A Wideband Low Voltage Offset Attenuator by Neil Robin, WA7NBF Revised: April 7, 2010

Overview:

A little background, first. 

Anyone who's attempted VHF or UHF hidden transmitter hunts with a standard hand held transceiver or scanner recognizes they're not well shielded.  When you get close to the hidden transmitter, the signal begins to enter the radio even when you add attenuators between the antenna and radio.  The problem is that the radio's are not well shielded because its not a requirement for typical use.  Afterall, the manuafcturer needs to keep costs low to remain competitive.  A strong RF source will blast right through the plastic case  and reach the detector circuits directly.  Your "S" meter is saturated, you've lost directivity from your yagi or whatever you're using for a directional sensor. 

On the otherhand, a well shielded radio has a related problem.  Usually it can be used effectively with external passive attenuators  but many have found having a continuously variable attenuator allows you to adjust the signal level as a manual gain control so the "S" meter reads half scale.  Then as you move closer, the meter will move higher and you'll be able to bring it back to the middle range by adjusting the manual gain to a slightly lower level.  Making note of the gain settings as you progress tells you how close you're getting to the fox.  With experience, you'll be able to judge your distance at any time by observing the gain control setting.  Of course, effective radiated power of the transmitter can vary but I've found that it doesn't change that much particulary when the power of the transmitter is stated beforehand and you take that into consideration.  However, most of us don't have well shielded radios unless they're designed for activities like foxhunting.

Enter the Offset Attenuator (OA)

A very clever solution developed many years ago has been to mix the incoming RF signal with a LO (local Oscillator) to create the sum and difference frequencies.  The intensity of the output signal can be well controlled (maybe a slight over statement) by the amount of LO injection into the mixer.   The very strong fox transmitter RF cannot get into the receiver because the receiver is now tuned to the sum or difference frequency and not the F_in.  If the resulting mixing products are too strong, simply reduce the amount of LO signal and you can bring the "S" meter into the mid range position.  This solves the two common problems found in handheld receivers listed above.

Most Amateur Radio Direction Finding (ARDF) is on 2 meters.  A common foxhunting frequency is 146.565 Mhz.  If our offset attenuator has an LO frequency of 2.00 Mhz  then the output will be:

F_out1 = 146.565 + 2.0 = 148.565 Mhz
F_out2 = 146.565 - 2.0 = 144.565 Mhz

Now you'll need to tune your radio to one of these two frequencys to hear the fox signal.  Note that one is outside the amateur band but many radios have receive capability for out of band signals.  Any other signals that may be residing on these frequencies will probably get through w/o much attenuation and can be an interference issue.

This is all common knowledge for foxhunters who have tried to use handhelds for hidden transmitter hunts.  What are some of the problems?

Problems to solve or improve upon

My first OA worked great but had several undesireable characteristics:

1. As I turned down the LO gain control it would reach a point where you couldn't lower the sensitivity any further.  The LO seems to be leaking through even with the pot all the way down.  Unlike international ARDF rules, some foxes are so concealed that you have to practically touch the antenna to find them.
2. It used a 9 volt battery which when left on runs down after a few days.  Do I really need such a high DC voltage?
3. It was not broadbanded.  I wanted to be able to use the OA on other VHF/UHF frequencies in addition to 2 meters.  Particularly the marine, wildlife and aircraft bands.  Their may also be times where you want to chase interference outside the amateur bands.
   
4. At times the sum and difference frequencies land on an undesireable frequency which may have interference.  The ability to change the LO frequency easily would be desireable if this problem surfaces in the field.
5. With a strong LO signal the chance of beating against a harmonic  increases.  IE,  F_out = F_in + 2F_LO or F_out = F_in - 2F_LO.  This increases the chance of interference.
6. With no selectivity on the front end, strong out of band signals can blast their way in causing intermod and general overdrive.
7. Accidently transmitting into the attenuator will usually take out the mixer diode.  A serious problem when in the field....it can ruin your day of hunting.
8. Need to use a metal housing.  Undesireable signals will be unwelcome without it plus you want to keep the LO oscillator tightly controlled

Solutions:

I attempted to solve a few of the problems but theirs still plenty left if you want to try your luck.  Referencing the items above:

1. In looking into the overdrive/leakage of the LO I realized that another effect was taking place.  As the incoming RF from the fox was getting strong, it would act as the LO and mix with the very weak internal LO signal. Two ways to solve this problem:
   1. Reduce the LO amplitude to as low as possible.  A few hundred millivolts is enough.
      I started exploring low voltage oscillators and found a chip that can be powered with as low as 0.7 volts DC.  Its the ALD110900 E-pad dual MOSFET which is designed for zero threshold.  Mouser part #585-ALD110900APAL which costs about $2.35 each.  The complete LO can be built with one of these 8 pin DIP chips.  I decided to run it off of 1 1/2 volt AAA battery.
   2. Shield the LO completely.  You can place the entire oscillator inside a small shielded box or make something out of raw PCB material.  So far I havn't resorted to this but the next time I build an OA I will consider it
2. From #1 above I now use 1 1/2 volt battery and single cell holder.
3. This design works well within all the VHF and lower UHF bands.  How, I avoided the use of any frequency sensitive circuit elements.  The first unit I built had an inductor for the return DC path in place of R3, see schematic.  By using high quality capacitors with low X_c at all the frequencies of interest along with resistors only you enable broadband performance.  The path that handles the wideband signals is made up only of C1, C2 and D1.  Keep these leads short with low stray capacity.
4. I built the unit with a socket for the crystal.  I carry 1, 2 and 4 Mhz "rocks" in case I need to change LO frequencies.  Sure you could put in a switch but that would complicate the shielding and open the possibility of more LO leakage.  I've found that 2.00 Mhz works well most of the time.
5. If you keep the LO as low an amplitude as possible to start with and don't provide excess feedback for the oscillator you will minimize the harmonics.  This has been a non-problem for the type of hunting I do.  If you were concerned about this you could add a bandpass filter in the LO path...but do you really need it?
6. In situations like this a helical filter for the two meter band with bandwidth and "Q" such as to act as a good filter can be plugged into the input.  I 've also used a tapped 1/4 wave transmission line made out of RG-58 and shorted on the low impedance end near the tap point.  The line is less bulky and alot easier to build but the helical filter is easier to tune.  This is usually only a problem when you're hunting near strong radio transmitters like on a hilltop.
7. I've done it...everybody does.  1N4152 diodes are cheap.  Pin diodes and clamping diodes back to back have been suggested but you'll probably blow them as well.  So far, I've done nothing to correct this weakness...but I have lots of spare diodes.
8. Enough said, always use a shielded box

Schematic:

The complete schematic can be found here.

All the circuitry below R2 potentiometer is for the oscillator and power.  R2 is available from Radio Shack and inexpensive.  D1 is the mixer diode with the LO injected via R1.  The return path is through R3.  As the wiper of R2 travels more negative a small DC bias is placed on the diode to help its conductivity.  When the wiper is in its low sensitivity position, the bias is removed but a strong RF signal coming in will bias the diode with C1 holding the offset voltage.  Generally, its prefered to keep the oscillator circuit ground separate from the chassis but you can try it either way to minimize leakage of the LO.

Summary:

No circuit board has been made but I'd be happy to give guidance for anyone who wishes to build it.  With a directional antenna, a receiver and an OA you have all the equipment you need to be a serious foxhunter or interference sleuth.

KA7CSZ looking for foxes	Offset Attenuator velcro'd to Yagi antenna boom