A Software Controlled Radio Preselector
High Performance Software Controlled Radio Preselector. Tunable Band Pass Filter for HF Bands SCR PRESELECTOR. High Performance Software Controlled Radio Preselector. Dec 10, 2013 - Elektor 's software defined radio (SDR) has been an unprecedented success in terms of publicity and sales, especially in the USA. The original publication in May 2007 was graced not just by favorable feedback from the radio community but also by follow-up projects like a simple tester, a VLF add-on and. High-performance Software Controlled Radio Preselector. 1.8MHz to 50MHz tunable Band-Pass Filter bank. Digital tuning. The SCR Preselector-CAT is suitable to work with SDR or Analogue radios.
-A high-level accessory front end for the HF amateur bands Here's an antenna selector / Preselector / attenuator / preamplifier accessory for every HF amateur transcelver. It can improve your receiver's IP2 for out-of-band signals and yield good flexibility. My father, I4FAF (an old-timer) and I both very much like Amateur Radio as a lifetime endeavor.
We do not have backgrounds in electronic engineering, but we do have a lot of practice. My father is a fast builder ofAmateur Radio projects, from printed-circuit artwork drawn by him with CAD software to working units in our home laboratory. Being retired now, he has more time and I help him from time to time. Our goal is to get more from the Amateur Radio equipment available to us. Our gear is average, not toppriced. I want to improve my skills with DX or weak signals, while I operate in crowded bands during some international contests. Lately, I have discovered the low-frequency bands.
They have added more fun. In 2000, we started to put up (at a flat, country location) a short vertical antenna by Butternut, the HF2V It has the 160-meter coil kit, is top-loaded with four wires (each about 5 meters long) and has six ground radials about 40 meters long for 160, 80 and 40 meters. In winter 2001/2002, we started to test some receive-only antennas, with better signal-to-noise ratios and some directivity, in comparison to the 360° radiation pattern of a vertical antenna. With the exception of Beverage and a four short vertical system presented by W8JI on his Web site, that still require much space for only -6 to -11dBi; most of the receive-only antennas we have considered are in the low-output category-in the range from -6 to -35 dBi.
We have worked with Beverages, EWE, the delta-EWE by K6SE -a variation of pennant-flag antennas and K9AY loops. We have seen them presented by our trusted teachers in recent artieles in Amateur Radio publications. For example, ON4UN's Low Band DXING (third edition, don't miss reading the new chapter 'Special Receiving Antennas'), QST, the Antenna Compendium series and K1ZM's DXing On the Edge-all published by ARRL. There has also been some follow-up on the Internet and on the top-band reflector from W8JI, WA2WVL (EWE antenna), K6SE (delta-EWE and other pennants), WA1ION (pennant with remote variable control of the resistive, in-line termination), K9AY (K9AY loops, now also with remote variable control), K3KY (his Web site has a full collection of contributions, links, about low band antennas), W7IT-TV (rotatable flag) and other well-known authors., There is a lot of interest and newcomers frequently ask, 'What is the best receiving antenna for the low bands?'
I like the Beverage very much, hut my answer must be that I don't know, simply because, until now, I have not been able to test them all. Read K1ZM's book. He agrees that it's better for hams to have more types of antennas available on 160 meters. That is true because of the variable and peculiar propagation conditions on that band., Our beverage and other receive-only antennas We tested a 177-meter-long, unidirectional Beverage configuration (for USA), up about 2 meters above ground (rural terrain), with a 500 Ω end load (two 1000 Ω resistors parallel connected to a ground rod) and an input impedance ratio of 1:9. The transformer was made of seven quadrifilar turns in parallel using #20 AWG or 0.8-mm-diameter enameled copper wire. The core is an Amidon ferrite FT114-F with a permeability of 3000.
Remember that this material, manganese-zinc, has a low bulk resistance, so it is best to cover the core with a thin layer of Teflon tape before winding the wire on it. See John, ON4UN's, third-edition book for a transformer picture and photo on page 7-17, Fig 7-18: 'Modified transmission-line transformer'. With our ground characteristics, we had better matching results with this ferrite mix than with the type 43 (permeability of 850) proposed in the book. The thing was tested by us with help of a new MFJ-269 and confirmed by our friend's laboratory-grade spectrum analyzer and tracking generator. ON4UN usually uses high permeability type MN-CX, which was not available to us.
A very nice description of how to get a Beverage system to work properly is contained in K1ZM's book, on pages 12-1 to 12-6. It had been up for only a few days of tests in the winter of 2001/2002 when a 160-meter CW-contest weekend came along. My impressions of performance were very favorable.
I had a great time with this antenna. Around 10 US states were worked in one night and with a very clear copy over my short vertical as the transmitting antenna. I still remember those nice signals; of course, most of them were from well-equipped contest stations. A simple Beverage antenna alone provides about -11 dBi. I don't know exactly the gain of my short vertical but the relative difference in level in decibels does agree quite closely. A phased system of two might have a higher output, at about -6 dBi, as per ON4UN.
The other receive antennas' outputs can be, at worst, about -30 to -35 dBi for a delta-EWE loop; see more detailed data in the new chapter in ON4UN's book. We have tried the K6SE delta-EWE, a pennant receive-only antenna that has reasonable dimensions: total wire length is 72 feet, 28 feet on the base side and the high apex is about 17 feet from the base side. It is easy to make it rotatable. In that case, the transformer is differently placed in one lower corner; in the opposite corner is a 950 Ω series resistor and it matches 50 to 950 Ω. K6SE suggested a FT140-43 ferrite with primary and secondary wound at the opposite sides of the core. The primary is 8 turns and the secondary 34 turns with about 990 µH to 1 mH using 20 AWG enameled wire. Remember that the directivity is in the opposite direction from the termination corner, toward the feed point, unlike the Beverage.
Fig 1 - (left, A) A block diagram of the system. (B) PTT and front-panel controls.
The bandpass filters are selected by relays and a front-panel band switch. Preamplifier gain is controlled by a front-panel switch. (C) An additional transmit antenna may be controlled with an added relay and switch. The Need for an Antenna Processor We immediately realized the importance of making frequent checks on the receiving antenna and on the transmit/receive antenna to get more flexibility from. That is, to avoid overloading the inside equipment switches when the same functions (antenna 1, antenna 2 and receive-antenna selections) are already built into some recent radios. If needed, you can make maintenance of such switches easy if they are in an outside home-built unit. Now, we don't have the problem of switching among more than two antennas!
Preselector Filter
In practice, we felt immediately that we needed a complete independent accessory for our transceiver as an outboard tool to deal with the issue of better selectivity in the receive chain. So we stopped our antenna tests and started to think about the design of a complete HF front-end unit with bandpass filtering for our amateur bands only, not for general coverage. Just to simplify and avoid wasting time durmg contest activity with peaking controls, we decided as a practical tradeoff to choose fixed band-pass filters without variable controls. That's why we agree with G3RZP when he wrote in QEX May/June 2002, on page 40: 'Are our receivers too sensitive?' The answer is 'Probably, but. Construction contracts hinze pdf. There are some imponderables. On the LF bands especially, the use of separate receiving antennas producing much lower-level signals hut also lower levels of noise means that requirements may exist at times for the low noisefigure levels that are typically seen in modern receivers'.
Obviously, the use of pre-mixer selectivity has a major effect on the performance requirements, although at 7 MHz, the proximity of the broadcast band offers little possibility of really effective filtering in conventional circuits. US conditions seem a lot quieter than those in UK'.
Thus a variable antenna attenuator has obvious advantages, but the attenuation steps need to be much smaller than the 6 or even 20 dB steps provided by commercial transceivers. I must confirm that it is very hard for us in Europe. We work split on 40 meters to listen to DX and North America among very powerful broadcast AM stations in your portion of the band from 7.150 to 7.300 MHz. The test Peter performed was with a FT-102, which is not a general-coverage receiver but has some ham-band preselectors in it. 1 don't want to use too much attenuation first if the rig used is even poorer.
First, I would like to try a bandpass filter in front of it with a moderate insertion loss, narrower than the internal one. Then, eventually, I will add more attenuation if needed. A resistive or PIN-diode attenuator is by nature broadband. Insertion loss in a band-pass filter is already an attenuation of RF signals. Outside the filter passband, attenuation inereases on both sides.
A practical preselector is desirable in the front end of a receiver to protect all the following stages of the receive chain. In-band insertion loss shouldn't be too high, but 4-6 dB is acceptable since in many cases, you don't need the full sensitivity of your modern receiver. Only when band conditions permit can you switch in one preamplifier to compensate insertion loss. I do believe that a good receiver must be designed for low IMD in all stages and should haven arrow filter from the beginning of the chain so all the following stages are protected. If not, you need a better following chain.
Some system gain-distribution consideration could be done, with one preamplifier switehed in, attenuator off, as shown in Table 1. Filters oss (dB) Atten (dB) Preamplifier gain (dB) Stage gain -5 0 +12 Total gain -5 +7 Stage gain 5 0 3. Table 1- Preliminary Gain-Distribution.not measured. To calculate cumulative NF and cumulative input intercept, please look at Chapter 4, 'Receiver Design', in W E. Schoenike, Single Sideband Systems & Circuits, 2nd Ed., MeGraw Hill, now also 'HF Radio System & Circuits'.
ARRL laboratory test have reported about the good performance of the Elecraft K2 receiver with respect to 5 kHz spacing, two-tone IMD test in comparison with some higher priced commercial equipment. Thereby arises a question: Why? A first answer could be that it has a narrower first IF filter and maybe a better first mixer as well. Most up-conversion, general-coverage receivers for amateurs have allmode capability and one roofing filter around 70 MHz, and wide enough for FM. A switchable first IF filter to narrow the bandwidth while on CW or SSB is desirable, but that adds to the cost.
In addition, you might need to change the whole architecture since such VHF first-IFs are not compatible with narrow band-pass filters. Maybe a secondary effect must be considered: problems in the area of signal delay to synchronize a conventional noise gate for an effective noise blanker. I do remember, some years ago, someone complaining about less effective blanking action with a Drake R4C receiver after the replacement of the first-IF crystal filter with a narrower one. A well-designed front end with band-pass filters around our band segments is an added bonus to improve our equipment's IP2. A preamplifier is not always needed. Fig 2 - A schematic of the high- and low-pass filters.
The 100-kΩ resistor near the RX ANT connector provides a path to ground for electrostatic discharges.You may want to add a surge supressor in parallel with it. Our Front End Now, our accessory needs to be an external independent front-end unit with its own filtered power supply. It should be easily connected to any transceiver (new or old) with:. A variable attenuator from 1 - 20 dB with a bypass switch;. Modular ham-band-only band-pass filters with relay switching (no diodes to avoid IMD), the inputs of unused filters are shorted to ground;.
Two stages of preamplification;. A push-pull, broadband medium quiescent-current amplifier configuration with low IMD and a reasonably low noise figure with some kind of RF feedback. The variable attenuator and preamplifiers are protected by the band-pass filters, since they are placed after them in the receiver chain.
This is a reinterpretation of a highlevel receiver front end, as we see it, adapted for our use. Lt is a system made with well-known circuits as building blocks. You can modify what you want, since every unit is modular. Improvements are welcome.
The unit must be capable of some switching among different antennas: ANT 1 RX/TX, ANT 2 RX/TX, RX-ONLY ANT. For each antenna selected, the receiveonly signal path is always routed through the band-pass filters. See K5AM's article (QEX, Nov/Dec 2001, p 40) in which he pointed out different IP2 performance when measured at the main receiver terminal or at ANT RX ONLY input, leaving some hope to the home builder for better performance. The front end should be useful in casual DX operating, in single-operator contests and in multi-operator contests with the receiver signal parallel routed to two receivers. These would be a main receiver and a secondary receiver with a second operator who can tune independently. It should be capable of work in low-frequency amateur bands but with modular construction that can be upgraded to cover all HF bands including WARC bands.
When used with full-sized antennas, it should provide benefits as well, since the band-pass filters are designed and aligned with sharp bandwidth and excellent shape factor. We use more space than most embedded band-pass filters.
Equipment manufacturers must tradeoff cost, dimensions and the Q of components. I'm thinking now of an Amateur Radio system composed of one antenna with multiband coverage with only one feed line to the rig that covers 10, 15, 20, 40 meters and the WARC bands, like log-periodics. Friends with such antenna systems told me about more IMD problems in their receivers during evening hours because of the high-level signals present in the broadcast bands around 40 meters. The situation is a bit better for those who use monoband antennas and separate antennas with separate feed lines for 40 and 80 meters.
Again, this problem seems to be worse in Europe, as pointed out by G3RZP. Think about radios with receiver general-coverage capability. If the number of band-pass filters is 10, they must be around 3 MHz (30 MHz/10) wide at -3 dB, and of course much more at -60 dB.
These so called 'suboctave-width. Band-pass filters' are a limited form of preselector filtering but they are still helpful. We have tried to select band-pass filters with bandwidth of around 400 kHz, except on 28 MHz because the band we can use is 1.7 MHz wide. The filters are centered at the middle of each amateur band with an acceptable insertion loss. In that way, the improvement in bandwidth we achieve is about 6 to 1 (3 MHz/0.4 MHz) and we believe that everything before the first IF roofing filter is a bit better protected from strong, out-of-band signals. Fig 3 - Schematics of the band-pass filters.
(A) shows a 1.8-MHz band-pass filter from an old Handbook. L1, L2 and L3 are T68-2 powdered-iron toroid cores with 40 turns of 0.5-mm (#24 AWG) enameled wire. Capacitors are dipped mica parts. Variabie capacitors are Philips film/Teflon units that are 10-mm in diameter.
(B) shows 3.5 to 28 MHz Butterworth filters with inlout capacitive dividers. See Tabie 2 for each band's design values.
Connect the relay 12-V control lines together, and control thern with circuit in Fig 1 B. Decouple the control lines for RF at both ends. (C) shows 3.5 to 28 MHz Cauer filters.
See Table 3 for each band's design values. Band (MHz) Toroidal cores L1-2-3 Turns Wire diameter (mm) µH C1-C8 (pF) C2-C7 (pF) C (pF) C4 (pF) C5 (pF) CR1-2-3 Var.
(pF) CR var C (pF) 1.83 For 160 m data, see ARRL Handbook 1980 p. 8-43 and refer to our dedicated board.