The 14m receiver box contains 3 receivers, for 2.2 GHz, 4.8 GHz, and 6.7 GHz. These are all mounted on a sliding platform, so that each of the 3 receivers can be positioned at the focus of the dish.
Things which are not on the sliding platform:
Things on the sliding platform:
This is a custom PCB with a PICAXE 40X1 microcontroller, and a bunch of other glue logic. The PICAXE could be replaced with a 40X2 with only minor changes to the code. The git repository for the PICAXE code has a branch labelled ‘master_40X2′ with these changes.
The entire receiver box is controlled through this PIC board via RS232. It has the following functionality:
These are the commands that the PIC accepts over RS-232 (at 2400 baud). Each command is a single byte lowercase letter. The commands to move in to a position could take 10 seconds or more to complete.
‘Print’ means the PIC will send it over RS-232.
The long and short status lines that the PIC can send are:
States “M” and “U” should never happen.
Interfacing with the ADC
The output from a ‘dump’ command is 35 bytes, corresponding to the 35 ADC channels which it can measure. There is a QBasic program on Eric’s computer in “My Documents\14m_analogue_interpreter\” which shows what each ADC channel does. Also see the PIC Board Headers section.
The PIC board has two separate 5V regulators: one to power the 5V relays, and one to provide a stable supply for all the digital components on the board.
The relays on the PIC board are:
The PIC uses 4 output pins (bits 3, 4, 5, 6 of PORTB) to control the relays. These 4 pins connect directly to a 4-to-16 line decoder, so that one of the 8 relays can be turned on at any given time. Eight of the outputs of the line decoder are connected to an 8-way transistor array, which can sink enough current to actually power the relays. (Only 8 of the outputs of the line decoder are used, so there is only one 8-way transistor array.) Due to the way the line decoder and relays are wired, if you write the value 56 to PORTB, then the relay (56) will turn on, etc… Setting PORTB=0 will turn off all relays.
In order to switch the grey relay, you must first activate either (40) or (56) to set the remnant relay, then pulse (64).
The PIC board can measure 35 different analogue voltages, but the PIC only has 7 analogue inputs. The PIC analogue inputs A0, A1, A2, A3 are each connected to a 8-way analogue switches, giving a total of 32 analogue inputs. The PIC analogue inputs A5, A6, A7 just measure one thing each, for a total of 35.
The 8-way analogue switches have 3 control lines (2^3 = 8), which are connected to the PIC PORTB pins 0,1,2. Thus, to read one of the 32 analogue inputs, first write a value 0–7 to PORTB, then read one of A0, A1, A2, A3.
The PIC board has a MAX232 chip, which converts RS-232 voltage levels to TTL and vice-versa. This is used for a second purpose, in addition to its intended one. The MAX232 has a charge pump circuit to generate ±7.5V for the RS-232. On this PIC board, those voltages are also used to power an op amp. (This op amp converts the −20V supply voltage of the receiver box to a voltage that the PIC’s ADC can measure.)
The board was designed in DipTrace, and the file is located on Eric’s computer at “My Documents\14m_pic_schematic\”. There are a few bodge wires on the physical board, but the DipTrace designs have been updated to make it equivalent to the physical board with its bodge wires.
The PIC board has 5 IDC connectors. Going from top-to-bottom, then left-to-right:
These two headers (at the bottom of the PCB) connect various voltages (on the RF hardware) to the ADCs of the PIC, so they can be measured.
Each FET servos |FET servo board contains 3 FET servos, and each servo has 2 measurements:
For instance ‘D2′ refers to the drain voltage of the second servo.
The following table contains the same data as the table above, but rearranged in order of FET servo board. The 4 FET servo board also have a colour (yellow, green, red, blue) which is the colour of the heat shrink on the cables going from the servo to the LNA.
The PIC is programmed using PICAXE basic. The code with comments is on Eric’s computer, at “My Documents\14m_drive_program\”. If you want more details, read the code and its comments.
There are a few other branches in the git repository:
The PIC program must be able to determine the position of the platform, but is only able to measure the state of the 5 position switches. In order to keep track of the platforms position, it implements a state machine, with possible state numbers:
There is an exception to the ‘multiple switches pressed’ rule. It is physically possible for both the 4 GHz position switch and limit switches to be pressed. In this case it is treated as if the platform is in the 4 GHz limit (not 4 GHz in position). This is why there are two ‘1′s in the state table to the right. Similarly for the 6 GHz end (there are two ‘9′s in the table to the right).
All this may seem a overly complicated, however it makes the system robust.
It also confines all the complicated code to the PICAXE BASIC function ‘getPosition’, leaving the functions ‘putPlatformInPosition’, ‘motorForwards’, ‘motorReverse’ relatively straightforward.
In the figure to the right, J4 and J7 are IDC headers on the PIC board, and AS1, AS2, AS3, AS4 are the analogue switches.
The receiver has 3 RF ‘boards’ (which are aluminium plates to which some RF components are attached):
In order to select a receiver, some RF relays must be set to:
This generates a signal to be mixed with the RF to produce the IF. It produces two identical outputs, which go to the two RF-IF mixer boards.
Input: 10 MHz reference.
Output: 2.65, 5.3 or 6 GHz.
There are only two oscillators: a 6 GHz one and a 2.65 GHz one. For 2.65 and 6 GHz output, the relays just switch in the respective oscillator. For 5.3 GHz output, the output from the 2.65 GHz oscillator is fed into a frequency doubler, and then into a filter (which is presumably a 5.3 GHz band-pass filter).
The letters (C, P, J, S, T, V) in the figure to the right refer to the pins of the ‘main connector’.
Also note the grey relay connected to T and V, which is a bit tricky to drive.
There are two of these boards, one for each polarization.
For each receiver, and each polarization, the signal from the sky goes through:
The amplified sky frequency signal from each antenna then goes through some RF relays, which choose one of the 3 receivers. Then, the chosen signal goes through:
The letters (G, F, E, H) in the figure to the right refer to the pins of the ‘main connector’.
All of the RF relays are latching.
All the RF relays (with the exception of the grey one) just require a pulse of either 5V or 24V on the correct pin. The 5V relays actually contain some control circuitry, and they can accept a TTL signal.
In the figure to the right, the relays are named A1, A2, B1, B2 are on the RF-IF boards A and B, and select one receiver to use. These two boards are (practically) identical, and operate in parallel, so there are only 4 control lines (G, F, E, H) to these 4 relays.
The relays ‘blue’, ‘grey’ and ‘black’ are on the local oscillator board, and select the LO frequency. The 6 control lines (J, S, T, V, C, P) control these 3 relays.
These letters refer to the pins of the main connector.
The figure to the right shows which control lines (and which relays) must be activated in order to select a given receiver.
This relay has only two control pins (T and V), and requires a polarised voltage across them.
To change to channel ‘1′, ground pin T, and apply +24V to pin V.
To change to channel ‘2′, ground pin V, and apply +24V to pin T.
Note that you have to specifically ground one of the pins (it is not connected to ground internally).
Originally, the RF relays and platform drive motor were controlled manually. In the cabinet below the telescope there was a 19″ wide piece of aluminium angle with some switches on it. The switches were wired directly to the RF relays via the ‘main connector’.
The figure to the right shows the switches of this panel, and how they were wired. The letters (G, F, E, H, J, P, S, T, V) refer to the pins of the main connector. The only tricky thing is the relays connected to the letters T and V, which go to the grey RF relay.
This manual selector panel has been replaced by the PIC board, and each switch on this panel has been replaced by a relay.
The parenthesised numbers (8, 16, 24, 32, 40, 48, 56, 64) in the figure to the right refer to the number which must be written to PORTB in the PIC code, in order to switch on the respective relay.
There are 3 receivers, each with 2 polarisations, so there are a total of 6 low-noise amplifiers. The 2.2 GHz receiver has 2 off-the-shelf LNAs, which just take +15V.
For the 4.8 and 6.7 GHz receivers there are a total of 4 (custom made, I think) 3-stage FET amplifiers. The quiescent current and voltage of each FET is controlled by a servo circuit, and they have all been tuned for minimum noise. There are 4 FET servo boards, each with 3 independent servos for the 4 3-stage LNAs.
The voltage and current for each of the 12 FETs can be measured by the PIC.
The connectors which go out the the box are:
This is a 19-pin connector, which was once served several purposes. However, now only 5 of the pins are being used, which provide DC power to the box.
The other pins connect (inside the box) to the RF relays, so it could be used to control the RF relays manually, without using the PIC.
Before the creation of the PIC board, the RF relays were controlled (via this connector) by a control panel covered in switches.
There is a sliding platform which holds the 3 receivers (2.2, 4.8 and 6.7 GHz), and allows any one of the receivers to be placed at the focus of the dish. This platform does not have full ‘xyz’ control of position, in fact it does not even have ‘x’ control. Instead, the platform has a little tooth which can press one of three switches to indicate that a receiver is in position. Additionally, there are 2 limit switches.
There are 3 relays, which are controlled by 3 pins of PORTC on the PIC:
The switches can be controlled independently.
Example: say the platform is on a position switch, and we want to drive it forwards to the next position switch.
The switches do 2 things:
When the platform drives into one of the switches, power will be immediately cut to the motor.
The PIC is able to turn on a relay (‘GO’) which overrides these switches.
The limit switches are not overridable, instead they are wired in parallel with a diode, which allows the platform to get out of the limit.
On the digital side of the switches, there is a 220 Ohm pull-down resistor connected to each switch.
Without these, there would only be a 10 kOhm pulldown (located on the PIC board).
The purpose of these is to allow some wetting current to flow when the switch is on.
Without these, the switches were unreliable.