|
|
|||||
|
|||||
|
The Radio Telescope is
composed of a frontend unit which
includes the low noise
preamp and a cylindrical feed horn for 1420 MHz. The frontend
is connected to the backend rack unit by user supplied low loss coax
cable with type "N" connectors. We recommend a very low loss cable
such as LMR-400 flexible cable or a Belden 9913 type derivative at
the very least. This cable should have a loss of 6 Db or less at
1420 MHz. Depending on the length of cable run it may be necessary
to add an additional low noise preamp in the line. Extremely low
loss hard line such as Heliax will allow runs of over 100 feet
with no additional preamp needed. It can not
successful operation of the telescope. Please do not skimp on the purchase of this material. Price certainly reflects quality and durability. Expect to pay at least $1.00 per foot for good material. It would also be a wise investment to purchase this cable with type "N" connectors already installed. If you are not familiar with installing these then you run the risk of water damage to the cable due to infiltration. A small fee will be charged by the cable supplier to install these correctly.
|
|||||
|
The feedline carries the low noise amplified signals to the telescope backend. The backend electronics are enclosed in a standard 19" instrument rack which should rest on a table top or be enclosed in a rack mount chassis. It is important to maintain the rack in as temperature stable environment as possible. This will keep electronics gain drift to a minimum. The back end is powered by a 110/220 vac supply which produces a nominal +/- 12.5 vdc for the electronics in the rack as well as a rear panel output to power the preamp and an optional noise source. The instrument communicates with an IBM compatible CPU via an RS-232 dB9 connector. The 1420 MHz signal from the frontend enters the rear panel of the back end and is fed to a 1420 to 70 MHz dual conversion down converter. This converter has approximately a 12 MHz bandwidth with the Hydrogen rest frequency at 70.0 MHz. This 12 MHz wide IF signal is passed through a programmable gain IF amplifier and then split between the continuum square law detector and the spectrometer third conversion mixer. The programmable gain IF amp is used to compensate for feedline losses and to place the signal in optimum range for the square law detectors.
|
|||||
|
The spectrometer third mixer converts a 2 MHz wide band of the 70 MHz IF centered on the Hydrogen Rest frequency to 21.4 MHz via a pll local oscillator which is stepped in 5 KHz increments under computer control. This gives a 400 data point sweep for +/- 1 MHz from rest. In reality most of the Hydrogen clouds observable with a smaller dish will reside in a Doppler shift range of +/- 600 KHz from rest. The software chooses this range by default but it is operator selectable. The 21.4 MHz third IF is passed through an 4 pole crystal filter with a 3dB bandwidth of 15 KHZ then amplified and detected with a similar square law detector as the continuum channel.
The outputs of both the continuum and spectral detectors are feed to a microprocessor controlled data collector which contains programmable DC Gain, DC Offset and signal integration. These functions are set by commands from the IBM and interpreted by a Basic Stamp CPU on the collector board. The conditioned signals are then sampled by a dual 12 bit a/d converter and sent to the IBM as requested by the software.
The square law detectors cover approximately an 20 dB range. This should be more than ample to accommodate most amateur dishes when aimed at the Sun. A typical 10 foot dish gives about 10 dB of sun noise under "normal" sun conditions. The square law range of the spectral channel is of small concern, most of the Hydrogen spectra are only 1 or 2 dB out of the noise. This is a very limited dynamic range. The sun is the only object which can tax the square law detectors range. This is the primary use of the built in variable gain if amp, to adjust the conversion gain to ensure that the sun signal stays in the square law range.
A simple method to "gauge" the square law response is to observe the "cold sky" with the dc amplification set to X1 and the dc offset at 0 vdc. The channel of interest should read between 2.0 and 4.0 volts for optimum square law range. For sun observations choose an attenuation level to give a quiet sky output at the low end of this range while for weak signal sources it would be best to set the if gain for an output at the high end of this range.
|
|||||
|
A brief description of the DC signal path follows. This will help the operator understand the purpose of the dc gain, integration and dc offset controls.
The dc output of the square law detectors has a no signal level of about 2 vdc. A typical source might cause a .2 vdc level change on the output as it passes through the beam. This would be a very difficult to read change on the screen. An obvious solution would be a dc amp. A X10 dc gain would give a 2 volt signal change on the screen... a very easy to read difference. But wait! Your signal has a bias of 2 vdc which would also be amplified by the X10 gain dc amp. Now the 2 volt signal we wish to display is riding on a 20 vdc "bias". This will not work! The solution is to incorporate a subtraction of the "bias" to bring it to a very low level and then dc amplify to our hearts content. This subtraction bias is called the dc offset and can be adjusted by the operator for both continuum and spectral channels.
In the above case let us assume that a dc offset of 1.9 vdc is subtracted from the detector output. Now the baseline is at .1 vdc with a signal change of .2 vdc above this level as the source crosses the dish beam. If we amplify by X10 we have a 1 vdc baseline with a 2 vdc trace to represent the source crossing.
This is a very appealing display of the data on the screen. There is one pitfall to this operation and to prevent it requires careful attention of the operator. This is the case when an incorrect dc offset has been entered by the operator. As in the case above let us suppose that the operator entered an incorrect dc offset of 4 vdc. The 2 volt detector output minus 4 volt offset gives -2 vdc and then a X10 gain results in a -20 vdc baseline. This will not work either! Care must be used to ensure that the dc offset is less than the X1 signal level output of the detector or as in the above case, the dc offset should never be set at 2 vdc or higher.
Fortunately once a favorable dc offset has been determined for each channel there will not need to be a significant change made from day to day unless the if gain value is changed.
It is a good idea to move the dc offset in small changes while starting and stopping the scan to observe the results. The dc gain can then be increased and a "fine tune" made to the dc offset to bring the baseline to a comfortable area on the screen.
|
SpectraCyber
Hardware Description Version 1.1