields and
provides ground
through the 7 mounting screws. The PCB is a double sided 0.1” spacing multi-hole
project card (left). The layout requires both layers; a single-sided board will
not work. The jumpers on the PCB have to be floated above the surface of the
top layer to prevent shortening to ground. This PCB should be widely available.
Mechanical work on the PCB
Before assembling the PCB, all the mechanical work has to be completed. The PCB is 5.2” (132mm) by 2” (51mm). Drill all the holes with a 3mm (1/8”) drill into the PCB first (all holes are on a 0.05” grid) and then use the PCB as a template for the bracket. Make sure the PCB mounted DB9 connector fits the holes in the bracket before drilling (if the DB9 connecter is not exactly the one used in the prototype, adjustments will need to be made). If all mechanical work is completed prior to populating the board, the final assembly will be easier.
Power Supply Section
Components: C48 to C51, U5, D5 – D7, T5, R30, R31,
Test: Apply 12VDC to TB5 and measure voltage on C48. If the meter reads 5V and the power LED light up the circuit is working.
RS232 Level Converter
Components: L3, C52 – C58, T7, T8, U4
The pins 6 to 9 of TB8 are not on the 0.1” grid, so they have to be cut off before the connector is mounted.
Test: Apply +12V to TB5. Measure the voltage on U4 pin 2: +9V, U4 pin 6: -9V
Microphone Amplifier:
Components: R1 – R10, C1 – C14, TB1 – TB4, Q1, Q2
C1 and C2 solder right onto the microphone connector and are terminated right at the front plate which is grounded through a 10 gauge (1.5mm2) grounding cable to the ground lug.
Test: connect +12V to R10 and apply a 20mV 1kHz sine wave to the microphone input. Turn P1 until the amplification is set to maximum. Measure with an oscilloscope on the ring of TB3 and confirm that there is an amplified (+40dB) sine wave present. Put 1Vss 1kHz signal on the tip of T3 and measure the audio output on the headphone output.
RF Combiner Circuit
Components: L1, C15 – C17, C28 to C30, D1, D2, Q4, R11 – R13, R19, R20, T9
Solder the wire that supplies 12V to the mic amp from the power supply.
Test: Apply +12V to TB5 and measure the voltage on the anode of D1 (approx 5V). Measure the voltage on the cathode of D2 (approx 4.4V). Ground T7 pin 1 and measure approx 9V.
The cathode of D1 it is exactly the opposite, approx 9V in RX and 4.4V in TX mode.
RX Down Converter
Components; R21 – R29, C31 – C47, P4, L2, U2, U3, Q5, Cf2, Cf3,
When soldering U2, do not use an IC socket. While soldering the oscillator circuit, be careful not to use too much solder. This will keep the stray capacitance low and the capacitor C34 should be left out. If the frequency is too high and does not tune to 467kHz with C33, a appropriate value can be added in later.
Testing: With the oscilloscope, measure U2 pin 6. There should be a sine wave of 0.2Vpp and a frequency of about 467kHz. Now measure the emitter of Q5; again the voltage should measure 0.2Vpp. If using a frequency counter, attach it to the emitter of Q5 and measure the frequency. Tune the frequency with a tuning fork (do not use a screw driver) to 467kHz (+/-5Hz). Now apply a 455kHz 1mV signal or the output from the option filter port of the radio to TB9 and measure the signal on U3 pin6 with a oscilloscope. There should be a down converted version of the IF (represented in amplitude and frequency) present.
TX Up Converter
Components: C17 – C26, R14 – R18, U1, P3, Q3, Cf1,
When soldering U1 do not use a socket.
Testing: With an oscilloscope measure U1 pin 6 and confirm that you have a 467kHz 0.2V sine wave present. Apply a 12kHz (+/-3kHz) 0.2Vpp sine wave to the ring of TB4. Attach an oscilloscope to TB9 and ground pin 1 of TB7 (setting the Bi-LIF into up-converting mode). A 455kHz sine wave must be seen. Adjust the amplitude with P3 to 0.1Vpp. Now center the sine wave with P2 so that no clipping or the same amount of clipping occurs on the positive and the negative rail of the sine wave.
This completes the assembly of the PCB and the preliminary function testing.
7. Mechanical Assembly
The bracket that holds the PCB is absolutely necessary to provide proper grounding. Aluminum with about 1/16” (1.5mm) thickness or a prefabricated L-profile is ideal. The hole sizes for the connectors in the back is dependant on the type of connectors available. Pre-drill with a 3/32” (2mm) bit and then use a uni-bit to expand the hole to the right size. A uni-bit is fast and creates a nice, clean, round hole. The PCB is mounted with 7 M3x20mm screws and 5mm metal stand-offs. All measurements are in decimal inches.
8. Required Cables and Connections
RS232 Cable
The RS232 cable is standard straight through 3 lead (RX,TX,GND) with a male DB9 on one end and a female on the other. For computers without an RS232 port, a USB converter can be used. The default setting for the port in the software is COM1. Connect the computer com port to the upper female DB9 (TB7) connector of the Bi-LIF. Specific port settings are set by the software automatically. The radio has to be set to 38400bd.
Accessory Cable
The accessory cable has to provide “GND on TX” to switch the Bi-LIF from the down-converter into the up-converter mode. It also connects the TTL level com port (RX, TX) to the radio. It has a male DB9 on the Bi-LIF side and a connector that plugs into the auxiliary port of the radio. See schematics for details on the pin-out. When the radio is keyed up with the mic PTT button, the red LED on the Bi-LIF must come on.
Audio Card Cables (Line in, Line out)
Two standard 3.5mm stereo shielded phono cables with a maximum length of 2m should be used; the shielding of the cables is absolutely necessary. The cables are crossed-over; Line-out from the Bi-LIF connects to the Line-in of the audio card and vice versa. For better RF-shielding take a #10 gauge (1.5mm) copper grounding cable and connect the chassis of the computer to the station ground. This will actually lower any interference from the computer.
Option Filter IF in/output
This cable connects the option filter port to the combiner circuit of the Bi-LIF. As previously described, an SMA male connector has been installed into the radio. Use 6” (20cm) of shielded RG174 cable with a female SMA and a male BNC connector on the Bi-LIF side.
Ground
The Bi-LIF has to be grounded to the station ground for best performance with a short #10 gauge (1.5mm) copper cable.
9. Headset and Microphone
Using a computer headset with a built in microphone is possible, but not all headsets will be suitable for modification. Access to the microphone and the earphone solder connections is necessary. One of the greatest issues with using a computer headset is the low quality of the cable and connectors. Because changing out the cable can be very time consuming, it is worthwhile paying for a higher quality headset (with a higher quality cable). To be able to shield the RF and avoid microphone interference, a higher quality cable will provide better results.
The following steps need to be taken:
· Solder a 4.7nF SMD capacitor across the mic capsule
· Solder a 4.7nF ceramic chip capacitor across each of the earphones
· Set the microphone to mono (tip and ring wire are connected together)
· Cut off the 3.5mm phono connectors because they do not provide appropriate ground to properly shield the RF from the headset. A better 4 pin radio microphone plug with screw-on lock is required and has to be soldered onto the headset cable instead.
10. Final Tuning and Testing of the Bi-LIF
The quality of a 16 bit audio card is more than enough to provide a clean and functional digital modulator and demodulator. The challenge in using a sound card as an analog to digital converter is the software interface that Windows provides. In Windows, under audio devices, the left and the right channel of the input and output are required to have a similar audio level. The design of the Bi-LIF takes this into account and the tuning of each channel has to achieve this balance. The aim is to get the best audio quality and sensitivity; it is not designed to provide absolute levels and test equipment accuracy.
Setting up the Audio Properties in the Control Panel
The audio settings in the control panel have to route the audio signals through the modulator-demodulator software (MDSR). The settings may vary from sound card to sound card. A little experimenting will bring the best results. Avoid audio loops which will create an echo-like effect.
The following steps need to be taken:
· In recording devices, select “Line In” or “Analog Mix”; turn the volume to 70%
· In playback devices, select “Line In”, “Wave” and “Play Control”; set all to 70%
· In the MDSR, select the sound card icon (just below options) and select your sound card, set the volume to 70%
· Click on the green arrow to start the MDSR; the receivers’ static noise should now be heard in the headset
Before proceeding please repeat the assembly tests as described in section 6.
Testing and Tuning of the Receiver
Two settings are important here, the frequency and the S9 input level. Set the MDSR to AM, BFO = 0Hz, Hi Pass = 300Hz, Low Pass = 2400Hz, IF = 100%, Volume = 20%, Squelch = 0%, AGC = “on” and let the Bi-LIF warm up for at least 30min.
The following steps need to be taken:
· Tune the transceiver to 10MHz with antenna attached
· In the spectrum window the signal from WWVH should be seen. Tune with a
tuning fork C33 until the 10MHz carrier is in the centre of the display
· Tune the transceiver to a weak AM station with a known signal strength, match
the reading on the RX signal meter in the MDSR by adjusting P4
Testing and Setup of the Transmitter
The important setting here is the 0db carrier output. Connect a wattmeter and a dummy load to the RF output of the transceiver. Set the MDSR RF Gain and the Mic Gain to 100%.
The following steps need to be taken:
· Turn the “Carrier” button to ON
· Press the “PTT” button, the radio will start transmitting
· Adjust P3 starting from a lower reading until the nominal power output is
reached
· Turn off the carrier by deselecting the carrier button (no output)
· Talk into the microphone and the output level should change; the reading on the
“TX Audio” meter in the MDSR should also follow the level of the voice
· Clear TX by releasing the “PTT” button
This concludes the setup of the Bi-LIF.
11. Description of the TX software
The MDSR builds on the software and extends the capabilities of the “RX only” MDSR to a fully functional transceiver interface. The MDSR completes the development in replacing hardware based modulation and demodulation.
The graphic user interface appeals to radio amateurs and the functionality of the controls matches those in the hardware counterparts. The TX component may seem simple, but with the support of computer based audio processing, it rivals some of the most expensive commercial transceivers in audio clarity and talk power.
MDSR TX Software Flow Diagram
a.) AF Processing
The pre-amplified signal from the Bi-LIF is digitized and then sent on to the TX signal processor of the MDSR. A volume control adjusts the proper audio level for the user. The compressor unit has two settings: ‘off’ and the signal is only clipped if it exceeds 1.5Vpp, ‘on’ and the signal is amplified (scaled) with an un-linear interpolation graph. This curve (graph) causes small amplitudes (>0.5Vpp) to be amplified and large amplitudes (<1.2Vpp) to be attenuated, never allowing signals to go beyond 1.5Vpp. This adds additional spectral lines into the audio spectrum before it gets modulated. More spectral lines within the transmission raise the signal level without increasing output power.
In order to meet the specifications of the SSB signal (300-2400Hz) an additional voice filter is added, removing any unwanted frequencies.
The pre-emphasis raises the high end of the voice band to give a crystal clear audio effect that makes the transmission easier to detect and to understand.
b.) Modulator and Filter
The pre-processed microphone signal is multiplied with the internal oscillator and a DSB signal results. Two individual side band filters assure a suppression of the unwanted sideband of at least -80dB. The filter settings are controlled by the RX demodulation logic. In the TX RF Gain module the RF-gain of the 12kHz LIF is user adjustable which allows the user to set the total output power of the radio. Also in this module, the carrier is added to the LIF if the carrier button is set to on. This is useful when a manual antenna tuner is used.
c.) Output Monitor
The microphone audio level is displayed on the TX Audio meter to allow the user to adjust the best level before the audio is sent to the modulator. This also affects the RF output. Too much gain would overload the mixer and unwanted frequencies would result. In the audio processing, this is prevented by limiting the output.
d.) Control
The slide controls control the microphone gain as well as the TX audio level. Button controls are used to activate the DX audio compressor, the PTT and the carrier “on” button for testing and tuning. The PTT is activated by sending an RS232 command to the transceiver.
12. The MDSR Graphic User Interface
The MDSR Graphic User Interface offers simple, big and user friendly controls. It also features a spectrum analyzer to enhance the tuning in RX mode and an oscilloscope to display all the AF signals of the RX and TX side. The RX s-meter has a resettable maximum hold indicator.
The interface is clean and simple because most of the work is done by the computer. This gives the user the freedom to concentrate on the QSO and not on the setup of the radio. However, the user maintains full control of the outgoing and incoming signals.
13. Conclusion
The computer (PC) has proven to be an excellent choice for this project. The audio processing of a sound card is superior to any hardware. The ability to suppress a carrier with a simple line of code or to provide a mixer by just multiplying two signals together is amazing. Using the computer clock to synchronize all oscillators in the MDSR is simple and superior when compared to complicated hardware designs. The steepness and flexibility of digital filters is impossible to achieve in hardware designs alone.
The software radio is here to stay and will eventually take over the market, replacing conventional transceivers. What is demonstrated here is only a very small portion of the possibilities that can be programmed into this setup.
I will offer technical support to any HAM who builds this Bi-LIF. Software will be available as a download on this website.
About the Author
Alex Schwarz (VE7DXW) is an advanced HAM and a graduate of the HTL, Innsbruck, Austria. He moved to Vancouver (Canada) in 1990 and has since been involved in professional communication systems (LDR trunking) and digital point to point wireless network systems. In 2005 he started work in the Biomedical Engineering Department at C&W Hospital in Vancouver. He can be reached through his email address: alexschwarz@telus.net.
MDSR Kit Building de AH6EZ