This project started in the autumn of 2002 with a search for a 'simple', low cost but practical design for an AM receiver to be used in 2m IARU ARDF ( Radio Orienteering ).
It follows a traditional format used for ARDF i.e. single conversion, free running vfo and an IF with ceramic filters, but is less conventional, being single conversion direct to 455kHz.
This receiver should really be called 'Google' because it is a product of the generous content of ideas and circuits posted by Radio Amateurs on the WWW!
The inspiration for the design came initially from Steve Hagman's Simple Aircraft Band Radio, however, not convinced by the single conversion concept. I built several double conversion prototypes using the SA602 & SA605 ICs. It became evident that without really good front-end and interstage filtering image rejection was still likely to be poor, with little return for the additional cost and complication of the extra mixer stage.
So instead, I opted keep the design simple and accept the inevitable image signal ... BUT
if you look at the ARDF scenario, you can see this is not as bad a decision, as might seem at first.
In Europe the IARU ARDF band is 144.5-144.8 and when the VFO is offset lower in frequency, the image is in is a segment unlikely to cause too many problems.
Even if the image does fall inside the 144-148 Mhz band, the offset is 910KHz which is not an exact 25kHz channel step.
Traditional 10.7 IF designs use relatively wide bandwidth ceramic filters (typical 110Khz at -3db) and will have problems with both adjacent channel and image signals over a number of channels. By using narrow width 455Khz filters any image problems will be limited a single channel.
In practice, image rejection has proved to be a 'non-problem'.
At the outset of this project there was also a question about the stability of the VFO when used with narrow band filters, but again, in practice, this was never found to be a problem with a SA602 oscillator, which has proved to be extremely stable.
The design also uses a simple means of reducing gain when very close in to the transmitters, by dropping the supply voltage to the RF and SA605 IF chip. This method is so effective in killing the signal there is less need for expensive screening and enclosures. The downside is that a separate SA602 VFO is required with a regulated supply.
Output from the SA605 is via the high speed RSSI pin, this gives 0-5v DC output for a 90db signal strength range and, since the RSSI voltage is unfiltered, demodulated AM. The DC drives an audio S-meter (by G4ODM), the AM modulation is taken to a conventional NE386 amplifier.
Earlier designs incorporated a front end stage to reduce in-band radiation from the VFO into the antenna, (rather than provide any gain). This has been removed, because the RF output from the SA602 VFO is very low, and the potential for interference in a contest is considered to be remote.
The downside of removing the front-end stage was the slightly degraded performance of the RX when very near a TX (i.e. closer than 5-10m). In a normal Orienteering event, where the TX control is quite visible, no problems have been reported, but if the TX is hidden, then the optional attenuator will allow you to reliably DF, 'inches' from the TX antenna!
The pcb board will run with an IF of 10.7 Mhz (to be published later ) as an alternative to 455Khz.
Enclosure
A lightweight, low cost folded aluminum box is now used for the enclosure. The box allows good access to the components, the PCB is nice and rectangular. It provides more-than-adequate RF screening. There is plenty of space under the board for user variations of batteries, switching and s-meters etc.
The box is clearly not watertight. To protect from rain simply, cover with a small plastic bag or cling-film!
The SA605D IC
The main IF chip is only produced in surface mount form. Do not despair! working with SMT components is not that difficult. You just need soldering iron with a small bit ,some tweezers and a MUST HAVE is - de-soldering braid. Tin the SMT pads first, remove excess solder with the braid. Tack one leg of the IC using the tinning, then solder in the IC, tidy up the inevitable solder mess, again using the braid... check for solder bridges ... its as simple as that.
VFO Tuning
The capacitor C13 (see schematic) mainly dictates the bandspread of the VFO.
A bandspread of 1Mhz of less will mean the tuning is not be too sensitive, even using a small 20mm knob.
A narrow tuning range will give the best overall performance, fine if the receiver is solely used for IARU ARDF band.
As the image frequency is not suppressed you can use this 'feature' to good use i.e. if the VFO tunes approximately 1Mhz, you can cover most of the 2Mhz band by tuning the VFO from approximately 144.5 - 145.5Mhz.
I used this feature to allow a ROX to be used for club mobile foxhunts on 145.425 in addition to the Radio 'O' band of 144.5-144.8Mhz without re-aligning the receiver . It is an unconventional solution, but works well!
The Enclosure
The board is mounted on 20mm standoffs. The tuning and gain potentiometers are 17mm conductive plastic types mounted vertically through the board. Use surplus component leads to connect the lead-outs to the board. Miniature carbon potentiometers types can also be used. A pot with 20mm dia body should just fit.
A rechargeable PP9 battery fits snugly under the board with a soft plastic pad sandwiched between. A small strip of Velcro on the end of the battery is also used for extra security.
The receiver consumes 15-18mA so a cheap fully charged 120mAh NiCad PP9 will last a radio 'O' contest with spare capacity.
A 2.1mm power socket is wired across the battery for charging.
A PP3 battery lead fitted with 2.1mm power plug can be used in conjunction with a wall charger BUT THE LEADS WILL NEED TO BE REVERSED e.g. RED to the negative terminal of the power plug . Make sure the lead is clearly marked *** FOR CHARGING ONLY ***.
The mechanical filters that I have successfully tried are TOKO type HCF2455D and HCF2455E.
These are 4 element types, respectively 10 and 12 Khz wide / 6db BW.
If you use other types R3 (12K) may need adjusting for best gain /stability.
Producing the PCB.
I use 'Press-n-Peel' for transferring the track pattern to pcb.
Drill the four corner mounting holes with a 2mm drill ( enlarged later). These holes can be used to accurately align the top and bottom pcb patterns. (see below for Press-n-Peel tips).
Component lead holes are 0.8mm unless shown otherwise. The holes for the mic socket , coil cans, stand-offs, and potentiometers will all need enlarging.
Cut the notch for the antenna socket.
Solder in the SA605D and other SMT components, it's easier to do with plenty of space around them.
The SMT de coupling capacitors can be replaced with small ceramic lead thru' types, but be warned, it likely to be a long and fiddly job.
Next the components that have leads soldered on both sides - C4 C9 R3 C30
It may be necessary to raise the capacitors slightly to ensure the top solder connection can be made.
Then the remainder in any sensible order!
Don't forget the varicap soldered on the underside of the board. This is the smallest component used in the design (and in retrospect, larger alternatives might have easier to fit!)
Setting up.
Some part of the VFO tuning will always be in the 2m band, so you can use a 2m receiver to align the VFO. e.g
For operation in the 144.0 to 144.8 segment, the VFO will be 144.800-0.455=144.345Mhz at the top of the band. Note that the VFO adjustment of is quite coarse, so it is easy to miss the oscillator signal using a 2m receiver, unless you adjust L3 very slowly
Peak L1 in the centre of the band spread.
Adjust RT1 for a comfortable volume at maximum gain.
Adjust RT2 so it just stops the squeaker from running, when there is no signal.
Using.
The squeaker will only trip-in when the RSSI is over 0.7v. This is useful to 'turn off' the squeaker, but a signal at this level is still very audible
When taking a bearing use the gain control to get the squeaker to trip-on in one direction, this makes it easier to assess the transmitter location.
Otherwise, with only two controls there is no more to say!
In-use reports (last updated 12th June 2003)
Real time testing has been very successful, mostly on local mobile foxhunts The ROX was never intended to be used on FM foxhunts, where the search area is 25km X 25Km, but it has now replaced my aging modified 2m handy as the primary receiver. I have no hesitation in saying, used in conjunction with the right brain and body (!) it would be very competitive in an IARU Rules contest.
- August 2002 to February 2003. Design period followed by extensive local field tests.
- 15th Feb 2003, Snelsmore Common, Newbury, Demonstration for Orienteers. Several units used very successfully by newcomers to ARDF
- 23rd Feb 2003, BARC Mobile Foxhunt (FM). Used by G3ZOI from start to finish and came second.
- 23rd Mar 2003, Full ARDF event, Kinver Million. All 5 TXs found (in overtime !) RX proved to be highly effective. Lack of a balun on the antenna resulted in a severe squint, now corrected.
- RSGB contest at Star Posts Bracknell. ROX-1 used by 2nd and 4th placed contestants.
- 11/04 RSGB contest at Blackwod Forest. Won by a ROX-2 user (Mike Dunbar)
- 03/05 RSGB contest at Burghfield Common. Won by a ROX-2 user (Dave Williams).
Downloads for version 2.02, in PDF format.
CLICK HERE for latest corrections and additions
Press-n-peel Tips
All the prototype boards were produced with Press-n-Peel iron-on sheets.
Ignore the articles about using paper or overhead transparency sheet as a cheap alternative, they do not work! Overhead transparency sheet deposits a invisible film of plastic on the board which is more resistant than the ink print!
The pdf template is formatted for an A4 sheet, but only requires one third of a P-n-P sheet cut lengthwise, (to make the expensive sheets last longer!)
Photo Left
Tape the P-n-P strip to an A4 paper backing sheet on the top edge, with a 5mm top margin.
Without the backing paper the heated P-n-P sheet can distort and stick to the laser printer rollers and believe me, it's a big task to remove it. With the backing paper, no problems have been experienced.
A small travel iron has been found to be easier to use than a regular sized iron.
Aligning the top and bottom sheets onto the board requires extra care.
Align the Press-n-Peel sheet using the four mounting holes and temporarily secure to the board with a spring binder clip on one edge. Remove the clip after the ironing process has stuck part of the sheet to the board.
I attached and removed the pattern, one side at a time.
For best results the pc board has to be ultra clean. Track pattern are usually reproduced without problem even the very fine track for the SMT components but larger areas are more prone to problems. Some touching up with a etch resistant pen is likely. Touch up AFTER both sides have been printed, otherwise the ink will run when heated.
| 
