The first thing I wanted to do before starting this project was to get a general idea of the system layout, hence the “part 0” thing above. Also, this will serve as the master post, linking to the other pages when the time comes.
Extracting a simple connection diagram always helps to simplify things later on when tracing and verifying signals, and also when a sudden space-time-vortex comes along and throws all those neatly organised cables into a messy heap. Might happen to be a cat, too. So let us first cycle through the different modules and their approximate functionalities. Click the thumbnails to get a better view and more detailed explanations in the picture subtext. Also, be advised that my interpretation of things is not necessarily consistent with the original service documents.
Oven-controlled crystal oscillator (OCXO)
Contained in the bottom right module block, consists of a styrofoam-encapsulated 10.000 MHz temperature-controlled oscillator and some clock distribution buffering. It delivers the main TTL clock that is also available on the backside ports as an instrument reference. The picture shows the whole top side of the module block, but the actual OCXO and distributor PCB are on the right. The rest of the circuits belongs to the next module, the RG.
Reference generator (RG)
This subsystem is strapped to the OCXO in the same module case. There are three particular circuits here: The 10 MHz TTL buffering and switchover for external references, a 10.7 MHz IF generator (PLL+VCO+Mixer) and another phaselocked VCO for a widerange signal that fine-tunes the RF output later on.
FM modulator (MOD)
Contains the FM modulation frequency generator for FM and PM, the mixing circuit and a XTAL+PLL octupler to provide an exact 80 MHz TTL reference from the 10 MHz base ref. The actual frequency and phase modulation is mixed onto the 10.7 MHz IF signal coming from the OCXO and then sent off to the VCO. Also hidden in this block is a downmixer that shifts an input between 80.1 and 100 MHz down to 0.1 to 20 MHz by mixing an external input with the 80 MHz reference. We’ll see what that is good for later on.
Voltage controlled oscillator (VCO)
This module features 8 separate VCOs that start at frequencies of 480, 560, 600, 640, 760, 880, 920 and 960 MHz. That’s right, no direct digital synthesis (DDS) in this generator – well, with the exception of the modulation audio signal. The VCOs are switched according to band and mixed with the 10.7 MHz IF from the MOD unit (containing the f./p. mod.) as well as the wideband variable signal from the OCXO, forming the unrefined RF output. You may have noticed that the VCO does not span the entire output range of the device, alerting us that several other tricks must have been used here. An important feature for diagnostics on this module are the two red/yellow LEDs on the PCB strip next to the rightmost coax port. These signal the unlocking of oscillators and should be off at all times.
This is where a lot of the magic happens. The 4002 uses three different methods of synthesis for different frequency ranges: Direct output, integer dividing and downmixing the VCO. An RF switching circuit on the right PCB selects the appropriate source/mode via PIN diodes, followed by a matching LCLCLCL lowpass filter to suppress its harmonics. The integer division is also included here. Passing the filters and the bridge to the left PCB, several buffer amplifiers, two power metering circuits and a PIN-diode based attenuator stabilize the amplitude of the output signal to +13 dBm. Instead of being fixed, the reference level can also be set from outside the module, providing around 20dB of linear attenuation range if this feature is enabled by pressing the button “CONTINUOUS” on the front panel. The mechanical attenuator is then kept constant while the PIN attenuator does all the work.
Mechanical attenuator (ATT)
A typical mechanical attenuator that switches pi- or T-type attenuator segments using magnetic actuators. I did not look inside since these devices are easy to break and mine is working. The attenuation values are conveniently written down, as is the internal signal path. The one special thing about this module is the pair of send/return ports that allow extracting and reinjecting the signal to/from another module if the corresponding switch is enabled. The doubler block is attached to this port. From the output of the attenuator, a semi-rigid cable runs directly to the RF front panel output.
Frequency doubler (DBL)
A frequency multiplication circuit that I have not yet inspected too deeply since it seems to work. I would expect the usual doubling circuit using a nonlinear component and appropriate band- or lowpass filters at the output. This module most certainly contains another AGC stage and is enabled when frequencies above 1080 MHz are selected. Note that the 4002 will not allow the user to input frequencies above 1000 MHz if the module is not connected! Sorry for the dull picture, I seem to have forgotten to take a good one when I had the block open. I’ll make up for that later on.
All of the modules are plugged into a backplane that carries the power and control lines, which are transferred by ribbon cable from a second backplane in the left side compartment. This one carries the control circuits on up to eight euro-format PCBs. In my case, six modules are inserted:
- Microcomputer (MC6802/MC6821/MCM68A10)
- DDS (generates the modulation audio)
- D/A interface
- Display interface
- Memory (battery-buffered RAM for stored profiles)
- IEC interface
The collection is completed by the modulator control unit that fills most of the bottom compartment of the case. This board contains the external modulator input circuitry and AM modulator.
Next up will be the block diagram that I worked out during my search for the different faults.
 Written in past tense because the device is currently already working, less some adjustments and calibration. Write-up of the parts will take some time.