DL4YHF's Amateur Radio Archive:
QRP Homebrew Projects

Main site: http://www.qsl.net/dl4yhf/qrp/index.html
Backup: http://www.mydarc.de/dl4yhf/qrp/index.html


QRP Indoor Loop
Miss Mosquita (40 m QRP rig) with digital frequency display

Last updated: May 2012

QRP Indoor Loop

For QRP operation with limited space, a compact and lightweight 'magnetic' loop antenna was built. For the first design (which leaves room for improvement), a long thin piece of aluminium sheet was used (3.14 meters total circumfence, 50 mm wide, 2 mm thickness):

Unlike most classic loop designs, the tuning capacitor and the smaller coupling loop are both at the base of the antenna. The tuning capacitor with approximately 20 ... 250 pF should be replaced with a 'butterfly'-design, but these are either difficult to find or too expensive, so a 'normal' variable capacitor scrapped from an old antenna tuner was used. With this arrangement, the resonant tuning range was 7 to 21 MHz. With a "better" capacitor (lower minimum capacity), tuning above 21 MHz could be easier.

The small neon lamp on the right side of the capacitor initially acted as a tuning aid. One of its legs is connected to the outer loop, the other leg is only capacitively coupled to the other electrode of the tuning capacitor. It later turned out that adjusting the loop for resonance 'by ear' brings the SWR low enough before transmitting (at QRP levels!), the 'fine tuning' should be done with an SWR meter.
The coupling loop's optimum length (circumfence) was found to be 0.6 meters, which is roughly 1/5th of the outer loop's circumfence, which agrees with many other 'magnetic loop recipes' (for example Rothammel's Antenna Book).
Note that unlike other designs, the small coupling loop of this antenna does not touch the larger outer loop ! If it would, the only point where both loops may touch would be the topmost point, which is 'electically neutral' (voltage minimum of the outer loop).
Due to the lack of an electric connection between the loops, there is also no need for a balun !

The plastic 'insulator' at the bottom holds both loops, and the tuning capacitor in place. The entire loop (including insulator) can be rotated on the wooden base, using a single screw visible in the center of the last photograph.
To eliminate the need for a vernier drive, and to stay away from the tuning capacitor, an plastic lever is used to operate it.

Ideas / possible improvements

The electric conductance of the aluminium loop (measured at DC by passing 2 amperes through the loop, and measuring the voltage - 3 mV - across the loop close to its ends) was found larger than it should have been. Calculation:
       sigma = ( Current * Length ) / ( Voltage * Conducting_Area )
       sigma = ( 2 A  *  3 m ) / ( 0.003 V * 0.002 m * 0.05 m ) = 20 * 10^6 Siemens / meter
     (In Germany, 'sigma' [σ] denotes the electrical conductance.
      In the English literature, 'kappa' [κ] seems more common.)
The electrical conductivity of pure aluminium would be 36.6 Mega-Siemens per meter; and a loop made of pure copper would be even better (sigma = 58 Mega-Siemens per meter).
Since most of the HF current only flows in the outer few micrometers of the conductor, the loop doesn't have to be "solid". A loop made of a long (bent) piece of plastic, carrying a long strip of copper foil (say 70 micrometers thick, and 50 mm wide) would be even lighter than this loop. Experimentation will tell if the reduced loss in the loop is worth the effort, considering the loss in the capacitor. A thin copper loop in combination with a 'butterfly' capacitor, or -even better- a variable vacuum capacitor will help to cut the losses in the loop even further. But at some point losses in the vincinity of the loop (walls, floors, ceilings, eddy currents in metal objects, etc) will dominate, and it possibly doesn't pay off to reduce the losses in the antenna itself.

Miss Mosquita

Miss Mosquita was the first QRP transceiver which I didn't design myself. But this little cutie by DL-QRP-AG caught my eye at a hamfest, so I bought the kit (for a very fair price) and assembled it in some hours on one weekend. The construction is very simple, and if you follow the detailed step-by-step instructions there should be no problem to get it running - my Miss Mosquita worked properly right from the start. Only for personal taste, I made two minor modifications which are described below.

But first of all, this is how the transceiver looked (with the prototype of the first 'DL4YHF-' LED frequency counter attached) during the first QSOs - no enclosure yet, because I wanted to get "on air" with it as fast as possible :

The seven-segment frequency display is my first prototype (details about the counter are here; the counter which can be ordered as a kit from QRP-project is basically the same with a slightly modified layout).

The counter (-prototype) fits easily inside the transceiver's housing (which is available from DL-QRP AG), there was even enough space under it to mount the jacks for headphone and morse key on the front panel under the display:

The boards are mounted on a copper-clad board which was inserted in a slot of the aluminum housing (like a "motherboard" holding the transceiver- and the TRX board in place). Three screws soldered (!) to the "motherboard" are used as bolts which hold the Mosquita PCB in place. I didn't remove the black adhesive stuff from the ex-photographic PC board material.

The top view shows the Mosquita board with PA transistor mounted on the rear panel, and the counter mounted at the front side:

The final result looks like this, with red plexiglass in front of the 7-segment display to increase the display contrast in a bright environment (the photograph was taken in the light of a strong halogen lamp):

Mosquita modifications

Again, Miss Mosquita worked right from the start. I made the following modification just for personal taste:

The CW pitch was a bit low (audio center frequency about 400 Hz). Since I prefer 650 Hz for CW, I modified the BFO, TX-mixer a bit, and reduced the audio bandwidth to something about 400 Hz:

  • reduced C11 (series capacitor for crystal in the BFO) from 100 pF by 56 pF, which increased the audio note and removed the "grumbling" low-frequency tones;
  • changed L9 in the TX-mixer from 15 uH to 4.7 uH, because after the BFO-mod described in the previous step, the TX frequency was a bit off frequency (this made RX and TX "transceive" again)
  • increased C7 and C8 in the 4 MHz IF filter from 220 pF to 330 pF to reduce the CW filter bandwidth a tiny bit
  • replaced C45 (100 nF) in the TX-driver stage by 2.2 nF, because the gain of T4 and T5 (TX preamp and driver) was a bit too high for an easy adjust the output power. With 2.2 nF (= -j*10 Ohm) in the emitter, there is less gain, but still enough driving power to reach 5 watts, and the power is a bit easier to adjust with P3 now.
  • replaced R10 (22 kOhm in the audio amplifier) by a series-combination of 1 k + 18 k, with two antiparallel silicon diodes to ground. This reduces loud bursts from strong stations, before the AGC kicks in. 100 nF parallel to the diodes give an additional lowpass to reduce broadband noise from the IF amplifier :

The VFO can be easily tapped for the (LED-) frequency counter between R15 (820 Ohm) and P3 ( 250 Ohm to ground).

Digital frequency readout for Miss Mosquita (DL4YHF counter)

I found the oscillator in the TX mixer running at 3.9990 MHz, so I added this frequency in the table of "preconfigured" offset frequencies in the PIC16F628 firmware (use "counter2.hex", also for the counter kit from QRP project). If your counter is preprogrammed with an IF offset of ZERO, here are the steps how to get it running properly with Miss Mosquita's 3999 kHz TX-mixer:

  • Connect the pushbutton to (DL4YHF-) frequency counter. You only need that button temporarily, so there's no need to install it in the front panel !
  • Turn Miss Mosquita (and the counter) on. If there is no IF offset programmed yet, the counter may show something like "3.0310" (MHz) now.
  • Hold the counter's button pressed until the display shows "Prog", then release it
  • Shortly press the button a few times until the display shows "tAbLE"
  • Now press the button a bit longer, until the selected menu (here: "TABLE") starts flashing. This takes you into a sub-menu where different preconfigured frequency offsets can be recalled.
  • Shortly press the button a few times until the display shows "3.9990". If you miss that entry, you have an older firmware (which doesn't have this "Mosquita-specific" IF offset - see note below on how to use that firmware for Miss Mosquita)
  • Now press the button for a second (until the display flashes "3.9990"). Release the button when the display flashes. After this, the counter asks you if that frequency shall be added ("Add") or subtracted ("Sub"). For Miss Mosquita, the IF offset must be added, so hold the button pressed for a second while the display shows "Add". When "Add" begins to flash, the counter has permanently saved the new offset. Release the button, and you are back in normal operation mode.

If everything went right, the counter shows something like "7.0300" (MHz) now. Bingo.. connect an antenna and listen for QRP stations, or start calling CQ on that frequency ;-)

If your counter is equipped with an old firmware, there may be no entry for "3.9990 MHz" in the preconfigured table of frequency offsets.
In this case, you can enter the offset as described in the firmware documentation for DL4YHF's PIC-based frequency counter (temporarily connect the counter to the oscillator in the TX-mixer, which delivers the 3.9990 (+/-) reference). If the relative link doesn't work, try this absolute link.

Additionally, you can activate the power-saving option in the counter's setup menu (toggle "PSAVE" and "NoPSV" in programming mode). If the power-saving option is enabled, the display will be turned off if the VFO frequency doesn't change by more than a few dozen Hertz within 15 seconds, and turned on as soon as the VFO frequency changes again (=operator turning the VFO knob). During power-saving mode, the display flashes up once every 10 second to show that the transceiver is still on. This mode can save up to 15 mA of battery current. In the author's prototype, the average total DC current during RX for receiver+counter dropped from 45 mA to 32 mA with the power-saving mode.