This is a 70 metre diameter dish, much used by NASA.
The first picture was taken in September 1994. DJJ is in the foreground. The second picture was taken in December 1998 and shows the primary feed structure, the 10 metre diameter sub-reflector, and the supporting gantry. The third picture shows a closeup of the subreflector taken in November 2000.
When a first attempt was made to photograph the dish in December 1998 using a Canon Z115 automatic compact camera, the camera electronics failed. This may possibly have been an EMC problem, as the camera was facing the lowered dish at an angle of about 30 degrees to the boresight direction of the primary feeds. Considering the 400kW final HPA power there could have been very significant field strengths at the camera, had the dish been transmitting. I wonder if NASA can be persuaded to refund the $120 repair costs for the camera electronics? Anyway, you have been warned about this problem if you wish to take your own pictures. The one posted below was taken with an ancient Pentax spotmatic mechanical camera.
NASA provides information about this dish . This page is part of a new (Dec 21 2000) site at CDSCC, the Canberra Deep Space Communication Complex.
Operating frequencies are listed as
1.628 - 1.708 GHz in L band
2.270 - 2.300 GHz in S band
8.400 - 8.500 GHz in X band
12 - 26 GHz in K band
The pointing accuracy of the dish is listed as 0.005 degrees or 0.3 minutes of arc.
The calculated boresight gain at 8.45 GHz for this dish, using the formula gain = 4 pi Area/(wavelength squared), is 39,200,000 or 75.9 dB. The wavelength at 8.45 GHz is 3.55 cm.
The actual boresight gain is somewhat less than this because of the efficiency factor due to spillover, blockage, antenna surface profile errors and resistive losses, and so on.
The main beam diverges, at this frequency of 8.45GHz, at an angle of about 2 minutes of arc, in theory. A tilt of the dish up or down by 1 cm at the rim is thus sufficient to move the beam off-target. That is why there is so much bracing below the dish. We remark that if the dish rim moves by 0.88 cm forwards along the boresight, the diametrically opposite side moves back 0.88 cm. There is then a half wavelength path difference at the feed between the signals reflected from opposite sides of the diameter of the dish main reflector. Thus it is not surprising that a motion of 1 cm, which is a little bit more than this, is sufficient to steer the beam off-target.
If we assume that the receiver system overall noise temperature is 60 Kelvin for this dish, and if we assume a receive channel bandwidth of 10Hz and a carrier frequency of 8.45GHz in X-band, then the carrier wavelength is 3.55 cm and the sensitivity of the receiver for a 3dB signal to noise ratio (the signal is just the same size as the noise) is 1.65E-20 watts. The receive signal strength at the 70 metre main aperture is therefore 4.3E-24 watts per square metre.
Now if we assume a probe out beyond Pluto with a transmit antenna dish diameter of 1 metre and a transmitter power of 3 watts, the gain of the transmit dish is 7,800 and the effective isotropic radiated power is 23,400 watts. At a distance R metres between probe and ground station the received signal strength is therefore (23,400)/(4 pi R*R) and equating this to the receiver sensitivity of 4.3E-24 watts per square metre we find a range of around 2E10 kilometres. By comparison, the orbit of Pluto is at 5.9E9 kilometres from the sun, so this antenna reaches at most 3.4 times the distance to Pluto and so is nicely designed to cover the outermost planet of the solar system with not much to spare. The absolute maximum data rate at the orbit of Pluto would be about 2-4 bytes per second. Imagine transmitting a picture of Pluto at such a data rate. One could conceive of a compressed picture size of about 10k-Bytes so the picture would take about an hour to transmit.
If you browse around on the web pages about these kinds of dish you will come across a unit of field strength referred to as the Jy. The Jy is named after Karl Jansky who did the pioneering work on radio astronomy, and who was the first person to observe radio waves emanating from beyond our Solar System. The Jy unit is 10^-26 watts per square metre per Hertz of bandwidth. Thus for the 70 metre diameter dish which has area nominally 3849 square metres, the signal strength necessary to provide a signal equal to a 60 Kelvin noise temperature is about 21.5 Jy.
If the probe loses its orientation the on-board dish antenna may not be locked onto the direction of Earth and so we have to assume communication is into an omnidirectional antenna on board the probe. Again, at X band frequency 8.45 GHz the transmit gain of the Tidbinbilla antenna is 39,200,000 so the effective isotropic radiated power is 39.2E6 * P where P is the final amplifier power of the transmitter. The received signal strength at 2E13 metres is therefore 7.8E-21 * P watts per square metre and the effective area of an omnidirectional X band antenna at wavelength 3.55 cm is no more than a quarter of the (wavelength squared) or 3.15 square cm. Thus we can expect a received signal power of 2.46E-24 * P watts compared with a noise power into the receiver of kTB watts (k = Boltzmann's constant, 1.38E-23 J/K T = effective noise temperature of the receiver and B = bandwidth in Hz). Again we can assume a probe receiver noise temperature of 60K.
Assuming we would like a S/N of 10 at the receiver for a bandwidth of 100 Hz for telecommand purposes, that puts the required transmitter final amplifier power at about 330 kilowatts.
Looking at the specification of the transmitter in the NASA url listed above, we see that the transmitter is rated at 400 kilowatts.
The lesson to be learned is, never underestimate the power of a link budget calculation. We now also see why it was necessary to increase the dish diameter from 64 to 70 metres to communicate with the Voyager probe.
Of course, this kind of link budget is only an estimate. First, we haven't accounted for pointing accuracy errors. Second, the assumption of 60K receive noise temperature may be a bit optimistic. Third, are the assumptions of 100Hz and 10dB S/N realistic? But nevertheless the calculation is instructive, especially on the orders of magnitude of signal needed and on the tradeoffs involved.
These are my calculations and may possibly be in error....
Incidentally, just across the road from the antenna farm is the nature reserve, which has a Koala enclosure. Here is a sleeping Koala at the Tidbinbilla Nature Reserve in November 2000.
