Progress on the RF stages

Back

These are my design notes for the development and construction of a dual band radio telescope. It should be noted that this is not the official BAA-RAG 2.695GHz telescope which is developing under the guidance of Terry Ashton. This telescope is a total power receiver and will be used to monitor the RF noise produced by the Sun.

The telescope is entirely modular. I have chosen this approach so that units can be swapped out as the design progresses. The medium term objective is to integrate the RF sections with the BAA RAG Starbase system. In the sort term, I want to get a system in place so that I can prove that the S band and VHF modules work to a satisfactory level. It can also be used to baseline performance against the BAA-RAG Radio Telescope. It is intended to make circuits and layouts publicly available for reproduction by competent engineers, however, it will not be made available as discreet assembled modules

The two bands that I am currently focussing on are centred on the Radio Astronomy dedicated frequencies of 2.695GHz and 151MHz. Both these frequencies are fairly close to the Amateur Radio allocations of 2.4GHz and 144Mhz and to accelerate the development time, existing proven published circuits from these bands will be reused. Full credit to the original designer will be acknowledged.

Phase 1

This phase comprises of the production of two modules. A 2.695GHz to 151MHz down-converter and a 151MHz to 28MHz down-converter.

The Starbase system utilises a 38MHz IF, so it is likely that the 151Mhz converter will (hopefully) become redundant in a short period of time. However, I have a very sensitive 28MHz receiver (RATS [Receiver Alignment Test System]) designed by Roger Blackwell G4PMK and published in RadCOM, (Journal of the RSGB) in July and August 1995. This will be ideal to setup and prove the higher frequency RF stages.

The 2.695GHz Down Converter and Local Oscillator

For the prototype, I have decided to use a receiver that incorporates a sub-harmonic mixer with external oscillator. The benefit of a this type of mixer is that the local oscillator frequency is doubled at the point of mixing, therefore for a 151MHz intermediate frequency:

2695MHz-151MHz = 2544MHz (local oscillator)

Because it's a sub-harmonic mixer the LO needs to be half of the wanted 2544MHz = 1272MHz. About 10dBW of RF is required. This is much easier to produce than at 2544MHz. And as a modular approach has been adopted, different oscillator designs and configurations can be utilised without having to redesign the artwork for the amplifier and mixer stages.

Initially, I am using a WDG016 kit (available from Charles Suckling G3WDG) for the Local Oscillator. This uses a x12 multiplier to achieve the desired output frequency, so 1272/12 = 106. A 106MHz 5th overtone crystal is required and this has been sourced from Quartslab. This oscillator provides a very stable and clean output. This type of oscillator is perfect where spatial resolution of the received signal is required, for example, in a H line receiver. However it is overkill where the requirement is for a total power receiver. So I am investigating other less critical, lower cost alternatives, possibly based on DRO (Dielectric Resonator Oscillator) technology. These provide a clean output and are used in the LNB’s of many domestic satellite TV systems.

The RF and LO stages will be contained in a waterproof box and located directly behind the dish. I expect that the Noise Figure will be < 1dB with about 36dB of system gain.

The 2.695Ghz receiver layout and gain distribution is estimated to be as follows:

The 2.695GHz down converter will use a pipe-cap filter as an image rejection filter. I have documented my experiments with this type of filter on my web site at http://www.czd.org.uk/astro

If required, additional gain blocks will be added between the 2.695GHz and 151MHz converters.

The prototype converter is constructed on 0.8mm PTFE based PCB material with an Er of 2.5. The layout and circuit diagram will be available following system testing.

The 151Mhz signal will be brought down to the radio room using RG213 coaxial cable to connect to the 151 to 28 MHz down converter.

151MHz to 28MHz Down Converter and Local Oscillator

The design of the 151MHz module is more demanding. The allocation at 151MHz is very small and there is a great deal of adjacent noisy users, such as the BT radio paging service. Propagation from VHF sources tends to more omni-directional than microwave sources and there is far more scope for interference. Additionally, because the wavelength at 151MHz is about 18 times longer than at 2.695GHz, there is less scope to build a more directional antenna (unless you have the space for a 12 metre diameter dish!) and the gain capability of the average aerial will be about 12dB less. Therefore a receiver with more selectivity is required to suppress interference from other sources. But a higher gain with lower noise front end is also required.

Amplifiers by their nature are fairly broadband so will amplify the local unwanted signals as much as the wanted signals. And to compound the problem a total power receiver is inherently broadband as we are measuring the received power across as wide a spectrum as we can get away with! Additionally, the effects of mixing from the image frequency should not be underestimated. To achieve a 28MHz output from a 151 MHz input requires a Local Oscillator of 123MHz. But 123MHz - 28MHz = 95MHz which is right in the middle of the medium power VHF FM local radio broadcast band! Unless these are totally suppressed at the receiver front end, they will simply add to the total power received and passed to the detector. They will comprise a far bigger component of the signal than the signal received from the sun. With a 38MHz IF, the problem is not quite as bad. A local oscillator of 151MHz - 38MHz = 113MHz will be required. The image will be at 113MHz - 38MHz = 75MHz. I think this allocation is currently used for commercial VHF vehicle communications, but this is much easier to filter out. Ideally, a dual conversion receiver may be a better option and this may be considered for the future.

You can see why the design of the 151MHz converter is tricky! In hindsight, I think using the RA allocation of 408MHz as the IF may have been a better strategy.

Usually, the selectivity (the ability to resolve the wanted from the unwanted signals) of a receiver is carried out in the detector stages, however careful design and filtering at the higher frequency stages will clearly help.

I have identified an existing and proven (144MHz to 28MHz) converter design that is no longer available in print. I have modified the circuit to use SMD technology as many of the original discreet active components are no longer available. The artwork and the circuit diagram have been produced and the design is currently being built. This module will be enclosed in a fully screened case and fastened to a standard sized Eurocard PCB to locate in the Starbase chassis.

These two modules will be aligned using the RATS and a standard amateur band 28MHz receiver.

If this is successful, the VHF converter will be redesigned at Phase 3 to provide a 38MHz IF output.

Alignment

Alignment of the 2.695GHz system will be by the use of a comb generator. A Crystal controlled oscillator running at 96.25MHz will have its 28th harmonic exactly on 2.695GHz. A narrow band receiver should hear this is a warble free whistle. The stability of the oscillators can also be tested using this method.

151MHz Low Noise Preamplifier

The prolific German designer, Rainer Bertelsmeier DJ9BV has produced a number of very low noise preamplifiers over the last few years.  In Dubus Technik IV he detailed a MGF1302 based amplifier designed for 144MHz. I plan to build this and locate it at the 151MHz antenna. This will be followed by a 20dB gain block using MAR-6 devices and then fed to the 151 to 28 MHz down converter using RG213 coaxial cable.

Thus ends Phase 1.

Phase 2

This details the introduction of a switched noise source and calibration system. I also need to devise a method of getting the aerials to follow the sun. More on this and the effect of temperature in the next circular.

Phase 3

Integration into the Starbase system.

Progress on this project along with circuits and PCB layouts will be made available on my website as they become available. An update will also follow in the next circular.