A NEW Mode S Downconverter for A0-40 - G3WDG040
A new downconverter has been developed for Mode S. The is based on the well-proven WDG010 2.3GHz transverter, but uses an on-board local oscillator, based on the WDG016. All the components associated with transmit have been removed to save cost.
The 2 prototypes and 4 beta test units are all working well, exhibiting 0.65-0.75dB noise figure, with about 38dB conversion gain. The very low noise figure means that an external preamplifier will not be required, unlike some of the converted surplus units around. The conversion gain could be reduced if desired, by not fitting the IF amplifier modamp.
There is provision in the design to power the unit up the cable, eliminating the need for a separate power cable. The unit requires a +12V supply at approx 120mA. If used outdoors, the unit needs to be housed in some kind of waterproof box .
As of early August 2001, the prototyping and subsequent beta-testing phase of this design have been successfully completed, and we are now taking orders for shipping in September. Cheques will not be cashed until orders are shipped and we will be sending orders out in the order in which they are received (by postmark). We anticipate a strong demand for this product and please bear with us if it takes a few weeks to ship the first units out.
As with all our products, we will endeavour to help you through any construction/tune-up difficulties if any are encountered, so you end up with a working unit. Provided units are built to a reasonable standard, we will get them working or check performance for the price of the return postage and any components needing to be replaced.
Construction of the unit is well documented in the writeup which accompanies the kit, which is similar to the details posted elsewhere on this site.
Photo of an early prototype converter - production models have been simplified!
Downconverters tested on air!
The prototype converter has now been tested on air using the beacon signal from AO40. The test was done with the satellite at about 40,000km range and with a low squint angle, and was about 10 deg above the local horizon at the time. Using a very small antenna (a WA5VJB 2.1-6GHz pcb log periodic), a signal from the beacon was easily detected owing to the very low noise figure of the downconverter. The spectrum of the signal as displayed by the ao40rcv soundcard telemetry decoding software is shown below, together with a photo of the downconverter and antenna.
PCB Revisions:
G3WDG S-Band RX: Original Prototype
G3WDG040:
Omitted negative voltage generator
G3WDG040_1:
Small changes to layout of IF amp, added ground pads for LNA
screen
G3WDG040_2:
Increased width of LNA gate bias line to improve process yield
Biasing up the IF line
The pcb has the provision for biasing the unit via the IF line. This removes the need for a separate power cable. Two extra components are needed: a 10nF chip capacitor and a ferrite bead. The bead is used to make a homemade inductor by winding 2.5 turns of 0.5mm enamelled copper wire through the hole in the bead. The circuit and layout of the bias network is shown below. A prototype unit was modified in this way, and performance was unaffected.
The +12V output from the +12V line pin is taken to the input of the 78L09 regulator.
Installation in a waterproof housing
Some constructors of the WDG040 have asked for details of how to put the WDG040 into a waterproof housing. We have recently turned our attention to this issue and the notes below outline one way of doing this.
Waterproof boxes are available from a number or sources. We opted for a gasket sealed plastic box rated to IP54/IP65 on grounds of lowest cost and weight. This was obtained from a local electrical retailer as an "outdoor junction box".
The photo below shows a WDG040 installed in this box.
The unit was fitted in the box by drilling a clearance hole in the lower box lid to accommodate a bump in the box moulding, and then clamped by a strap made from aluminium sheet. The strap was secured to the box using M4 screws. Existing bushed holes in the box were tapped M4. Suitable self-tapping screws could probably also be used (cut to length or with washers to prevent the screws from breaking through the bottom of the box.
The RF input and output connections were made via semi-flexible/flexible 50 ohm cables. Special attention was paid to the 2.4GHz input connection, in order to minimise discontinuities. The classic solder tag for ground and length of unscreened inner conductor is not adequate at this frequency, but is OK for the 144MHz IF. The 2.4GHz transition between the input N-connector and the WDG040 was made using a small pcb and length of semi-flexible 50 ohm cable. The launch pcb is shown below:
At the top is a strip for grounding the outer of the cable. Connection to the groundplane is made with two veropins, fitted as per the instructions for the grounding pins in the WDG040. There is then a short length of 50 ohm microstrip. The shape below ths indicates the area that needs to be cut away in order to accommodate the protrusion on the back of the chassis-mounting N-connector. A sharp scalpel is recommended for cutting this away, followed by filing if necessary to make the pcb a good fit on the back of the connector. The groundplane of the pcb is then soldered (with a large iron) to the protruding metal on the back of the N-connector at either side. Front and back views of the pcb mounted on a connector are shown below. The connector spill (inner conductor) was filed to leave about a 2mm square 3mm long. Tests on the transition showed that it introduced less than 0.1dB loss, including the expected length of cable.
The connection between the pcb and the WDG040 is made using a short length of low-loss 50 ohm cable. We have obtained a high quality cable similar to 085 semi-rigid, but with a tinned braid outer conductor. The easiest way of fitting this is to make the connection initially to the WDG040, leaving a "tail" sticking out which is then joined to the launch pcb after fixing the WDG040 in the box. Careful marking out of the position of the N-connector is important, to make sure that the cable runs straight from the WDG040 to the pcb. A small amount of mis-alignment can be taken up by forming the connector tail to the correct shape before it is soldered to the pcb.
The cable should be prepared by sawing off about 2mm of the outer conductor and then trimming off the insulation to leave the inner conductor protruding. The cable is then offered up to the hole in the WDG040 box and the inner soldered to the input microstrip of the WDG040. The outer can then be soldered to the outside of the box. Note that the outer does not need to be splayed out (unless the hole in the box is too large). If the tail is made slightly too long initially, it can be trimmed to length to be an exact fit to the pcb once the position of the downconverter has been fixed in the plastic box. After trimming to length, the other end is prepared as above and finally soldered to the pcb on the N-connector, outer and inner. The inner may need to be bent down slightly to touch the 50 ohm microstrip.
We are able to supply all items needed for this, including plastic box, launch pcb, veropins and the semi-flexible cable. Please contact us for pricing.
If connectors have already been fitted to the downconverter, they can be removed to allow this type of interface to be added. Alternatively, cable-entry chassis-mounting connectors could be fitted to the plastic box and plug/socket connections made to the WDG040. We do not have a source of such connectors and have not tried this.
The IF connection as noted above is not so critical. The installation above used a length of thin flexible 50 ohm cable to join a UHF chassis-mounting socket to the WDG040. The ground connection at the connector end was provided by a solder tag under one of the mounting screws. The choice of IF connector is open - BNC, N, UHF etc should all be suitable.
Susceptibility to RFI has not yet been evaluated. It may be that the short unscreened length of microstrip in the launch pcb could pick up more transmit signal than would be the case if a fully "connectorised" system was used. If this does prove to be an issue, it may be necessary to add a screen over the microstrip section. Feedback on this would be welcome.
Final note - the efficiency of the waterproof box has not been tested by us. Boxes supplied are rated waterproof but we are not able to verify this from personal experience! It may be advantageous to apply some form of sealant around the connector bodies where they contact the box. Checks after installation might also be wise. To quote a famous microwaver: "Take all steps to make sure outdoor housings are fully waterproof, and then drill a hole at the lowest point to let the water out!"