Active Antennas 4/28/2017

Transcription

1 1

2 Active Antenna??? What Why Theory Loops Verticals Dipoles Amplifier Considerations VE7KW experiments 2

3 What Simply an antenna with an amplifier - Automotive application - My first exposure to active antennas was in Newfoundland where a company was making an amplifier to overcome the poor performance of the printed window car broadcast antennas - Small low frequency antennas - I had built a passive 1m dia transmitting loop for 80m - Converted to tuned receiving loop for 470 khz by inserting a ribbon cable connected as multiple turns and added amplifier/upconverter 3

4 What (cont d) - Broadband antennas - I became fascinated by the mini-whip used at the University of Twente (this is now over 10 years old!) - This led me into a year long project to improve upon this antenna for use as a diversity antenna for the K3 and as a 100kHz to 30 MHz antenna for my Cloud-IQ SDR. Active Antennas, Receivers and Pre-amps Active antennas overcome some problems 4

5 Why Match to receiver Most small antennas are highly reactive and definitely not 50 ohms Overcome low antenna gains (-20 db) Antenna sensitivity is basically a function of size so a small loop or vertical has very low gain Better IMD performance Broadband receivers (and amateur receivers in a multistation or contest environment) have to withstand extremely large signals with minimal effect on the weak signals desired. 5

6 Why (cont d) Broadband performance (100kHZ to >1 GHz) Modern SDRs have coverage from 9kHz to >2 GHz in a single device so what do you use for an antenna? Reduce transmission line losses and noise Mast head amplifiers at VHF and up Common mode reduction LF-HF 6

7 Theory An antenna is a transducer (EMF to voltage) The larger the antenna the more effective (gain) For a small (untuned) antenna, the output to the receiver will go up with frequency The output of a transducer is optimised when matched 7

8 8

9 Theory Cont d What we want is signal to noise (s/n) No point in amplifying noise. Most amateurs turn the RF to max and I have even seen pre-amps being used on 160m with a full size vertical! The rule of thumb is that you should just hear a small increase in noise when you connect the antenna. For 80m with a full-size antenna this normally means attenuator on and some RF gain reduction. A low output antenna can have lower inherent noise Directional (loop) Remote (verticals and loops) 9

10 Theory cont d The transmission line shouldn t be an antenna Balanced antenna, Differential amplifier Common mode on the outside of coax Twisted pair (cat 5 especially if shielded, CAT6) can be quieter The amplifier has to withstand very large out-of-band signals e.g. Broadcast stations (AM or FM) without causing Intermodulation distortion (IMD), the amplitude modulation of signals containing two or more different frequencies, caused by nonlinearities in a system. The amplifier looks like and draws power like a QRP TX amplifier! 10

11 Theory cont d We can have low output broadband antennas and then optimise the matching, IMD and common mode performance to achieve improved s/n Consider the antenna as a probe either Electric field (small verticals or dipoles) or Magnetic Field (small loops, including ferrite loops) Output is determined by the size of the probe relative to frequency. At low frequencies can be small with poor sensitivity as the frequency increases the sensitivity improves 11

12 Theory Cont d The amplifier bandwidth is determined by: Amplifier gain/bandwidth (50 MHz easy, 1 GHz??) Input Capacitance (reduces the signal for electric field probes) which reduces high frequency sensitivity. Feedback can help increase bandwidth 12

13 Loops (untuned) Sensitivity (signal out from magnetic field) Increases with loop size Increases with frequency Increases with number of loop turns (but capacitance will limit bandwidth) Is directional Inherently balanced so differential amplifiers/baluns required. 13

14 Verticals Sensitivity (signal out from Electric Field) Increases with size Increases with frequency Capacitance (depending upon diameter and length) forms a voltage divider with the amplifier input capacitance. Larger element capacitance and smaller amplifier input capacitance is better. Omni-directional Unbalanced (Where is ground ) Pickup of noise on transmission line 14

15 Dipoles Sensitivity (signal out from Electric Field) Increases with size Increases with frequency Capacitance (depending upon diameter and length) forms a voltage divider with the amplifier input capacitance. Larger element capacitance and smaller amplifier input capacitance is better. Omni-directional?? Definitely if vertical Probably if horizontal and small Balanced (differential amplifier required) 15

16 Amplifier Considerations Broad bandwidth (high gain/bandwidth required) Low input Capacitance Differential desirable for loops/dipoles High IMD rejection Normally requires high power Favourite devices (E.G. 2N5109) were from CATV amplifiers and are now not manufactured Feedback circuits can help with bandwidth Are exposed as normally at the antenna Ability to drive CAT5/6 helps reduce feedline noise 16

17 VE7KW experiments 1m dia loop K9AY/SAL30 PA0RDT mini-whip Mini-dipole Current Mode Amplifiers Cross-Country Wireless mini-antenna 17

18 VE7KW Experiments Initially the interest was in exploring the new 470 khz band and other LF activities Looked at using the loop for noise reduction Then the interest became in having an antenna suitable for use as a diversity antenna with the K3 (160/80m) Needed reasonable sensitivity Omni-direction was desirable I became interested in noise monitoring with an SDR Wide bandwidth ( Mhz) Remote receiver 18

19 470kHz loop This was a 1m dia (octagonal) loop made from ¾ dia copper pipe and 45 degree elbows originally intended as a magnetic loop transmitting antenna. Ribbon cable was inserted and multiple turns obtained by staggering the connection of the ribbon cable. It had too many turns to be tunable and broad band was desirable. It was used with an amplifier/up-converter to 4 MHz Provided adequate coverage of the LF band 19

20 Noise Reduction Because loops are directional, the idea is that a noise source can be nulled to improve s/n if the desired signal is in a different direction This was tried with the octagonal loop and with a ferrite loop. In my Burnaby location, it was ineffective due to the noise being in all directions. I later did a campaign to clean up the noise from my own house (wall warts, cheap routers/hubs, long CAT runs) which was far more effective. A sense antenna can be phased to steer the loop but this null will be broader and less deep. 20

21 Diversity Because this was to be used for MF reception on Dxpeditions a relatively high sensitivity was required along with some directionality. The K9AY and SAL arrays have been used. The directionality comes from selecting and phasing the loops. As these are quite large, the signal can be sufficient for high IMD capable receivers such as the K3 but an amplifier is provided in the K9AY system mainly to ensure high IMD capability. Although broadcast band filters are used, a Dxpedition is likely to have simultaneous 160m and 80m 1KW operations. 21

22 Diversity cont d I built an amplifier for the first K9AY and Adam VA7OJ assisted me with testing it. I did not have suitable high power transistors and was concerned about reliability so obtained a DxEngineering amplifier which Adam also tested. In Tonga (A35T) we used an Array Solutions shared apex loop array, with the supplied amplifier which I believe is similar to the DXEngineering amplifier (2N5109 push-pull) 22

23 Diversity cont d I wanted to have diversity capability at VE7KW and thought that an active antenna could provide this. The mini-whip fascinated me because of its simplicity and small size <0.5m. I built this antenna and felt that it did not have the required sensitivity so decided that I could improve upon the design. 23

24 Improved? Active Vertical Dipole I noted that the University of Twente appear to have gone to a vertical dipole configuration and moved the Antenna to the roof and clear of local noise sources. Increase the size/capacitance by using longer and larger diameter elements to increase signal I thought that going to a current mode integrated circuit amplifier could provide the low capacitance, differential, broad bandwidth, high common mode amplifier desired Use CAT5e or CAT6 shielded cable for the transmission line Isolate power supply 24

25 use the K3 dynamic range of 106 db say 100 db. minimum input voltage 0.1uV. Maximum signal = 1V. A dipole element 3cm by 10cm each side will be 30 pf. The input capacitance is 3 pf so we only lose 0.9 db due to the capacitor signal division. -100dBm is -24 dbuv/m = V/m. The antenna is 0.66m long so input voltage is 0.04 uv. Thus a preamp gain of 10 db is required. The equivalent input noise for the LM1385 is 3.2nV/sqrtHz. So for 500 Hz = 0.7 uv 25

26 Antenna Design Copper sheet is rolled into 2in Dia PVC tube configured as an ~ 1m dipole. The amplifier board is inside the copper sheet which acts as a shield The Cat5 is connected through a shielded waterproof connector via a short jumper. Low cost Cat5 CAT cable is used. CAT6 is now available. Hung in the trees at about 5m high 26

27 27

28 28

29 This is interesting, a low horizontal dipole is effectively isotropic and does not show low elevation lobes until > 1/4λabove ground. Figure 180m Horizontal dipole at 1m Figure 2 80m Horizontal Dipole at 40m A low vertical dipole starts off with low lobes and become more isotropic as it is raised. 29

30 Considering that I was likely to end up at 6m, these models caused me to change my mind about needing a vertical dipole. Figure 3 80m vertical dipole at 1m Figure 4z80m Vertical dipole at 40m A low vertical dipole starts off with low lobes and become more isotropic as it is raised. 30

31 Amplifier I went through a number of amplifier designs using various IC amplifiers. What I ended up with was Fet front end to act as an impedance transformer to give low capacitance, high impedance to the antenna and differential current to the Current Feedback video amplifier. Current feedback amplifer for high gain bandwidth Line driver amplifier for the CAT5 +/- 5V isolated switching supply CAT5 isolation transformers 31

32 40dB gain referenced to one input and flat to beyond 100 MHz. 32

33 Cat 5 transmission line Well known to be effective for noise reduction simplifies transmission of amplifier DC power Can be transformer isolated Nominal 110 ohm (can be configured for 50 ohm) Can have additional shielding Standard connectors; shielding and sealed more of a problem Low cost (even Cat6 now) 33

34 Cat 5 Transmission line cont d This became a major project I had a supply of CAT transmission line transformers but had a steep learning curve on their use Compensation for wide band Use with DC power Driving and receiving Decided ultimately that the transformers were probably not needed 34

35 35

36 Cat 5 Transmission line cont d When I started CAT6 (individually shielded with outer shield) was very rare and expensive. Now it is being used all over for new installations. I had some waterproof connectors and could find shielded connectors. Again this is much easier now. Bandwidth and matching was not a problem 36

37 37

38 Cat 5 Transmission line cont d 38

39 Amplifier Power Supply I was concerned about noise pickup. In addition, it appeared as if a +/- supply would be required. I thought that I had an easy solution where I would send up 12VDC and use a local isolated power supply. This became a major issue with separate shields and extensive choke/capacitor isolation required. 39

40 40

41 This shows my improved EFA1 antenna to the Miniwhip MW1 from 1 to 30 MHz 41

42 At this point I decided that I needed a Benchmark and ordered a Cross-Country Wireless Active antenna. This is a 1m dipole with a conventional 2N5109 amplifier and CAT5 transmission line. I found that this had similar performance in the trees as my active antennas. It improved when placed on the mast holding my VHF beam to where it equaled or exceeded the R9. Note the R9 is a multiband resonant vertical. It is down in comparison to my 160/80/60m dipole at 50 ft but works down to 60kHz. 42

43 43

44 44

45 db Active Antennas 45 s/n X-Country at 30 Ft s/n against WWV X Country R MHz 45

46 Conclusions Active antennas work BUT: They have to be away from any noise sources (house wiring), buildings or trees They have to be decoupled from their transmission lines They are not noise cancelling. They can make a poor receiver better and in high noise locations can offer a s/n improvement. Active Antennas are an attractive choice for broadband SDR receiver applications The SDR will have to be able to handle any strong signals in the antenna bandwidth 46

47 Conclusions cont d The amplifiers are non-trivial The traditional 2N5109 designs work well High gain bandwidth/low noise figure/high 3IP devices CAT5/6 can work and is a good choice for transmission lines The improved EFA has potential but, at this time, is on hold as the improvement would be small. 47

48 Here, as stated by W8JI, is the bottom line stated for larger loops but applicable in general At -140 dbm and 250 Hz noise bandwidth, the system would require a 1 db noise figure front end. The only place negative gain antennas that require more than ~ 20 db gain with a normal receiver at a quiet location will work into the external ambient noise floor generated outside the antenna is in a location blanketed with strong local noise. Besides that, if the gain is so far negative the coaxial cable will easily become more of an antenna than the thing we call an antenna So, if you already have a decent receiver, antenna and location, there is no improvement to be had. 48

49 The EFA projects gave me an opportunity to drag my self into the 21 st century world of modelling, (EZNEC, LTSpice), layout software (CircuitMaker) and using modern surface mount components; ICs (CFA, Isolated switching modules), transformers. Next day delivery is a true wonder! I would like to thank Nick Massey VA7NRM for assistance with the modelling and circuit layout. Dave Miller VE7HR facilitated my prototypes and encouraged me through the dark times with CircuitMaker. 49