Sep Oct MAR
Previous capture 9 Next capture
2013 2014 2016
2 captures
9 Oct 14 - 10 Mar 16
Close Help

Pa0nhc miniwhip active receiving antenna (30kHz LW-MW-SW 30MHz wideband).
(Vastly improved version of the design from pa0rdt)
Last rev. 2013051
5 (PDF)

Download accurately printable PCB masks by <right-clicking> on "PDF" and click "Save as". Open in Adobe Reader and print.


Miniwhip antenna pa0nhc (earlier version).

The power transistor is soldered to the copper plane for cooling.


Pa0nhc version of the Miniwhip active antenna (L) and combibox (R).
The antenna PCB fits inside a 40mm PVC drainpipe.

In the combibox the mass surroundings on the single sided PCB are soldered onto the groundlugs of the connectors.
REM: For best results ONLY use recommended components (see partslist) and take denoising measures for the coax cable.


- FET and bipolair transistor are now optimally biased fro real class A operation.
    Largely enhanced strong signal and IMD properties.
    Max. input = 8.6Vpp, max. output 4.4Vpp (!).
    No spouriï at 7MHz during good evening CNDX, nor mixing due to very strong local MW broadcast stations.
    Enhanced stability due to reduced VHF gain and double sided PCB.
- Free choice of the FET to be used as long it can draw 40mA peak. 
    Instructions for choosing wich one and how to bias are included (important).
- Easy construction with standard available through-hole parts
    No SMD, or only one "big" SMD part (SOT103-type dualgateMOSfet is preferred).
- Power/output-coupler without transformer
    Really wideband RC coupling. No IMD caused by a transformer.
    No difficult to make narrowband transformer
    Flat frequency response with constant sensitivity from 50kHz to 28MHz.
- Practical suggestions for installing and de-noising the outside of the antenna cable for optimal S/N in city environments.
- Neat PCB design. 
    Sharp, uncompressed, easilly, very black printable 600DPI PDF mask files available. 
Dimensions (outer border lines): 114.935 x 34.29 mm2. 
    Fits inside a 40mm sewage pipe.
    Double sided version needs no cooling star (T2 soldered to copper surface).
    Can be used as as single-sided too (with extra cooling star).
- Can be made using "dead bug" construction if wisely constructed.

Miniwhip active part schematic diagram


Xray view. 
Red=bottom, green=top, blue=holes, yellow=parts.
PCB designed with freeware FREEPCB


pa0nhc Miniwhip Bottom copper (300 DPI).


pa0nhc Miniwhip Top copper


pa0nhc Miniwhip topsilk.



Pa0nhc Miniwhip combiner schematic (combibox).


Combibox bottomcopper 2400DPI for the single-sided PCB.



Combibox topsilk.



Miniwhip cable de-noising suggestions.

This cable screening filtering using ferrite clamp and coax air coil is highly recommended in order to obtian the highest S/N in places with high levels of man-made noise (cableTV, switching power supplies).

On top of that, put a ferrite clamp over the coax immediately at the connector of the miniwhip active antenna, AND immediately at the antenna connector of the receiver (at the end and beginning of the coax).

Do not spare on ferrite ! Use clamps with BIG (16mm) holes for best results.



The 2.1mm power jack is connected to a well filtered 12V to 15V DC supply.
RX is connected left, the Miniwhip ant. is connected right.


The antenna is RC coupled to the Rx, resulting in a constant sensitivity from VLF-30MHz.
No transformer is used, as it is very difficult (impossible?) to produce with really 30kHz-30MHz wideband properties.
But good care must be taken that noise-mantle-currents on the antenna cable canNOT run towards the antenna.

ver.20121105 pa0nhc Miniwhip Partslist. Only use the recommended types of electronical components for good results. 
Download PDF for printing



For other FETs like BF246A, B  etc.

See text

Drill a 5.5mm hole for T1.  Observe pinning!
A SOT103 type dual gate mosfet must be soldered at the UNDERside of the PCB.

A TO72 type dualgateMOSfet or TO92 type Jfet can be at the topside with their legs through te hole. 

IMPORTANT: if a Jfet is used, it should have a Idss of 40mA @ Vds=3V. SEE INSTRUCTIONS BELOW.



VHF small power transistor or switching transistor. For better cooling it is thermally connected to the insulated top copper field by soldering the rim at 4 points to the copper surface. Dissipation=ca. 800mW.



Fast switching diode.



1A diode.


PCB 90deg

Conrad 730387.   On active part.


Chassis type

Conrad 740632.   On combiner.


2.1mm Jack for +12V power.

Conrad 733946.    On combiner.


10uH RM5

Conrad 447209.   Peaks with circuit caps at abt. 29MHz for extended SWrange and better suppression of VHF.


100uH RM5

Conrad 433396.   Low Ri.


100n RM5



100n RM5


DO NOT use ceramic Cs for C2,4,5. 
Most ceramic C's vary in capacitance with varying AC/DC voltages applied. 
Using them in critical situations can cause IMD !
(Foil capacitors and NP0 type ceramic C's (Murata) are OK).


100uF 25V RM2.5

NB: Ud=ca.11.21V .



All resistors below are metalfilm 0.25W low induction, 
wires bend to RM10
If needed, adjust R1 for ca. 3.25V on T1s






R3 limits VHFrange and de-encourages oscillation (stopper-R).
(soldered directly onto T1g1/2).


150 Ohms

NB: Id must be ca. 20mA.  (Us=ca. 3.0 +- 0.25V @Uc=12.3V)


82 Ohms









100 Ohms

R8 limits VHFrange and de-encourages oscillation (stopper-R).
NB: Ub=ca. 6.44V. @Uc=12.3V.
(soldered directly onto T2b).


2x220+2x180 parallel

Total 50 Ohms 1W, induction free, NOT wirewound.
NB: Ic=ca.115mA @Uc=12.3V
Max output @200kHz: 4.4Vpp.





Teko 1B

ALU box lxbxh = 70x35x43 mm for combibox.
        Best performance:
The original design of this Miniwhip was of pa0rdt, but showed signs of IMD (mixingproducts at 72kHz and between 250 and 450 kHz). I measured ca. 250mVpp output from two very strong high-powered MW AMtransmitters at only 10km distance from me. This 100mVrms input allready caused mixing products (f.i. 72kHz) and distortion of the modulation of the 625kHz transmitter..
Furthermore, in the evening, when strong shortwave signals were received, a 5kHz raster of carriers showed up around the 7MHz amateur shortwave band.
I redesigned the original to improve its IMDfree large signal properties. Several test examples were made, with different antenna surfaces (for testing DX - NVIS properties) and different electronics. 
A Jfet has no protection on its gate. It cannot be driven with positive voltages. Most Jfets have a small input voltage range and low Idss. A dualgate MOSfet with both gates connected therefor is recommended, as it can easilly peak 40mA of Id, and withstand more than - or + 8V input voltages. Two fast diodes at the input further prevent damaging T1 when a nearby TX is active or lightning occurs.
The maximal input voltage of the last version with G1 and G2 in partallel is 8.6Vpp, and max. output is 4.4Vpp ( ! ! ). A so connected fet is very linear (i think). 
All my versions did not show any spouriï. The output was clean.
        Directivity ??
In my opinion a small, vertical, flat, thin antenna+mass surfaces as in the original miniwhip design, shows a dipole-like directivity pattern. It will nearly not receive signals from directly above (80m band) nor from the other thin sides (90degr. off).
Experiments with a vertically, 90degr. crossed antenna surface seemed to support this idea. A combined horizontal en vertical antenna (omni directional) surface over a horizontally placed mass surface did too.
If only DX signals from all directions are of importance, the Miniwhip electronics could be placed inside a metal pipe which is acting as “ground” surface, with an insulated metal pipe above it, as a non-directional antenna surface towards the horizon.
If regional signals coming from above (NVIS, 80/40m) are also important, a cylindrical antenna with metal lid on it (equal top- and side surfaces), above a larger flat horizontal surface as mass, can be used. This will receive NVIS and DX signals. Noises coming from below will be attenuated if the coaxcable mantle underneath it is decoupled from the antenna electronics.
For the latter configuration, a mantel-current-schoke should therefor be placed directly under the coax connection. The coax then will not be of influence on the directional pattern nor can couple manmade noise into the antenna.
        How to choose the best type FET:
The impedance of the base of T2 is very lowZ. In order to be able to drive T2 properly T1 must be able to draw more than 30mA peak current @ 3V Vds into a 150 Ohms source resistor. Originally a J310 Jfet was used for T1 with less current drawn in a higher value source resistor.
Years ago i allready discovered that a dualgate mosfet, with its gates connected together, acting as a MOSfet triode, is an exellent sourcefollower, capable of up to 40mA peak source current, and high input voltage and undistorted output voltage. In the junkbox i found several 40673 dualgate MOSfets, they performed well, as BF961's did too. 
In the given PCB a BF961 dualgateMOSfet (SOT103 style) is soldered upside-down onto the bottom of the PCB. 
A TO92 style Jfet or dualgateMOSfet could be placed at the top with its legs though the hole.
A BF246A or B Jfet should work well. One 2N4859A Jfet could work too. Two BF245C Jfet in parallel could be useable. 
If the a Jfet of your choice has an Idss less than 40mA, reduce Vg1 (R1 higher value) so that Id equals half its Idss (measure it first).
This is not recommended
as the max. input and output voltages will then be proportionally lower. But possebelly still high enough.

Measure the Idss of a Jfet as follows:
- Connect g to s
- Connect a mAmeter between -3V and s.
- Connect d to +3V.
- Measure the current. Is has to be 40 to 50 mA.

The 2N3970, 2N4091, 2N4391, 2N4856A and 2N4859A have higher input capacitances, possebely causing a little less signal. If one of these FETs is used, R3 must have a smaller value, f.i. 470 Ohms (Fc >= 50MHz). R3 is a “stop resistor”, preventing oscillation. It forms a lowpassfilter with the input capacitance of the fet. The -3dBpoint must be above 50MHz. 

        Biasing the FET:
You can use a Jfet or MOSfet of your choice, as long as it can draw 40mA @ Vds=3V & Vgs=0V (or Idss=40mA). Set Is to ca. 20mA (Us=ca. 3.0 V) by choosing the appropriate value for R1. For a Jfet R1 must possebely have a HIGHER value (3M9 - 5M6).

      VLF/LW/MW freq. response:
R1/2 form a high pass filter with the parasitic antenna capacitance (ca. 10pF ?). If R1 and R2 are ca. 10k, frequencies below 3.5MHz will be attenuated (6dB/oct). This could help if a very strong MW or LW broadcast transmitter nearby is overloading your receiver input circuit. 
L1 shoud resonate to 29MHz with the antenna and the electronics input capacitance, in order to keep good sensitivity up to 30MHZ. Above 30MHz the freq. char. rolls off steep. It helps to attenuate very strong VHF/UHF signals which could cause problems.

    PCB and circuit design:
You could build the Miniwhip using the"dead bug" construction (see earlier prototypes). But the (double sided) PCB is designed so, that it also can be used single sided (bottom copper only), and with different styles of fets.
The PCB is best made double sided, with (where ever possible) all mass connections soldered at BOTH PCBsides. This improves stability. Sertain components then must be lifted a bit at their mass connection, in order to be able to be besoldered at both copper surfaces.

The FET impedance converter is optimally biased for the best large signal handling. Again: the fet must have a Idss of 40mA @ Uds=3V. It is a source-follower, with a gain of -6dB. 
With a dualgate MOSfet and separate biased g1 and g2, the signal loss is less but the max output is far less. It is NOT the optimal solution. 
If you want more signal, then enlarge BOTH the antenna surface AND the mass surface of the electronics equally. Penalty: your receiver is easier overloaded. S/N will not be better.
T2 is a common VHF low power transistor: 2N3866. It also is optimally biased for processing the full signal levels delivered by T1. It is directly seating on top of the insulated top copper layer island, and sodered to it, thus cooled by the PCB. A cooling-star is then not needed.
R8 again is a “stop resistor”, it forms a lowpassfilter with the inputca[acitance of T2, lowers gain at rising frequencies, and discourages VHF-oscillation.
The signal splitter  (L3,4 etc) is placed directly near the receiver. It must be in a closed METAL box
It is wise to put ferriteclamps over all the cables there also.

For noise free reception, special care must be paid to the installation of the RG174 (RG58C) signal cable !
The impedance converter T1 can only discriminate the signalDIFFERENCE between the antenna surface and the mass surface of the electronics.
The mass surface of the PCB, and the outside of the screening of the signal cable act both as a “counterpoise” for the “antenna” surface on the PCB.
Any (man-made) noises induced onto the OUTSIDE of the cable screening, will try to run up, onto the mass surfaces of the Miniwhip.

Consequently, these noises will then be seen by T1 as a signal difference to the antenna surface. And WILL BE SEND BACK THROUGH THE INSIDE OF THE SIGNAL CABLE down to the receiver. 

Result: noise on the outside of the cable screening is heard in the receiver.

Prevent noises on the outer surface of the coax can run upwards from down the house upto the Miniwhip.

This can be done by placing mantle-current-chokes in the coaxl cable at the point where the cable goes through the wall outside the house, and at the lower end of the antenna pole :
A coil of six turns dia 10cm in the RG58CU coax itself, 
3 turns of the RG58CU or 5 turns RG174 coax through a 13mm ferrite clamp (Conrad 502137 - RKCF_13_A5) or a toroid with 13-16mm dia hole (use high Mu LW type ferrite).

After that only the part of the coax cable running down from the Miniwhip at the top of the mast, to the lower mastend, act as (noise free) counterpoise, thereby enlarging the signal level and the signal to noise ratio.

        Installation of the Miniwhip:
– Place the Miniwhip as high as possible, and far away from noise sources.
Do NOT use a metal antenna pole electrically connected to any metal structure. Better use a sturdy (composite garden-) pole of at least 2m length. At the top a (short) piece of white PVC or PPC pipe with the miniwhip inside it. (Older grey PVC contains graphite). Run the coax down along the pole.
- Read and use "Miniwhip cable de-noising suggestions". 
– Install two mantle-current-chokes at the lower end of the pole.
– Run the remaining coax far away from mains cables, power supplies, plasma Tvs etc.
– Use a separate, well filtered noise-free 12-15V power supply. Best use a simple (old fashioned) transformer-diodebrige-filter mainsplug type, 12.6-14.2V. Output voltage stabilisation is not needed.  
Small, cheap switched power supplies lack often of filtering, and make a hell of noise. Test them first.

In some experiments i found that, even if the signal insde the coax is clean, HF currents on the outside of the coax cable could be heard in the receiver. Strong noises on the OUTSIDE of cable screening (from switched power supplies or plasma TV inside the house) therefor can still be heard, even if the antennasignal INSIDE the coax is clean.

Test it by connecting a long piece of coax running through the house, terminated with a dummyload. Listen for signals. Then put a coax ferrite clamp to the cable directly near the receiver. See what the difference is. This is the difference obtainable in manmade housing noisefloor.

Place also a 13 or 16mm ferrite clamp with 3 turns RG58U or 5 turns RG174 through it, directly behind the antenna connection of the receiver. It will block noises from the the house into the receiver. 

Also put such a ferriteclamp over the 12V supply cable near the RX.

Some experimental proto types:

A prototype-version with one flat antenna surface (fig.8 directivity ?).
(Antenna surface dimensions taken from pa0rdt).
The output transistor is cooled by hot-gluing it to th PCB.
The PCB hangs on a 40mm sewage pipe cap using a piece of Perspex.
The white cap fits over a 40mm PPC drainpipe.
The input impedance is low, to weaken signals below 2MHz.


A version with a enlarged crossed antenna surface (omni towards horizon ==> DX ?)

A test version with hor. antenna and mass surfaces (looks up, MW/80/40m NVIS).
The lower mast-part is an ALU flagspole. The upper 2m is a 40mm PPC drainpipe.
Between them a ferriteclamp with 3 turns coax in it can be seen.

The NVIS version in heystack construction with horizontal antenna- and mass surfaces.

The cap fits over a 40mm pipe.
Mark the ventilation hole (anti-condens).