R&S GB610 P1 avionics radio panel
I found a aircraft radio panel at a local (Zutphen) dumpstore. The panels were brand new and still in it's original packaging. The price was right so I bought a couple of panels. The control panels are supposed to be used as a control panel for a UHF communication transceiver. The frequency can be set between 200.000 and 399,975 MHz in 25 kHz steps. The image of the webshop advertisement is shown below.
After a little 'research' I found that the GB610 P1 control unit part number 614.3119.02 / 5865-12-199-8447 was a part of the XD 613 P1 transceiver. The transceiver consists of several modules as mentioned below. Behind the control panel is a synthesiser module, a receiver module, a transmitter module and a RF power amplifier. The part numbers are menioned below.
UHF - SENDE - EMPF. - GERÄT
XD 613 P1 - 614.2312.03
BEDIENTEIL . CONTROL UNIT - GB610 P1 - 614.3119.02
UHF-SYNTHESIZER GF611U3 - 701.6009.02
UHF-HAUPTENMPFÄNGER - MAIN REC. - EO610 H1 - 679.5000.12(?)
UHF-SENDERTEIL 10W - TRANSMITTER - VD 610 P2 653.8318.02
UHF-VERSTÄRKER 30W - UHF-AMPLIFIER 30W - VD613 - 614.4015.03
Unfortunately I have no other parts than the control panel nor documentation, so the device description is based on an educated guess.
The switch at the left bottom corner is probably the power switch. The switch has three positions: off, on with low transmit power and on with high transmit power. At the right is a toggle switch for (de)activating the audio squelch. The [VOL] know is probably used to set the audio volume. The [TONE] push button is likely for sending an audio tone. The two position [MAN/PRE] switch at the right is likely used for mode selection between manual [MAN] frequency selection and selecting one of the 30 (0...29) preset [PRE] frequency channels. There's also a hinged bracket visible so the transceiver can be pulled out of the panel.
The middle part of the control panel is for setting the desired frequency by five knobs. The most left know can be switched between 2 and 3. (After removing a locking bracket, '1' becomes also available.) The middle three numbers can be set eacht between 0 and 9. The most right knob is a four position switch for 00, 25, 50 an 75 kHz steps. So a desires frequency between 200.00...399.975 MHz can be selected. This results in almost 8.000 possible frequencies.
New use for the control panel
I contacted Rohde & Schwarz for technical information, but (as expected) I never received a reply nor information. R&S is a company with professional equipment and professional customers. Therefore don't expect any service or technical information as a non professional user. (Note: HP does share all the needed information like schematics...) Since the schematic is never to be found and reverse engineering is too complex for my intended use, I discarded the used printed circuit board and I planned to design a replacement board.
The idea: Controlling a commercial radio with this control panel
The idea is rather simple. I want to use the control panel to control a commercial transceiver like a Icom IC-705, Yaesu FT-897 or an Kenwood TH-D7E. I want to convert the button/switch information to CI-V, CAT or RS232 information. These commercial radio's can receive airband frequencies. When the switch information is converted to the corresponding rig command, the control panel controls the connected radio receiver. By expanding the frequency settings, other (ham)radio frequencies can be set if desired.
If the switch information is known by the (Arduino) controller, it's 'just' programming. The hardest part is to redesign the electronics. So let's focus on the hardware first. My design strategy is shown below.
A microcontroller can 'read' a high (1) or low (0) signal applied on an input pin as shown below. If 5 VDC is applied to a button/switch, the other side is 0V/low of the button/switch is open or the signal is 5V/high is the button/switch is closed. So the controller 'knows' the button/switch state and the desired command is sent to the transeiver.
Since the control panel has more than one button/switch, more inputs of the microcontroller are needed. This situation is shown below. The advantage is that it's rather simple. The disadvantage is that the input pins of a microcontroller is limited. Since more than 40 wires are needed, a microcontroller with a huge amount of input pins is needed. This crude approach is simple, but not the best way for this situation.
A more 'real world' approach is shown below. An eight bit (74HC595) shift register is used to expand the input pins. A shift register uses three main signals. A high/low data [D] signal, a high/low clock signal [CL] and a high/low latch signal [L]. By applying a clock pulse at the CL input, the data bit is shifted into the register. If the data pin is low, a 0 is shifted into the register at the clock pulse. If the data pins is high, a 1 is shifted into the first position of the register at the clock pulse. There are eight bit positions available and therefore eight bits (one byte) can be set from 0000.0000 to 1111.1111. The shift register outputs are still all unchanged though. By sending a pulse to the latch pin, all the bits in the register are 'sent' to the outputs. This makes is possible to apply 5V to the connected switches.. The other side of the switches are all 'tied' together and goes to the controller input. If for example 0100.0000 is sent to the shift register outputs, there's only 5V applied to the switch at position 2. This results in 0V at the 'sensing' pin; thus the switch is not in the second position. If 1000.0000 is sent though, the 5V signal is flowing through switch position one and the sensing pin is high. To 'know' all the switch positions, each of the eight outputs has to be set high momentarily and simultaneously 'read' by the controller. This 'reading' of the button/switch positions can be done very fast and continuous. And 'only' four pins are used to find out eight switch/button positions. Beware that shift register outputs are 0V or 5V if not floating. Forget the three state floating possibility for now and assume that the state is 0V or 5V. If two buttons are pressed together, one of the register output pins is set high (this the other low). There's a short circuit since the 5V flows trough both buttons to the 0V output pin thus damaging the chip. Therefore it's mandatory to use the drawn diodes so no current can flow to the output pins...
The next and final step is to use more shift registers. In theory there's no limitation to the number of cascaded shift registers. But each output requires an extra pulse that takes time. So for practical reasons is the number of maximum bits limited. All the clock pins are connected together and all the latch pins are connected together. The 74HC595 shift register has also an output pin that can be used to connect to the next data input pin. So the bit that 'falls out' of the shift register, can be read by the next shift register creating one virtual giant shift register. By applying five registers, 40 bits can be set. In the example below are two shift registers drawn. In this case not only 11 button/switch positions are connected to read the state, also five leds are connected. Since each bit can be set as desired, not only one bit for reading the switches is possible, also other outputs can be set 'random'.
Documenting switch information
After removing the pcb from the switches, the switch connections are documented. On the images below is all the connection information available.
10 position rotary switches
In the image below is the switch information of the 10 position rotary switch. There are five rotary switches installed. Two pieces for the channel number where only position 0...2 is used for the first channel number digit. The other three rotary switches are connected to the three middle knobs of the frequency setting. Pin [C] is the common pin. In position , no connections are made. This is quite odd since there's no way to determine if the rotary switch is set to zero of the switch is broken/missing... Altough five pins are switches, only four are needed. Pins 1, 2, 3 and 4 are used for binary encoding. Pin 5 is not needed for determining the switch position. So four bits are needed for determining the switch position.
4 position rotary switches
The frequency steps are 25 kHz steps. Therefore 'only' four switch positions are needed instead of ten. So a much smaller rotary switch is used. All four positions are used. The switch has two independent switch parts. For determining the 00, 25, 50 or 75 kHz steps, four wires can be used. The minimum amount of bits is three instead of four. If it isn't one of the first bits, the switch is in position four. This reduces one bit, but if the switch fails there's no fail-safe check. Since number '0' is also not fail-safe as mentioned above, three bits should be enough I guess...
5 position rotary switches
There are also (three) five position switches used next to the four position switch. But not all the switch positions are used. The switch movements are limited by mechanical locking pins that moves in slots. The  for 1××,××× MHz is usually blocked so only 2××,××× and 3××,××× is available in normal conditions. These rotary switches are dual type switches and are used for the mode switch, on/off and transmit power switch and the first digit of the frequency selection.
Button, toggle switches and pots
For storing and reading information a three position switch is used. The switch is sping loaded and returns to the middle position. For switching the squelch (SQL), a two position toggle switch is installed. For sending a tone a normally open momentarily push button is installed. For the audio volume knob is a wire wound 5K Ohm potentiometer installed. For the 'main' is a 10K trimmer pot installed. The use of the 'main' pot is unknown, but I assume that the squelch level or transmitter power can be set.
There's also a 7-segment display available. It looks like a old style BCD/binary 7-segment display and I have no purpose for reusing this so I didn't spent time for determining the connections...