Barry Davies Memorial Receiver
Barry Davies was a long standing member of the Medium Wave Circle who sadly passed away in 2021. He left a generous legacy to the Circle which has enabled us to purchase a remotely located receiver at one of the best DX locations in the UK which members can access and use via the Internet. The Circle has installed a KiwiSDR at MWC Treasurer’s Martin Hall’s location at Clashmore in North West Scotland and it is fed from one of his low noise Beverage antennas that pull in weak signals from around the world.
Don’t equate using the KiwiSDR with listening to recordings made of the whole MW band. The Kiwi experience is more like using a regular analogue receiver as we did on DXpeditions in days of yore, instead sitting in your own home. You will be at the forefront of the propagation experience, observing changing conditions as they happen in real time – with the ability to simultaneously record the audio. Because you may hear signals on almost all 10kHz channels it is easy to quickly build up a mental map of the prevailing propagation conditions. We hope you enjoy using it.
Users searching to check out UK LPAMs or local stations should be aware that Clashmore is well away from the more densely populated parts of the UK, with the nearest MW transmitter being more than 200km away. The facility is more suited to MW DXing.
An obituary for Barry appeared in MWN Vol 67 #05 October 2021, and maps showing Martin’s location and antenna details have been described previously in the article on beverages which appeared in MWN Vol 67 #03 July/Aug 2021 and aren’t repeated here. If you don’t have copies of these issues, they may be downloaded from the MWN Archive.
KIWI SYSTEM RF Signal Routing
The diagram in Figure 1 shows the current receiving system and highlights the path of the RF from the antenna towards the KiwiSDR. Note that the Devantech board used to control the antenna switching is under the control of the Jaguar software installed on PC1. More on the scheduling of antenna switching is provided later in this article.
The RF signals from the splitter pass to the KiwiSDR via a 9dB attenuator, for reasons described in the next sub-section. The Kiwi also receives a GPS signal for frequency accuracy. Data from the KiwiSDR is passed via CAT cable to a secondary modem which provides a Virtual Private Network (VPN) over which the data travels. The data then flows through a domestic 4GEE modem to a remote server, through which up to 4 users may access the information from the Kiwi and control its settings using their browsers. Here is some additional information that might be of interest:
- The VPN is necessary to provide a “tunnel” through the vagaries of the 4GEE, which uses a Mobile Data link to the nearest phone mast near the Stoerhead Lighthouse, and from there through the network to the Remote Server, which provides a fixed IP address through which to access the Kiwi.
- The 4GEE router also provides access to the internet via wi-fi for all the PCs, laptops, mobile phones, printers etc in use at the house for home office, amateur radio and MW DXing purposes, including Zoom meetings.
- We typically experience 50-70 Mb/s download speed and 20-30 Mb/s upload speed, although it’s lower at times due to other local users. It is anticipated that this is sufficient to meet the demands we wish to place on it – the upload bandwidth required for the KiwiSDR is only 320kb/s with all 4 channels active. The speed compares favourably with the internet access provided by the landline, which gave only 500 kb/s download speed at the time I terminated the contract. We are fortunate in this area that our 4GEE system is continually being upgraded to support the Emergency Services Network (ESN).
RF Sensitivity and Dynamic Range
Receiver front ends need to be sufficiently low noise to receive weak signals – the lower the noise generated in the receiver front end, the greater the sensitivity. Bjarne Mjelde’s “Sensitivity Figures on Some Receivers” (Ref. A) shows comparative sensitivities at 1000 kHz for various receivers: KiwiSDR -102dBm, Perseus SDR -103dBm, JRC NRD545 -114dBm, AOR7030+ -106 dBm (Method: AM mode, 6 kHz bandwidth, 30% modulation signal at 400 Hz, for a 10dB S/N ratio).
The figures show that the Kiwi and Perseus are pretty much on a par, and although their sensitivity isn’t brilliant compared with the NRD545, it’s adequate for most locations. This is because local QRN/QRM received on the antenna is higher than the receiver generated noise, and therefore determines the S/N you hear. In a very quiet location like Clashmore there are occasions during the day, in mid-winter, when this is not the case, and a greater receiver sensitivity is required to stop the front end noise masking the wanted signal.
It’s worth noting at this point that a better S/N doesn’t necessarily result in better intelligibility – there is no doubt that the audio quality of the Jaguar/Perseus combination is superior to that of the KiwiSDR, giving it the edge in intelligibility.
System sensitivity can be improved by inserting a low noise preamplifier in front of the receiver, which is a satisfactory solution when the noise received on the antenna is very low. However, the associated gain can increase signal levels to the point at which the receiver becomes overloaded, particularly during the hours of darkness when signal strengths increase dramatically. The overload mechanism in Software Defined Receivers (SDRs) is different from that in analogue receivers, so there is little point in comparing specifications. Both the KiwiSDR and Perseus use 14-bit Analogue to Digital Converters (ADCs) in the front end, which determine the dynamic range of the receiver. So how can we avoid receiver overload at night and still have maximum sensitivity during the day?
Perseus provides a solution by giving the option to switch in 10, 20, or 30dB of internal attenuation before the front end, to avoid overload at night. The Jaguar control software can be set to do this at pre-determined times, and from experience I set it to switch 10dB of attenuation in from 30 minutes before sunset to 30 minutes after sunrise. However, the KiwiSDR has no such feature, and so an external fixed attenuator needs to be inserted in the RF chain before the front-end to prevent overload, as shown in Figure 2. But what does this do to receiver sensitivity during the day?
And why is the Preamplifier necessary, and what’s its effect and that of the Splitters in the RF chain shown in Figure 1 on receiver performance?
A test was carried out during the daytime recently when the noise level was low, which showed that when listening on an empty channel if the antenna was replaced by a 50 ohm dummy load the noise level in the received dropped by approximately 2dB (from -117dBm to -119dBm in a 3kHz bandwidth). This confirms that the noise received on the antenna was higher than the KiwiSDR front-end noise, so there is no loss of receiver sensitivity. However, on even quieter days this might not be the case, and further tests will be conducted in due course.
Let’s get technical for a moment and carry out a theoretical analysis. The DX Engineering RPA-1 is a high dynamic range, high linearity, low noise preamplifier with a NF of 3.5dB and a gain of 17dB. It’s followed in the RF path by two splitters, each introducing a loss of ~3.5dB, leaving a gain of +10dB, after which a further 9dB of attenuation, determined experimentally, is introduced to avoid receiver overload. Work by John Fallows VE6EY (Ref. B) shows Perseus to have a Noise Figure (NF) of +17dB, so the KiwiSDR will be very similar since their sensitivities are similar. (Note that the lower the NF the better).
What effect does introducing all these components have on the sensitivity of the receiving system, and what happens if we remove the 9dB attenuator? Plugging the numbers into the online Pasternack Noise Figure calculator (Ref. C) gives the results shown below. Note that they are approximate and for illustrative purposes only. [It is conventional to reserve the term NF for active elements whereas it has been applied to the passive attenuation in this simple analysis. The internal mechanisms are different – in addition to the attenuation of the signal, thermal noise is also generated in the losses themselves, which means that (SNR(out)/SNR(in) is always worse than the calculated loss, i.e. the resulting NF is slightly optimistic in the above analysis].
- 9dB attenuator in circuit to avoid overload (as shown in Figures 1 and 2): the overall NF remains at ~17dB. Essentially, the amplifier gain compensates for the attenuation that follows, and the NF of the KiwiSDR determines performance.
- The 9dB attenuator shown in Figure 2 is removed: the system NF improves to 8.6dB, though at the expense of reduced dynamic range. The lower NF of the RPA-1 is providing some benefit.
- For comparison, when the numbers for the RF path to the Primary Perseus SDR are plugged into the calculation, the corresponding results are 14dB and 6.5dB respectively.
This tells us that the KiwiSDR system design is sound for all conditions except when they are exceptionally quiet.
Locally Generated Noise
When Martin first moved to Clashmore the beverages were terminated close to the house. Although he took steps to reduce the effect of noise generated by our TV, SMPSs, PCs, LED lamps etc, as well as noise imported on the mains, it became a losing battle as time went on, and QRN increased in level. After a few years he started to terminate the beverages further away from the house and successfully eliminated the pickup of noise generated within the house. All his beverages now terminate between 50m and 200m from the house, and the coaxial cables back to the house are buried to a depth of a few inches (~10cm). Care has been taken within the house with screening, earthing, and cable routing to avoid pickup on the cables before they reach the RPA-1 Preamplifier, so any pickup of locally generated signals is at a very low level.
The nearest house is a holiday cottage 100m away, which is only occasionally occupied, and has no modern electrical/electronic devices within. The nearest permanently occupied house is about 500m away, but only the African beverage points towards it, and this is infrequently erected. Any local noise picked up on the Mid-west and West Coast North American antennas comes from the nearby township of Raffin, about 1km off the ends of the antennas, and on the Asian antenna from occupied dwellings 200m from the end, although this isn’t usually significant.
Various antennas are erected at different times of year to take advantage of the seasonal changes in DX propagation. A rough guide to what you might expect is shown in the table below:
|Description||Bearing (True)||Length (m)||Load||Approximate Availability|
|Asia||46||420||680W||August to March|
|South America||236||460||560W||April to September|
|East Coast North America||276||490||560W||Sometimes all year|
|Mid-west North America||305||920||680W||August to April|
|West Coast North America, AH, HI, Japan||342||630||560W||August to March|
These are the most likely antennas to be erected, although other combinations are possible.
When more than one antenna is available the scheduler in the Jaguar software on PC1 is used to switch between them at appropriate times of day to maximise DX opportunities. Currently available antennas and the switching schedule between them is shown at the top of the KiwiSDR page in the form shown in Figure 5.
Whenever there’s a change to the antenna or to the switching schedule, Martin updates the details in this pane.
Operational Aspects and Propagation
It’s simple, just enter http://xx.xxx.xxx.xxx:xxxx/ into your browser (the exact URL is only available for MWC members) and you’ll be asked to enter a password, which will be updated from time to time and will be publicized via the Online Group and in Medium Wave News.
Since this service is provided as a membership benefit you are asked not to share the password with others. Please enter your name and location (and optionally your amateur call-sign and QTH Locator) in the box in the top right of the screen – if you can’t see it, zoom out to 80% on your browser settings while you fill it in, and then zoom back to 100% again when done.
Since the system is optimised for Medium Wave, you will notice that the waterfall is masked above 3MHz. Using the KiwiSDR itself is pretty intuitive, but if you run into trouble, please refer to the Introduction to Using the KiwiSDR (Ref. E).
Our target is for the KiwiSDR to be available as a service 24/7 throughout the year. However, please be aware that interruptions or disruptions might occur due to one or more of the following:
- Kiwi will occasionally need to shut down and re-start. This can happen automatically to install updates, but we will sometimes re-start it manually just to avoid any build-up of problems. This will only be done during the day when there are no users connected, though if something appears to be wrong it will be done as soon as the problem is noticed. Normally these re-starts shouldn’t last for more than a couple of minutes.
- Power outages – which can be due to the weather or maintenance of overhead power lines. They may last for a few minutes up to as long as several days.
- Short breaks in reception during routine maintenance when Martin disconnects the antennas to undertake a continuity check. He tries to avoid doing this near the top of the hour!
- Damaged antennas due to the weather, livestock, or wildlife – in which case Martin will keep members informed via the Online Group.
- Emergency Service Network (ESN) traffic will always take priority over all other network traffic, although we think this is unlikely unless there is a major incident locally. (See Ref. F)
- Although this is an electrically quiet location, local noise does appear from time to time. Reception can also be disrupted by noise due to passing hail/snow/sleet showers, during which all signals might be wiped out; fortunately, they don’t usually last long.
From time to time, you may experience interference from Martin’s ham radio activities, particularly when he’s transmitting FT8 on 1840kHz. He tries to avoid transmitting over the top of the hour. (FT8 will sound like 15 second bursts of noise alternating with 15 second periods of normal reception).
When propagation is poor you might hear very little in the way of DX.
If you experience any of these issues, “please try again later”.
Propagation conditions in the far north of Scotland are very different from those in the south of Scotland, let alone the south of England! The long dark winter nights and proximity to the terminator to the north throughout daylight hours brings in DX 24 hours of the day. In summer the long days and short period of semi-darkness at night restrict the period in which DX can be heard for several months over the Solstice.
Sea gain of up to 6dB and the ability to receive signals at low angles of elevation can at times bring in long-haul DX at good strength, though perhaps only for minutes at a time. Of course the variability of geomagnetic conditions is the determining factor in what can be heard in the way of DX, not just in relation to the strength of DX stations but also the strength of European QRM.
Here’s a brief summary of what you might expect to receive throughout the year:
The winter DX season starts in August or September with the reception of west coast stations, peaks just before Christmas, and peters out in late February. Stations from the Indian sub-continent are best in early autumn (fall). As the season progresses, once the east coast stations have faded out in mid-morning stations from the mid-west, Prairies and west coast are at their best in the late morning, followed by Alaskans and Hawaiians peaking around noon. Under exceptional conditions, stations in New Zealand have been heard around this time, when the co-channel European stations are subdued or in a fade. If conditions are good, propagation to Japan and the Far East then starts to improve, peaking just before sunset when the European QRM begins to increase. East coast stations can be heard throughout this period at good strength, and hunting for daytimers on an evening can be very productive, especially in the hour or two before their local sunset – and east coast stations continue to come in well right into the spring.
The summer South American DX season starts to build as the North Americans decline and continues into the autumn. The best times for listening are in the late evening in Clashmore, just before the station’s local sunset, and in the period around dawn in Clashmore. There is a good sea path between Clashmore and Argentina/Uruguay, so stations from these countries are well heard on the South American beverage. Stations in the Caribbean region are usually best heard on the North American east coast antenna.
However, don’t take our word for it, get on the Kiwi and find out for yourself. If you arm yourself with the DX Atlas software (Ref. G), centred on Clashmore you’ll be able to see the darkness areas, terminator (grey line), and plot paths from Clashmore to the DX. Best reception usually occurs when the path is at right angles to the terminator, or along the grey line. Good hunting!
Logs made using this facility will be welcome in your contributions to DX Loggings, subject to a few simple guidelines about what you should submit. Most of the stations reported so far to the Online Group are commonly heard in Clashmore, and we want to avoid being overwhelmed by multiple reports of, for example 590 VOCM and the other Newfoundland stations, or the powerhouse 50kW stations east of the Mississippi which are regularly heard here, so some restrictions are necessary (see below).
It’s also worth noting that it’s quite normal to hear 590 VOCM early in the evening in winter, and also 1400 CBG, although hearing them late afternoon is unusual. North Americans are often heard 30+ minutes or more before their local sunset – hearing them 2 hours before would be noteworthy. Under good propagation conditions, don’t be surprised to hear many stations coming in very strongly with armchair copy. Look out for the stations lurking under the dominant powerhouses, and for daytimers, if you want something new. Remember that you’re more likely to hear the less common stations when the powerhouses aren’t coming in too well due to disturbed conditions, during which propagation can be very focussed on relatively small localised geographical areas.
You can check what is commonly reported and what is less common by reference to my own reports to DX Loggings, or to the UK All Time Lists (Ref. D).
Take your time to get used to what can be heard, and there’s a good chance you will hear something that I haven’t reported recently, or that is even a UK First.
To summarise the key points above and already made to the Online Group:
- Logs will count as UK logs and be eligible for entry in the UK All Time Lists* since the receiver and antennas are located in the UK.
- In submitting logs, please highlight those made using the Clashmore KiwiSDR by using the suffix /c after your initials, e.g. SW/c. In the table of contributors this will appear as, for example, JW/c John Williams, Hemel Hempstead, Herts (Clashmore KiwiSDR).
- Please restrict your contribution to DX Loggings using the Clashmore KiwiSDR to no more than 10 of your best loggings, unless you are reporting less common or unusual station operation (e.g. overnight operation of a station listed as a daytimer), or stations that normally only give network IDs giving local IDs (e.g. 1380 ZYI773 Radio Continental) in which case there is no limit. Use your discretion.
*To avoid multiple reports via the KiwiSDR distorting the statistics for received stations a station logged via the KiwiSDR by multiple users will only count once.
We’re still feeling our way, so these guidelines will undoubtedly evolve in the light of experience.
If you would like any enhancements to the system and the way we have chosen to implement it, please raise them on the Online Group for discussion, and bear in mind the practical implications of any suggestions. Martin has a few thoughts:
- Fine-tune the value of attenuator in series with the RF input to the KiwiSDR to optimise the trade-off between overload and sensitivity.
- Alternatively investigate the design of a system which would switch the attenuator out from 30 minutes after sunrise to 30 minutes before sunset without manual intervention.
- Consider dedicating one of the less-used beverages full time to the KiwiSDR.
- Introduce filtering to reduce interference from my ham radio operations.
Please note that Martin is unable to consider requests for specific antennas to be erected or for the switching schedule to be modified to meet your personal preferences.
Explore the facility and have fun – and hopefully you will enjoy some good DX!
- Sensitivity Figures on Some Receivers – Bjarne Mjelde. http://arcticdx.blogspot.com/
- Making It Up (Enjoying Radio and Maker Hobbies) September 18, 2018 – John Fallows VE6EY. play.fallows.ca/wp/radio/software-defined-radio/perseus-sdr-preamp-mystery-is-solved
- Cascaded Noise Figure Calculator – Pasternack. https://www.pasternack.com/t-calculator-noise-figure.aspx
- UK All Time Lists, maintained by Andrew Brade on behalf of the Medium Wave Circle. https://groups.io/g/mwcircle/files/2_MW%20DX%20heard%20in%20the%20Brritish%20Isles
- Introduction to Using the KiwiSDR. http://kiwisdr.com/ks/using_Kiwi.html
- Emergency Services Network: Overview – Home Office https://www.gov.uk/government/publications/the-emergency-services-mobile-communications-programme/emergency-services-network
- DX Atlas, available from http://www.dxatlas.com/DxAtlas/ for a small fee.
Why not join the Circle now and get access to one of the very best medium wave SDRs in Europe?