(C) 2013, G. Forrest Cook
This project involves building an isolated audio interface between a Kenwood TS-430s ham radio transceiver and a PC sound card. The interface includes automated push-to-talk (PTT) circuitry. It allows the TS-430s to be used for transmitting and receiving PSK31, RTTY, Olivia, JT65, SSTV and more. The open-source program Fldigi has been used for working with PSK31, RTTY and Olivia. This interface is an expansion of the example circuits shown in the Fldigi documentation.
The Kenwood TS-430s requires a simple modification involving the routing of the line-level audio output (known as av3) to an unused pin on the microphone connector. This mod provides two advantages. First, only one connector is required between the radio and the interface box. Second, the line out signal is tapped before the volume control so the radio's audio output can be varied or shut completely off without changing the levels going into the computer.
This interface should work with many other ham radio transceivers, it is advisable to work with a unit that has a stable VFO such as a PLL or DDS design.
The transmit signal comes from the sound card output's left channel (tip to sleeve) and is fed to a 600:600 ohm isolation transformer. The level is reduced via a 2.7K resistor and the 1K tx level trimmer, then fed to the TS-430s microphone input via the PC/Mic Select switch. The TX signal monitor jack can be used for monitoring the sound card output, either with an oscilloscope or an audio amplifier and speaker.
The push-to-talk signal comes from the sound card output's right channel (ring to sleeve). Fldigi is configured to send a sine wave to this channel when it wants to transmit. The PTT signal is fed to a 600ct:600ct ohm isolation transformer that is wired as a step-up device. The output of the transformer drives a voltage doubler/rectifier circuit that produces the gate drive for an IRF521 MOSFET transistor The IRF521 is used to switch the transceiver's PTT control line.
The PTT hang time control is a SPDT center off switch, it adds either zero, 500nF or 1uF across the 10nF gate delay capacitor for three values of PTT hang-on delay. The gate signal is limited to 12V by the 1N5242 diode and the 750K resistor drains the gate delay capacitor when the push-to-talk wave shuts off. The PTT signal is routed to the TS-430s via the Select switch.
The receive signal comes from the high side of the TS-430s volume control, see the TS-430s modification below. The signal is amplified by the 2N3904 transistor, which is configured as a class A amplifier. This stage also acts as a buffer so that the TS-430s audio line is not loaded down by the interface. The 2K gain trimmer and associated parts on the emitter of the 2N3904 provides a gain adjustment control. The resistors on the rx gain control circuit were chosen so that the midpoint setting gives a useful level.
The output of the receive signal amp drives a 600:600 ohm isolation transformer which has a 33nF capacitor across the primary. The transformer/capacitor acts like a low-Q bandpass filter which has a peak response at 1Khz and is 3dB down at 10Khz. This filtering action removes unwanted high and low frequency noise from the signal. The output of the receive signal transformer is attenuated by the rx level control before it is sent to the right and left inputs of the sound card.
The circuit was built into a 2"x4"x6" aluminum project box. One shielded multi-pin wire was used for connecting to the Kenwood TS-430s microphone connector and two stereo shelded wires were used for the PC sound card input and output wires. The three wires were rounted into the project box through strain relief grommets and were soldered directly into the circuit. Direct soldering eliminates several connectors in the signal path for better reliability.
All of the interface circuitry was soldered onto scrap pieces of blank circuit board material in the typical "dead bug" style. Perforated circuit board material could also be used. Most of the parts used in this project were scrounged from the author's junk box. Note that the internal photograph was taken before all of the circuitry was added.
The shield connections should be done carefully to avoid pickup of stray RF energy from the transmitter. First, all of the external wiring was done with foil-shielded wire. Beldfoil is a common wire type made by Belden. A video cable from an old CRT computer monitor can be a good source of multi-conductor shielded wire. The shield on the microphone cable is connected to the shield pin on the connector that plugs into the TS-430s mic input. The other end of the shield connects to the interface box and the interface circuit board ground.
The shields on the two sound card connectors are completely isolated from the transceiver shield, they only connect to one side of the three isolation transformers. This isolation prevents ground loop problems which can introduce hum and RF noise into the signals.
The push-to-talk circuit ground is also completely isolated from the microphone circuit ground, it was built onto a separate piece of circuit board material. This is probably unnecessary since the mic ground and PTT ground are connected together in the transceiver, but keeping the two circuits separate could be useful if the circuit is used with a different transceiver. Essentially, the PTT circuit acts like a completely isolated switch.
|Coax added to TS-430s Rf Gain/Volume board (gnd/av3)||Coax and ferrite bead to TS-430s Mic input board unused pin|
A simple modification was performed on the Kenwood TS-430s to bring the line-level audio output to the microphone jack. The high side of the volume control signal (av3 on the schematic) was routed to the unused pin on the microphone connector via a short piece of miniature coaxial cable. The coax shield is only connected on the volume control side and a ferrite bead was placed on the line at the microphone connector for RF isolation. The modification will not affect normal microphone operation since the pin is otherwise unused.
If your transceiver has an SSB narrow filter installed, it should be activated when using BPSK31 and other digital modes. If your transceiver has a CW narrow filter, there is a fairly simple modification for swapping the CW and SSB filter control lines so that the narrow CW filter can be used in SSB mode. This modification is essential for serious BPSK31 operation. With the CW filter activated, it is much easier to decode very weak signals or isolate a single signal in a crowded spectrum.
To perform the filter swap mod, locate connector #27 on the IF Unit board, it is next to the optional crystal filters (large silver rectangular boxes) which should be installed in the transceiver. Cut the SSN and CWN control lines (white-red and white-blue) in the wire bundle leading to connector #27 and connect the wires to a piece of miniature four-wire shielded cable. An old computer mouse is a good source for this kind of wire. The cable shield should be connected to ground on only one side. Heat shrink tubing should be used on all junctions to prevent short circuits. Wire the contacts of a miniature 12V DPDT switch to form a polarity reverser and connect the four control wires to the switch.
If the switch is wired correctly, the TS-430s will behave normally when the switch is in one position. When the switch is in the other position, the narrow filter switch on the TS430s front panel will select the CW filter when in SSB mode. The filter switch can be mounted on the back of the transceiver using a small L shaped bracket that is attached to the large heat sink.
The fan on the TS-430s RF output transistors has a known bug where it does not turn on until the finals become very hot. The heating problem is especially critical in the high duty cycle digital modes such as PSK31 and RTTY. This issue is well documented on the Internet, the recommended fix involves changing R36 on the filter board from 3.3K to 4.7K to lower the turn-on temperature.
Some articles suggest removing the entire RF output section and filter board to change the resistor. I was able to unscrew the RF output unit and do the mod on the mounted filter board without performing so much disassembly. One lead of the original R36 was clipped and a 1.5K resistor was carefully soldered in series between the remaining resistor and the resistor's old lead to form a 4.8K resistor. This is a tricky job that should only be attempted by someone with a lot of soldering experience.
Fldigi, or whatever software you use on the computer should be configured to output a push-to-talk sine wave on the sound card's right audio channel during transmit.
The PC's sound card mixer adjustments need to be set up. The sound card output level (typically the master and PCM controls) should be set once and left alone. Connect the transmitter's RF output to a dummy load during this process. An oscilloscope should be connected to the TX Signal Monitor jack and a psk31 waveform should be transmitted. The sound card levels should be set so that the signal on the oscilloscope is about 10% below the point of clipping (flat-topping). This level may vary between different programs, so be sure to monitor the sound card output with all of the software that you intend to use. Set the TS-430s mic input level to 50% and adjust the tx level trimmer so that the ALC meter is at the bottom of the ALC range.
Note that the sound card's output level also controls the amount of gate drive on the PTT circuit. If you send too low of a level out, the IRF521 gate drive voltage may drop below the activation level and the transmitter will shut off. A logic-level gate MOSFET such as an IRLZ44N can be substituted for the IRF521 if the sound card's maximum output level is limited.
The rx gain can be set to about 80% full and left alone. The rx level control is mounted on the front of the box. The rx level and sound card aux input level should be adjusted for a good decoding level while watching the waterfall display on Fldigi. Keep in mind that the signal and noise levels can vary widely depending on which band you are using, the band conditions and the transceiver filter settings. See the Fldigi documentation for more details on setting the levels correctly.
Connect the interface between the computer's sound card and the TS-430s microphone input. Tune into some signals and do some test monitoring to verify that the signals are being decoded correctly. I wrote a command-line Python script called MixerSet.py that can be run to set the initial mixer values and enable capture on the aux input. This will only be useful if you are running Linux (Ubuntu 10.10 in my case) and it will need to be modified to work with your particular sound card. MixerSet.py requires installation of the amixer command line mixer package.
The sound card's output level should be set to the sweet spot explained above, then left alone. The TS-430s mic input control can be used to fine-tune the transmit level. The rule of thumb is to adjust the transmit output power to below 70-80% of full power.
The mic input level should be set so that the modulation level is below the point where the transmitter's ALC circuit activates. Too much modulation will produce distortion and make your signal hard to decode, it will also cause excess heating of the RF power transistors and interference to other stations. If you watch the transmitted signal level on an SWR meter, the meter will reach a steady level when the ALC circuit activates. When running PSK31, I back off on the mic level until the ALC meter is somewhat below the ALC bar and the SWR meter wiggles more visibly.
The PTT Hang time can be adjusted for different operating modes. A fast hang time allows quick T/R switching for PSK31 and a slow hang time is useful for keeping the PTT line active during tailing CW identification tones such as those generated by QSSTV. The medium hang time setting was included for experimental purposes, it can be eliminated if you like.
Your author has made many successful PSK31, PSK63 and Olivia contacts with this circuit. It is a good idea to request signal to noise and modulation data from the remote station to make sure everything is aligned correctly. The interface has also been used for monitoring JT65 signals with success.
The interface has also been used to send and receive SSTV signals using QSSTV. Unlike PSK31, SSTV operation usually requires the operator to switch the input select switch between PC and Mic in order to use both voice and SSTV. The PC/Mic select switch makes this type of operation easy.
Some users may want to generate a CW keying signal for operating the transceiver's key line. This method generally produces a cleaner CW signal than sending a keyed audio signal through an SSB modulation system. Direct CW keying can be performed with an additional voltage doubler/MOSFET circuit or with a level shifter/current amplifier circuit between the PC's serial or parallel port and the keying line. The Fldigi documentation mentions that CW keying which is driven by an audio tone may produce better timing results than keying via a serial/parallel port handshake line.
Back to FC's Ham Radio Circuits page.