The introduction of DSP ultralight radios using the Silicon Labs Si473x chip about four years ago proved to be an experimenter's playground. The new design allows for a simply-wound ferrite loopstick antenna with a single coil winding, connecting to the radio's circuit board on just two soldered pads. No more complicated coil with a secondary local oscillator winding or multiple coil taps. The radio's chip tunes the single coil inductance for longwave through the top of the mediumwave band. Replacing the ferrite antenna with something else is now easily possible.
The common mod to these radios has been to remove the internal ferrite loopstick and replace it with a longer one, often mounted outside the radio on top. This new ferrite loop is then tuned by the circuitry of the radio just as the old loop was. The longer (or beefier) ferrite gives the radio much greater sensitivity. What about an air core loop?
Early on in reviewing the Si473x chip's documentation, I noticed the manufacturer showed an option of using an external, air core loop antenna as a substitution for the internal ferrite loop. Their suggestion was for a loop of minimal turns connected to the circuit board's antenna terminals through a 1:5 winding, 25x step-up ferrite core transformer, thus providing the correct coil inductance (180-450 micro-Henries). It was apparent that using a full inductance loop was also possible. This would also result in greater signal gathering ability.
So, let's get started.
Currently I own the Tecsun PL-380 and Eton Traveler III DSP radios. After removing the cases of the two and examining their internal ferrite loopsticks, it was evident that the Eton's loopstick would be the easiest to remove. One end is held down with some flexible sticky glue and is easily removable with a small jeweler's screwdriver. Pull the ferrite bar with coil out of the case and unsolder the two pads on the circuit board.
I decided to make this a totally breadboard-type operation, so I drilled two tiny holes through the top of the radio's case and soldered two short pigtail leads through them to the board. I then connected two micro alligator clips to the ends of the pigtails. These will be how we attach the radio to the loop we will build. Using clips, it will also enable easier experimentation with other loop devices.
Initial experiments with a 12 inch PVC loop of 16 turns (about 215 micro-Henries) showed excellent signal strengths. In researching optimum coil inductances for the Si473x chip, the figure of 400 micro-Henries seemed to be mentioned in the forums, so it was decided to build an 18 inch loop of 17 turns***, which calculated out to about 410 micro-Henries. The additional loop area (2.25x greater area) would also provide better signal gathering ability.
***17 turns proved to be way to many turns, as I will explain below.
About the simplest box loop frame to build is one from 1x2 inch lumber. Cut four pieces to make a square 18 inch frame and nail (I also glue) the corners. Using a fine saw or hacksaw and file, notch each corner about 3/4 inch as shown. This will contain the coil windings. You will need about 65-70 feet of insulated wire, about 22-24 gauge should do it.
Now, let's talk about wire. Unless someone gives you a quantity of insulated wire for free, the least expensive wire source I know of is to buy 4 conductor telephone cable. Remove the outer sheath and you have four solid, insulated 22 or 24 gauge copper conductors. Using a straight edge razor held between your fingers just right against a hard table surface, draw the sheathed wire under it, scoring it lengthwise, and you will find the sheath peels away quite easily. A 50 ft. roll will cost you about $8 at Home Depot. That's 200 ft. of wire for about 4 cents a foot. I have used this wire in loops for years, and in the loops described here. Regular hookup wire at Radio Shack is about double that cost.
Before starting the loop winding and to secure the coil ends, I drilled two small holes through the bottom of the frame. Near these holes on the inside, drive two small headed-nails which will secure the wire ends, as shown in the picture just below.
I decided to close-wind the coil, that is, to place each turn right next to each other. I haven't found this to be a huge detraction in passive loops. The close turns tend to increase the inductance a bit, and add a small amount to the inherent internal capacitance of the loop, but not as much as you might think.
Secure one end of wire to one nail through the hole, and start winding. I brought the last turn back through the other hole and secured it to the other nail. Be sure you've left two pigtails off the nails and strip the wire ends. Make a small loop at each wire end, 1/8 inch diameter or so. We will attach the alligator clips to them.
After winding the frame, I used electrical tape to wrap the sides in strategic places to keep the loop turns together. Over time, the loop turns will tend to relax some and get loose, and the tape will help to keep the loop turns corralled.
Last, cut two 18 inch side slats from 1x3 lumber for the bottom of the frame, as shown on the completed loop photo. They will provide a slot to hold your radio and also broaden the base of the frame so it will sit upright on a flat or nearly-flat surface.
You are done! Time to try it out. Place the radio in the slot and connect the micro clips to the loop ends. Turn the radio on (use headphones), tune to a known station and rotate the loop for maximum signal. This will be off the ends of the loop.
Disappointment at first. 17 turns wouldn't tune to 1700 KHz, or anything above about 1580 KHz. Signal strengths were also not much better than the 12 inch loop. Too much inductance, apparently. About all I could get at the upper end of the band was a weak KMIK-1580 (50 KW), in Tempe, AZ, at 139 miles. This station came in with tremendous signal on the 12 inch loop. Mexican station (ESPN Radio) XEKTT-1700 (10 KW), between Tijuana and Tecate along the US border (169 miles) was weak but copyable on the 12 inch loop. It was non-existant on the 18 inch with 17 turns. In fact frequencies above 1600 KHz were almost full quieting, indicating that the radio was not tuning this high. Recalculating, I found that 12 turns was about the equivalent inductance as the 12 inch loop, so I removed 5 turns. The loop now tunes perfectly between 530 and 1700 KHZ.
What a difference. Huge daytime signal increases over the radio's stock internal ferrite antenna are apparent using this loop, and a noticeable increase over the 12 inch loop. KNX-1070 (50 KW) at Los Angeles, at 237 miles distant now shows an RSSI strength reading of 44/15 and is beautiful copy. It was not receivable before with the stock ferrite antenna, at 15/02. Calculated field strength for KNX-1010 at this location is on the order of 0.03 mV/m, an extremely small signal, not receivable at all on a stock ultralight. The Panasonic RF-2200 and the Eton E1 will receive it fairly well. The Eton Traveler 3 now beats them both.
Nulls are well defined with this loop, and fairly narrow in beamwidth. I am able to receive KTIE-590 (2.5 KW), San Bernardino, CA at 178 miles and KSUB-590 (5 KW), Cedar City, UT at 284 miles by rotating the loop and nulling out each station. The stations are approximately 90 degrees to each other. Calculated field strengths here are extremely weak, at 0.02 mV/m for KTIE and 0.03 mV/m for KSUB.
KTNN-660 (50 KW), Window Rock, AZ at 327 miles produces a decent daytime signal as well. I have heard KKOB-770 (50 KW) at Albuquerque, NM (447 miles) on certain days. It's calculated signal strength is extremely low, at 0.01 mV/m. KALL-700 (50 KW), N. Salt Lake City, UT (515 miles) is an easy catch. Most of the big-gun San Francisco, CA stations, though usually weak, (about 525 miles) are receivable as well: KNBR-680, KGO-810. KCBS-740 is somewhat masked by semi-local KDIR-740 (1 KW) in Phoenix.
I had a small, plastic entertainment center receiver loop frame I've been saving. It measures 5.5 x 6 inches. You've seen them - they are everywhere until you want one. I found one at a Goodwill store some time ago for about 50 cents and stockpiled it. Calculating the inductance required, I wound 31 turns of wire on it and connected it to the Eton via the clips. Signal strengths were definitely down from the 12 and 18 inch loops. They were slightly above the stock ferrite loopstick, however. This little loop might make a usable antenna for playing the radio in the house.
Here's a curious tuning tip that may not be apparent at first. It occurred to me that when the radio is attached to the loop and powered on and tuned to a frequency, say 590 KHz, the loop should act the same way as a passive loop does which has an external variable capacitor across it tuned to 590 KHz. I placed a second radio in the slot within the loop and tuned it to 590 KHz. The pronounced signal at 590 KHz was also induced into the second radio. Other frequencies were not. If I tuned the Eton off 590 KHz, the second radio, still tuned to 590 KHz went quiet. Interesting, but it makes sense. Using this technique, we have a way to use this loop arrangement also as a passive loop for another radio, with the benefit that we know precisely the tuned frequency of the loop.
The 12 inch loop is made from 1/2 inch PVC. Purchase four 3-way corners. The pieces fit together tightly without glue and allow it to be disassembled. To each corner I glued a grey plastic, 3/4 inch x 1-1/2 inch sprinkler system nipple. We will use the thread indentations on these nipples to create an perfectly and evenly-spaced loop of 16 turns. Drill tiny holes through the nipples at each end to secure the wires. The wires are further secured with a zip-tie.
The 12 inch loop, with neatly-spaced turns, performs remarkably for its size.
If you are an experimenter and don't mind cutting into your DSP ultralight, try building a hardwired loop for it. It will open up a new world of listening. 12 or 18 inch loops will result in signal sensitivities exceeding those of the stock Panasonic RF-2200 and Eton E1 radios. Coupled with the excellent stock sensitivity and selectivity of the Eton Traveller III, you will have a great performer.
The common mod to these radios has been to remove the internal ferrite loopstick and replace it with a longer one, often mounted outside the radio on top. This new ferrite loop is then tuned by the circuitry of the radio just as the old loop was. The longer (or beefier) ferrite gives the radio much greater sensitivity. What about an air core loop?
Early on in reviewing the Si473x chip's documentation, I noticed the manufacturer showed an option of using an external, air core loop antenna as a substitution for the internal ferrite loop. Their suggestion was for a loop of minimal turns connected to the circuit board's antenna terminals through a 1:5 winding, 25x step-up ferrite core transformer, thus providing the correct coil inductance (180-450 micro-Henries). It was apparent that using a full inductance loop was also possible. This would also result in greater signal gathering ability.
So, let's get started.
Currently I own the Tecsun PL-380 and Eton Traveler III DSP radios. After removing the cases of the two and examining their internal ferrite loopsticks, it was evident that the Eton's loopstick would be the easiest to remove. One end is held down with some flexible sticky glue and is easily removable with a small jeweler's screwdriver. Pull the ferrite bar with coil out of the case and unsolder the two pads on the circuit board.
 |
| Ferrite loop removed. |
I decided to make this a totally breadboard-type operation, so I drilled two tiny holes through the top of the radio's case and soldered two short pigtail leads through them to the board. I then connected two micro alligator clips to the ends of the pigtails. These will be how we attach the radio to the loop we will build. Using clips, it will also enable easier experimentation with other loop devices.
 |
| Holes drilled and pig-tails and clips attached. |
Initial experiments with a 12 inch PVC loop of 16 turns (about 215 micro-Henries) showed excellent signal strengths. In researching optimum coil inductances for the Si473x chip, the figure of 400 micro-Henries seemed to be mentioned in the forums, so it was decided to build an 18 inch loop of 17 turns***, which calculated out to about 410 micro-Henries. The additional loop area (2.25x greater area) would also provide better signal gathering ability.
***17 turns proved to be way to many turns, as I will explain below.
 |
| Radio complete. |
About the simplest box loop frame to build is one from 1x2 inch lumber. Cut four pieces to make a square 18 inch frame and nail (I also glue) the corners. Using a fine saw or hacksaw and file, notch each corner about 3/4 inch as shown. This will contain the coil windings. You will need about 65-70 feet of insulated wire, about 22-24 gauge should do it.
Now, let's talk about wire. Unless someone gives you a quantity of insulated wire for free, the least expensive wire source I know of is to buy 4 conductor telephone cable. Remove the outer sheath and you have four solid, insulated 22 or 24 gauge copper conductors. Using a straight edge razor held between your fingers just right against a hard table surface, draw the sheathed wire under it, scoring it lengthwise, and you will find the sheath peels away quite easily. A 50 ft. roll will cost you about $8 at Home Depot. That's 200 ft. of wire for about 4 cents a foot. I have used this wire in loops for years, and in the loops described here. Regular hookup wire at Radio Shack is about double that cost.
 |
| 18 inch box frame showing notched corners. |
Before starting the loop winding and to secure the coil ends, I drilled two small holes through the bottom of the frame. Near these holes on the inside, drive two small headed-nails which will secure the wire ends, as shown in the picture just below.
I decided to close-wind the coil, that is, to place each turn right next to each other. I haven't found this to be a huge detraction in passive loops. The close turns tend to increase the inductance a bit, and add a small amount to the inherent internal capacitance of the loop, but not as much as you might think.
 |
| Bottom detail showing drilled holes and nails. |
Secure one end of wire to one nail through the hole, and start winding. I brought the last turn back through the other hole and secured it to the other nail. Be sure you've left two pigtails off the nails and strip the wire ends. Make a small loop at each wire end, 1/8 inch diameter or so. We will attach the alligator clips to them.
After winding the frame, I used electrical tape to wrap the sides in strategic places to keep the loop turns together. Over time, the loop turns will tend to relax some and get loose, and the tape will help to keep the loop turns corralled.
 |
| Loop winding completed. |
Last, cut two 18 inch side slats from 1x3 lumber for the bottom of the frame, as shown on the completed loop photo. They will provide a slot to hold your radio and also broaden the base of the frame so it will sit upright on a flat or nearly-flat surface.
You are done! Time to try it out. Place the radio in the slot and connect the micro clips to the loop ends. Turn the radio on (use headphones), tune to a known station and rotate the loop for maximum signal. This will be off the ends of the loop.
Disappointment at first. 17 turns wouldn't tune to 1700 KHz, or anything above about 1580 KHz. Signal strengths were also not much better than the 12 inch loop. Too much inductance, apparently. About all I could get at the upper end of the band was a weak KMIK-1580 (50 KW), in Tempe, AZ, at 139 miles. This station came in with tremendous signal on the 12 inch loop. Mexican station (ESPN Radio) XEKTT-1700 (10 KW), between Tijuana and Tecate along the US border (169 miles) was weak but copyable on the 12 inch loop. It was non-existant on the 18 inch with 17 turns. In fact frequencies above 1600 KHz were almost full quieting, indicating that the radio was not tuning this high. Recalculating, I found that 12 turns was about the equivalent inductance as the 12 inch loop, so I removed 5 turns. The loop now tunes perfectly between 530 and 1700 KHZ.
 |
| The final product, the 18 inch loop with 11 turns. |
What a difference. Huge daytime signal increases over the radio's stock internal ferrite antenna are apparent using this loop, and a noticeable increase over the 12 inch loop. KNX-1070 (50 KW) at Los Angeles, at 237 miles distant now shows an RSSI strength reading of 44/15 and is beautiful copy. It was not receivable before with the stock ferrite antenna, at 15/02. Calculated field strength for KNX-1010 at this location is on the order of 0.03 mV/m, an extremely small signal, not receivable at all on a stock ultralight. The Panasonic RF-2200 and the Eton E1 will receive it fairly well. The Eton Traveler 3 now beats them both.
Nulls are well defined with this loop, and fairly narrow in beamwidth. I am able to receive KTIE-590 (2.5 KW), San Bernardino, CA at 178 miles and KSUB-590 (5 KW), Cedar City, UT at 284 miles by rotating the loop and nulling out each station. The stations are approximately 90 degrees to each other. Calculated field strengths here are extremely weak, at 0.02 mV/m for KTIE and 0.03 mV/m for KSUB.
KTNN-660 (50 KW), Window Rock, AZ at 327 miles produces a decent daytime signal as well. I have heard KKOB-770 (50 KW) at Albuquerque, NM (447 miles) on certain days. It's calculated signal strength is extremely low, at 0.01 mV/m. KALL-700 (50 KW), N. Salt Lake City, UT (515 miles) is an easy catch. Most of the big-gun San Francisco, CA stations, though usually weak, (about 525 miles) are receivable as well: KNBR-680, KGO-810. KCBS-740 is somewhat masked by semi-local KDIR-740 (1 KW) in Phoenix.
I had a small, plastic entertainment center receiver loop frame I've been saving. It measures 5.5 x 6 inches. You've seen them - they are everywhere until you want one. I found one at a Goodwill store some time ago for about 50 cents and stockpiled it. Calculating the inductance required, I wound 31 turns of wire on it and connected it to the Eton via the clips. Signal strengths were definitely down from the 12 and 18 inch loops. They were slightly above the stock ferrite loopstick, however. This little loop might make a usable antenna for playing the radio in the house.
 |
| Mini-receiver loop. A fairly poor performer. |
Here's a curious tuning tip that may not be apparent at first. It occurred to me that when the radio is attached to the loop and powered on and tuned to a frequency, say 590 KHz, the loop should act the same way as a passive loop does which has an external variable capacitor across it tuned to 590 KHz. I placed a second radio in the slot within the loop and tuned it to 590 KHz. The pronounced signal at 590 KHz was also induced into the second radio. Other frequencies were not. If I tuned the Eton off 590 KHz, the second radio, still tuned to 590 KHz went quiet. Interesting, but it makes sense. Using this technique, we have a way to use this loop arrangement also as a passive loop for another radio, with the benefit that we know precisely the tuned frequency of the loop.
The 12 inch loop is made from 1/2 inch PVC. Purchase four 3-way corners. The pieces fit together tightly without glue and allow it to be disassembled. To each corner I glued a grey plastic, 3/4 inch x 1-1/2 inch sprinkler system nipple. We will use the thread indentations on these nipples to create an perfectly and evenly-spaced loop of 16 turns. Drill tiny holes through the nipples at each end to secure the wires. The wires are further secured with a zip-tie.
The 12 inch loop, with neatly-spaced turns, performs remarkably for its size.
 |
| The 12 inch loop. |
If you are an experimenter and don't mind cutting into your DSP ultralight, try building a hardwired loop for it. It will open up a new world of listening. 12 or 18 inch loops will result in signal sensitivities exceeding those of the stock Panasonic RF-2200 and Eton E1 radios. Coupled with the excellent stock sensitivity and selectivity of the Eton Traveller III, you will have a great performer.
