Thanks for finding that. Apparently the receive antenna has a gain of 0dbi. The use of PCM/FSK for the modulation schemes means you need to achieve a high signal-to-noise ratio on the receiver for the message to be coherent to the spacecraft. I don't know the spectral power density of a PCM/FSK, but this most likely describes everything you'd want to know about the scheme: http://ieeexplore.ieee.org/xpl/login.jsp?tp=&arnumber=109174... . The abstract specifically mentions NASA.
If the issue is that all of the existing gear is wideband and you need to send something narrowband - build a narrowband exciter, use an upconverter to drive the wideband amp.
PCM/FSK-AM/PM doesnt seem to be really a complicated thing. Its a pretty easy way to send data, and should be build-able.
They say they need to use the DSN to do the comms. This is a set of three stations, in the US, in Spain, and in Australia, that has a 70 meter dish each, plus some 34 meter dishes.
It's not regular ham equipment in many ways, obviously. Besides power, aperture, and accurate pointing, which have been mentioned here, there is also the need to hand off transmitting/receiving as the Earth rotates. You also need a receiver and transmitter velocity model to adjust for Doppler.
People are not giving the difficulty of this problem adequate respect.
How hard would it be for access to be granted to the DSN? The challenge to build the transmission equipment does not seem insurmountable nor even particularly hard.
It's the logistics and red tape to make it happen.
I'd also like to point out, that this satilite will be doing a near earth flyby - you could probably get away with something smaller, maybe even a 3.5m dish versus a you know, 70m dish.
The engineers at GSFC say you will need the DSN. The smallest DSN antenna is 26m and that is used only for LEO (a couple thousand km).
You talk about "red tape" like moving a huge radio telescope is some kind of formality, and you quoted a figure of 3.5m with no apparent engineering basis. I find this ridiculous.
There are lots of other issues involved, but the DSN picked up the carrier signal in 2008, and it's much closer now. The article also mentioned that people can listen to the carrier as it goes by. I don't think dish size is one of the problems.
well, other than 0db gain antenna on the sat. But, ideally, if we can hear it, all we need to do its match the equivalent ERP on this end that the sat is sending (gain of the TX antenna on the sat, plus the minimum gain of the RX antenna required to successfully receive) it should work.
The previously linked pdf puts the up/downlinks around 2.0/2.2GHz; the same band used for the Apollo missions, which figures given the era (http://en.wikipedia.org/wiki/Unified_S-band).
Currently, the entire Deep-Space Network is on ~8.4GHz.
Re-equipping a multinational network for S-band could quite possibly a significant budgetary issue given how little love have been NASA getting in recent years.
Yeah, it could be implemented inexpensively in SDR using GNU Radio and a USRP. Of course you'd still need the front end hardware; high gain antenna, LNA, and PA. Not something out of the ordinary for some hams though. It's probably NASA not having the manpower to devote to it.
Apparently the highest gain antenna NASA has is 61.7 dbi at the relevant frequency. If they used 5 watt transmitters, that would be like running 4.5 megawatts into a conventional dipole antenna. The reality is NASA probably used many tens or even hundreds of watts to contact the spacecraft. As an amateur radio operator I assure that is out of reach for most of us.
> Apparently the highest gain antenna NASA has is 61.7 dbi at the relevant frequency.
Don't forget angular resolution. Because of the antenna's geometry, it's misleading to describe its gain without also mentioning that the gain applies to a very small angle -- which, depending on the circumstances, may be a great advantage if it needs to reject interfering sources.
61.7dbi, while it represents a peak rather than a more descriptive statistic, does go a ways toward describing 'angular resolution', or rather, how tight you can make the beam of the transmitter.
It's easiest to think about it thusly: You steal part of the power that would normally go out equally to a sphere ("isotropic radiator") and you redirect that power to a smaller part of the sphere. If you can successfully redirect the entire power into only half of the sphere (1 hemisphere), you get +3db. Keep slicing that in half and you add +3db each time.
61.7dbi gain necessarily means that the power felt at the center of the target is 1545883 times as powerful as it would be if you were using an isotropic radiator, given the same number of input watts. If you have successfully concentrated the power that much, something at the target's range isn't going to detect sidelobes much at all.
This isn't a technical summary - radiation patterns are never absolute step functions and there is indeed a distinction between peak power and usable area... it's just not as relevant as the huge number represented by 61.7dbi, for any remotely gaussian distribution of signal.
I did a little searching and came up with some nice details about the ISEE-3 communications system here: http://mdkenny.customer.netspace.net.au/ISEE-3.pdf