Feature Articles
- Details
- Written by: Ken Gibson
- Category: Feature Articles
- Hits: 621
Completing the last page of my 32nd RSGB log book recently made me think of the information recorded within its pages and those of its predecessors.
I know that Sean EI7CV and I have completed 2033 weekend contacts over the past 45 years but what about the lunchtime net? What do my records show about that?
32 RSGB logbooks each containing 80 pages, with 26 lines per page. That's 66,560 entries, the first of mine being a top band AM contact with Eric Letts G3RXJ on the 13.10.1967.
My lunchtime net started on 1.9.1979, a 15m SSB contact with HO Townsend (then WA5MLT, later K5CX) when he was living in Norman Oklahoma. My eldest son David had just arrived at Oklahoma University to take his Masters Degree and HO promised to look after the 20yr old during his stay.
HO and I then started a daily sked, which developed into a net including MSARS members and others over the years. When David moved to Pensacola Florida, the net began to include local amateurs in the Serious Hams ARC (SHARC) including Mike N4MAD, and of course Bob, N4XAT who later moved to New Jersey and remains a very active member of the net today.
HO you may remember came over to Louis G5RV’s funeral as an official representative of the ARRL and met many MSARS members on that occasion. I’m still in touch with Anita N5AOK HO’s widow who is now living in Texas.
From 1.9.1979 to date is 43 years plus 2 months and another 10 days for leap years and at 5 contacts a week means that the lunchtime net has, so far, met on 11230 occasions with a few breaks for holidays etc. Think about that when you next log on to 21,345! There have been one or two forays onto 20m at the depth of some particularly bad cycles but in general the net has been on 15m for most of its life and I suppose you could say that we are very well known there.
My log books also contain records of contacts with many other long-term friends. Who remembers Dell Popplewell K4NBN (No Bad News) for example and how many of you have worked a King (JY1)?
Nowadays log keeping is digital and many amateurs don’t even keep them, but perhaps they are missing something in not being able to pick up an old logbook and remember contacts made many years ago.
Ken Gibson G3WYN
- Details
- Written by: Alex M0TOT
- Category: Feature Articles
- Hits: 1139
Revisiting a Budget QRP ATU Project
I first built this project in March 2016. Since then ‘hundreds’ must have been sold! However, I recently came across a reference to this particular ATU on a YouTube video, put together by Carol Milazzo KP4MD in July 2021. This video showed that the inductance had been connected incorrectly on the original schematic. The number of windings at certain points on the toroid were wrong. There was also a mistake on the PCB silkscreen (see references below).
So, I used my original PCB, but re-wound the inductance in accordance with the recommended numbers in the YouTube video.
I checked the antenna’s SWR values with an MFJ Analyser (without the ATU) and then tested the ATU and antenna with a miniVNA (The original antenna was a Windom, the present one, as tested, is a vertical). With the switch in the 'Tune' position, the LED on top comes on if the unit does not 'see' 50 Ohms (this is done by a Wheatstone Bridge).
Test Results
| Band (m) | Mid-Frequency (MHz) | SWR(1) | SWR(2) |
| 160 | 1.92 | >3.00 | >3.00 |
| 80 | 3.70 | >3.00 | 1.38/10 |
| 60 | 5.32 | >3.00 | 1.18/8 |
| 40 | 7.10 | >3.00 | 1.94/5 |
| 30 | 10.12 | 1.20 | 1.35/3 |
| 20 | 14.18 | 1.65 | 1.27/2 |
| 17 | 18.10 | 2.40 | 1.27/1 |
| 15 | 21.23 | 3.00 | 2.22/1 |
| 12 | 24.94 | 3.00 | 1.39/1 |
| 10 | 28.95 | 3.00 | 2.39/1 |
| 6 | 51.00 | 1.60 | 1.25/1 |
Notes:
SWR(1) = Results using MFJ Analyser, Model No 249, on antenna only.
SWR(2) = Results using miniVNA and QRP A.T.U. on antenna.
The secondary figure in SWR(2) column is the band switch position.
The antenna on test alongside this ATU was the Thunderpole SE-HF 360 Fibreglass Vertical Antenna.
I have a few spare PCBs if anybody is interested in taking on this small project.
Kind regards
Alex Henderson M0TOT
References:
- Details
- Written by: Maintenance Guy
- Category: Feature Articles
- Hits: 4049
Some of you know that I've been experimenting (or maybe that's just 'dabbling') with various SDRs (ostensibly the Kiwi and the RSPdx, but also the HackRF One and Nooelec NESDR SMArTee RTL-SDR) running on a succession of semi-home-made magnetic loop setups. To clarify, what I mean by that is a re-working of the 'loop' element of a couple of off-the-shelf mag-loop pre-amps and Bias-T units placed in various locations around my QTH.
I thought I'd post a brief description, along with a few photos of the antenna I ended up with.
I've been tinkering for several months now, to see what sort of results I can get. I was hoping that the club would invest in a Wellbrook ALA1530 loop, which is the de facto standard if you like, so that I would have a yardstick with which to appraise my results. Happily, one now adorns the rear of Cyprus Hall.
I'm going to build my own Web-connected SDR (again, from off-the-shelf bits) running the same OpenWebRx system that the Kiwi uses, and then pop my homebrew effort on a big stick in Cyprus Road car park so that I can perform a direct comparison.
In all honesty, I don't really need another antenna, but (spoiler alert...) these loops do provide a pretty good HF reception solution in a compact and practical package. I can tell that already from comparisons with my main antenna.
So I started off by buying one of the cheap Chinese MLA30+ loops, via airmail from the source.
It was tiny. In fact, it came complete with Bias-T, USB lead, wire loop and maybe 15m of very thin coax which fed straight into the weatherproof pre-amp box - no connector, and it all came in an A5 padded envelope. I ran it like that for a while, but when I saw the modifications that other people were making to them, I decided to follow suit and snip off the coax (and SMA connector at the Bias-T end) replacing it with a chassis-mount BNC female connector and soldering the remaining short tail of coax to the appropriate points now on the inside of the enclosure. I re-sealed the entry-point with Araldite.
Pushing the 9mm drill through the side of the remarkably well-made enclosure without damaging the circuit board or the large coil you can see above was quite tricky and took quite a lot of concentration. The electronics are very well sealed in whatever goop that is, and that includes the coil and its windings, so I have no concerns around the longevity of the internals.
The big plastic mounting tabs that protrude from the top and bottom of the box had some useful mounting holes. Less usefully, they were only around 4mm diameter - not enough to get the common-or-garden 4.8mm wide cable tie through. They were soon opened out to 5mm.
Without touching on the actual performance of the unit in receiving RF signals, the only other slightly negative thing I have to say is that the attachment bolts for the loop itself (be it the supplied 70cm diameter stainless steel wire loop, or any other loop construction you wish to attach) are too small at 4mm diameter. In fairness, I suppose they're OK for the wire loop, but it wouldn't have hurt (or cost the manufacturer any more) to have fitted 5mm or preferably 6mm button-head Allen bolts. I didn't bother changing them out, as there isn't a lot of room going spare in there, but they could easily have been accommodated pre-assembly and prior to the tags being soldered.
Turning then to the loop itself, I played with a number of options. I started with some very promising composite tubing given to me by Phil G4UDU. It's a sandwich of PE-X/Aluminium/PE-X and on the face of it, it looked perfect. Neat, stiff, lightweight and pre-formed into roughly a 1m coil.
I cut a suitable length and started work on fabricating something to make the connections to the pre-amp attachment points. I started by trying to strip the outer sheathing of PE off the ends of the tube. Nightmare. It was all bonded together like you wouldn't believe. Next, I made some cranked brackets out of 2mm thick aluminium sheet with a 4mm hole at one end and a 14mm hole at the other. I then drilled out the bore of the pipe with a 14mm drill to remove the inner layer of PE and screwed a 14mm stainless steel nut and bolt up into the ends of the loop, securing the aforementioned brackets between the nut and the head of the bolt, leaving a good inch of thread in the tube (ooh err, look at me mixing my imperial and metric units...). That was obviously going to provide a solid connection, except that it didn't. It worked OK at first, but then stopped working. Intermittently infuriating, before being thrown up the garden in a childish rage.
Next I tried some beefy coax with an obscene amount of copper in it, also supplied by Phil, but it too wasn't really lending itself to the connection approach I wanted to use, plus it wanted to form any other shape than a perfect circle. When I started to see various animal-shapes in it, it followed the PE/ALU/PE loop into the herbaceous borders. I will retrieve it at some point, as I'm going to use it for yet another mag-loop in the loft, and I don't care what that looks like.
I'd watched a YouTube video of some guy that had formed a 1m loop out of two lengths of 15mm copper plumbing pipe. Very sensible, except that he had joined them with one of those plumbing connectors pre-filled with solder, and then used other connectors and joining methods to attach wires to go to the pre-amp lugs. Basically, he had taken a really nice, simple concept, and then proceeded to turn it into a right dog's dinner. Spray paint, tape and dribbly solder everywhere. OMG.
He obviously hadn't found the 3m lengths of copper pipe in Wickes DIY store in Burgess Hill. Other DIY emporiums are available.
So the diameter of a circle is 3m of copper pipe divided by π, which equals a 1m diameter mag-loop. Thank you Archimedes.
Time to get baking. Yes, baking. Wet sand is no good for filling copper pipe, so I heaped one of the GLW's baking trays with soggy sand and popped it in the oven at 170°C for 20 minutes. I told the wife that I was making a tray-bake. I don't know what that is, but I can tell you that when the beeper on the oven went, and she went to help herself to a tasty treat, I had to do some pretty fast talking.
I plugged one end of the pipe, filled it with sand, then plugged the other. The sand stops the pipe from deforming and collapsing when you bend it of course. Once I'd wrangled the 1m diameter round kitchen table out into the garden and duct-taped the pipe to the edge, I was easily able to bend the copper around it. It went remarkably smoothly, and no, you can't borrow my table.
After emptying the sand, I bent a couple of tabs on the ends (annealing the copper mid-bend to stop it from fracturing) and drilled a couple of 6mm holes in them. A bit of filing and so on, and we were done. Simplicity is nearly always the best policy.
Why 6mm holes? Ahh, revelations follow.
Not a bad job, if I say so myself.
I made the top mount out of a white 20mm electrical conduit tee with a 13mm wide slot cut in the top to snap the 15mm pipe into. This was in turn held in place by a white BMX handlebar grip, cut down to be a really snug fit in the end of a length of the ubiquitous 40mm waste pipe.
A decent new male BNC compression plug and some Mini-8 coax feeds the signal away and up into the shack. The loop is in a compromised location currently, and testing is therefore also compromised, but I can tell you that it's remarkably good. Proper comparative testing will follow, as hinted at earlier.
Job done then.
Well, not really. Of course I now have the same problem that I had when the KiwiSDR was located here. Namely, a nearby HF transmit antenna spitting out 100W (or 6.1W ERP if you're the ICNIRP police). I started looking at several solutions to this on the various forums, and didn't have much confidence in any of them, until I came across an apparent consensus that the pre-amplifier produced by Cross Country Wireless does what it says in the sales literature i.e. it "provides protection which allows it to be used very close to transmit antennas without damaging the amplifier or the attached receiver".
Now, these units get mixed reviews in the various loop shoot-outs that I've read, but I pulled the trigger at £55.20 plus shipping and I haven't been disappointed thus far. Again, some hard-nosed testing between the Wellbrook, the MLA30+ and the CCW amplifier will follow, with the last two using the exact same copper loop, but that's for another day. No mounting tabs/lugs on this unit you'll notice, but it does have 6mm loop mounting bolts and wingnuts!
If you fancy an unobtrusive Rx antenna for your shack or even your backpack, then a mag-loop may be the answer. You could potentially avoid having to spend £264 on a Wellbrook loop too, if you're not super-fussy.
If you can't wait for my testing results, there are a number of quite objective studies already out there on the Internet - the best of them on YouTube.
For the price (around £35 for the MegaLoop MLA30+) I don't know why you wouldn't have one at least on standby. OK, so it's receive-only, but what a great pair of ears, even when used with a random length of wire draped over a nearby thingamajig. They even work indoors.
M0XYF
- Details
- Written by: Alex M0TOT
- Category: Feature Articles
- Hits: 916
Alex M0TOT has followed up on his uni-polar stepper motor project with some modifications and enhancements - it now uses a more powerful Bi-polar motor and a new driver. It's also now controlled by an Arduino Uno open-source microcontroller board. You can see the code ('Sketch' in Arduino terminology) below. It's quite easy to follow along and see how it works. Well done Alex.
This is what he had to report on his latest project. Thanks to Alex for his contributions as always.
I am not really into Arduino and Sketches but could not use the uni-polar motor P.C.B. from last time to run a bi-polar motor. It would have meant building two new 'H' Bridge PCBs.
The Arduino ‘Example Sketch’ gives all the ‘C’ Code lines for the motor:
// Example sketch to control a Stepper Motor with TB6600 Stepper Motor Driver.
// Reference: https://www.makerguides.com/tb6600-stepper-motor-driver-arduino-tutorial/
// and ‘Programming Arduino – Getting started with Sketches’ by Simon Monk (Second Edition).
//
// Compiler defines Stepper Motor connections and steps-per-revolution:
// Each step of 1.8 degrees is divided 8 x times = 0.225 degrees.
// 360 degrees is divided by 0.225 degrees = 1600 microsteps.
// Microsteps help to reduce the amount of vibration at slow speeds.
// In microsteps a phase is not fully 'ON' or fully 'OFF'.
// This may cause the motor to become warm/hot at low speeds.
int dirPin = 2;
int stepPin = 3;
int stepsPerRevolution = 1600;
void setup()
{
// Declare pins as OUTPUT:
pinMode(stepPin, OUTPUT);
pinMode(dirPin, OUTPUT);
}
void loop()
{
// Set the spinning direction CLOCKWISE:
// Spinning direction of the motor is set in
// the'dirPin', either HIGH or LOW.
digitalWrite(dirPin, HIGH);
// Spin the stepper motor one-revolution SLOWLY;
for (int i = 0; i <stepsPerRevolution; i++)
{
// These four-lines result in one-step:
// The speed of the motor is determined by the
// frequency of the pulses that are sent to the
// STEP pin. The higher the frequency the faster
// The motor runs. Pulses are changed in the
// 'delayMicroseconds()'.
digitalWrite(stepPin, HIGH);
delayMicroseconds(2000);
digitalWrite(stepPin, LOW);
delayMicroseconds(2000);
}
delay(1000);
// Set the spinning direction COUNTER-CLOCKWISE;
digitalWrite(dirPin, LOW);
// Spin the stepper motor one-revolution QUICKLY:
for (int i = 0; i <stepsPerRevolution; i++)
{
// These four-steps result in one-step:
digitalWrite(stepPin, HIGH);
delayMicroseconds(1000);
digitalWrite(stepPin, LOW);
delayMicroseconds(1000);
}
delay(1000);
// Set the spinning direction CLOCKWISE:
digitalWrite(dirPin, HIGH);
//Spin the stepper motor five-revolutions or 8,000 microsteps FAST:
for (int i = 0; i < 5 * stepsPerRevolution; i++)
{
// These four-steps result in one-step:
digitalWrite(stepPin, HIGH);
delayMicroseconds(500);
digitalWrite(stepPin, LOW);
delayMicroseconds(500);
}
delay(1000);
// Set the spinning direction COUNTER-CLOCKWISE:
digitalWrite(dirPin, LOW);
//Spin the stepper motor five-revolutions FAST:
for (int i = 0; i < 5 * stepsPerRevolution; i++)
{
// These four-steps result in one-step:
digitalWrite(stepPin, HIGH);
delayMicroseconds(500);
digitalWrite(stepPin, LOW);
delayMicroseconds(500);
}
delay(1000);
}
Bi-polar stepper motor specification:
NEMA 17 Bi-polar Stepper Motor, 4-Wires, 12 Volt, 0.4 Amp, 28 Ncm.
When the motor is running slowly it becomes warm/hot, so I had to attach some heatsinks. In fact the two together are still too small, but I did not have anything else at the time.
Alex M0TOT
- Details
- Written by: Maintenance Guy
- Category: Feature Articles
- Hits: 1936
Our very own WebSDR
I think most members are aware that we now have a WebSDR available to us (a link to which is available at the bottom of the website homepage). What you may not realise is that it's temporarily installed at my home while we test out suitable antennas, and then look to get the whole shooting match installed either at the club shack, or if that proves problematic, at a secondary location which is both RF-quiet and has a good internet uplink.
I'll be posting three or four articles on how the whole project comes together to provide a useful addition to our shack, and I'll write them in the style of a retrospective guide for those that come after us and who are seeking guidance. The implementation has certainly proved to be popular with members so far, with all four available access slots filled up every weekday lunchtime. On several occasions, I have been unable to snag one for myself, and if that trend continues we do have the option of allowing eight concurrent users, though only two of those eight will get the full graphical 'waterfall' experience. The rest just get an audio interface.
For this first instalment, I'll detail the physical build so that you can get a feel for what it looks like in the raw. I'll also cover the initial setup process and explain some of the difficulties I encountered along the way. The next article will give an overview of the back-end software configuration so that you can see some of the flexibility we have in that regard.
Other articles may cover wideband antenna testing, and possibly one on how to actually use it once you've loaded the webpage into your browser. That last one may just end up being a talk at the club or over Zoom, and would be a very high-level walkthrough to enable everyone to have some fun with it, regardless of experience. Most of you will already be well beyond that, I suspect.
Hardware - What's in the Box?
I purchased the KiwiSDR in kit form from Mr Lynch. I'd have been happy to have sourced it from anywhere, but they happened to be by far the cheapest. I also picked up a 5V power supply while I was at it (1.5A, although owners in the know often recommend a modified 3A Raspberry Pi item). You can also get a really nice aluminium enclosure, but it's around £45, and is not in stock anywhere that I could find. Thankfully the kit comes with an adequate DIY acrylic case which turned out to be fine as long as you don't plan to throw it around too much.
The box contained nearly everything you need to get started, and contrary to the instructions, the two main parts (the little BeagleBone computer board, and the KiwiSDR radio board) were already mated together via a subset of the GPIO pins. I pulled them apart again for the photos below.
There is lots of information on the BeagleBone computers at beagleboard.org. They've made several variants of this developer-focused board over the years, but apparently all the KiwiSDR kits come with the later BeagleBone Green (BBG) which is good news as these all include 4GB eMMC flash memory. The Green board (Green, as in the colour of the screen-printed background) is actually a Seeed Studio collaboration - an evolution of the original BeagleBone Black, and made possible by the open-source hardware philosophy much loved by the maker community.
The 4GB of memory is important, because ultimately it'll give us an upgrade path to Debian 10.x (Buster) from Debian 8.x (Jessie) which is now unsupported and therefore effectively a burning platform.
In an attempt to avoid letting the magic smoke out, I read all the info on the maker's website which proved very useful albeit a little bit daunting for a complete newbie. I needn't have worried, as most of the scary situations that beginners seem to find themselves in are as a result of just diving in and not reading up on the basics first.
Turns out that as long as the two boards have been connected together correctly, you can pretty much just plug in the various cables and then watch it all power up.
The only pre-requisites for a successful first boot appear to be a reasonable internet connection and an existing DHCP server set up on your network. The Beagle will need the latter to establish an IP address for itself, as you can't easily set this up manually until after it has performed a software update. You can change it to a static IP address once it's up and running, which is what I did. I like to know where my flock are at all times.
I built and tested the device before attempting the enclosure assembly. This seemed sensible in case there were any issues encountered, and because although I read the instructions before starting, I'm only human, and the case is boring!
I'm going to suggest that a logical pre-power-on procedure would therefore go something like this:
1) DON'T INSERT THE SD CARD. There is an Micro-SD card included in the kit, and a corresponding slot on the BeagleBone. It's for emergency recovery only, and is not needed unless you really know what you're doing, or the support guru on the official forum suggests it's a good idea. I'm going to guess that the Micro-SD card contains the self-same v1.2 image that's already pre-loaded onto the device.
2) Connect some sort of reasonable HF antenna (via adaptors if necessary) to the RF ANT section of the KiwiSDR board. There is a standard screw-terminal block for a High-Z antenna wire and ground, but I did what most people will do and connected a 50Ω coax-fed antenna to the commonly-found SMA socket. No antenna is supplied in the box.
Make sure you pick the correct SMA socket though, because there are TWO on the Kiwi board. The other one is for a 3.3V GPS antenna. The silk-screen printing on the board and the manual clearly identify which is which.
3) Connect the above mentioned GPS antenna, if required. The kit comes with a neat little GPS antenna, so although the whole thing is going to live under my desk for the time being, I connected it up even though it has no possibility of seeing clear sky in this location. I wanted it to be detected during any automatic software update and to have its driver correctly installed. Its presence gets flagged for inclusion on the receiver list at kiwisdr.com as well as being used in actual TDOA directing-finding functionality. More on that another day.
4) Plug in an Ethernet cable. Thankfully there is no WiFi capability on-board. Good old-fashioned copper twisted-pair technology for solid reliability and bandwidth. Woe betide anyone who thinks it's a good idea to use a cheap WiFi bridge as a permanent solution. My experience of these is mixed to say the least. The other end of your cable is obviously going into your router, network switch or whatever. This isn't a networking tutorial, though it might turn into a bit of a networking anecdote later on. An Ethernet cable is not supplied either, though I'm sure you have a whole bunch of them in those otherwise empty boxes up in the loft.
Now before we reach 5) Connect power supply, get yourself ready for bed, or find something else to do which will keep you completely occupied for at least two hours. You are NOT going to want to interrupt the initial start-up process until it completes normally, and it's almost impossible to resist the temptation to power-cycle your new toy every five minutes if you're just sitting there staring at it.
5) Connect power supply. Yes, you can do it now, as long as you definitely read and understood 1) DON'T INSERT THE SD CARD
Most, if not all BeagleBones which come in the Kiwi kits are pre-loaded with a Debian-based operating system which is ready to go out of the box. I have a feeling that my retailer of choice had performed this step on my behalf. The initial startup process (assuming you have an active internet connection) will soon sniff out the latest version of the Beagle/Kiwi software on Github and go get it. It will then compile everything on your device and then reboot itself. As mentioned, the pre-loaded version is v1.2 but the latest version (at time of writing) is v1.544, and there is no comparison. Security issues alone mean you must upgrade, but functionality is massively improved too.
If everything has worked as it should have, your KiwiSDR is now available on your local network. You can probably find it at:
my.kiwisdr.com
or
kiwisdr.local:8073
...if you're lucky.
I just dipped into my router's admin pages and found it's DHCP IP address, and then added the port number :8073.
Your admin page can then be found at yourlanip:8073/admin
Relevant password information can be found at kiwisdr.com.
I'm not going to cover the dynamic DNS setup, because either you know how to do this already, or you're going to do some research online and work it out. The Kiwi group used to offer a reverse-proxy solution, but no longer. Fortunately though, there is a DUC (Dynamic Update Client) included in the (downloaded) software and accessible via the Kiwi admin panel. Unfortunately, it only allows configuration of services from no-ip.com (so you can select a yourname.ddns.net:8073 address). Annoyingly, no-ip nag you to death every 30 days on the free version, so I used my own script instead of the regular DUC so I can peacefully enjoy the free services offered by duckdns.org.
You're also going to need to do some port forwarding on your router. You probably know how that works. Just set up a virtual server, where you map an external port number to your newly-assigned static IP address (you don't have to go static, but it's much easier IMO) and internal port number, then tell it to use the TCP protocol.
You can find generic information on how to perform these last two steps all over the place, including the manual for your router hopefully.
Once you've got everything working, you can go back to the acrylic enclosure build. Instructions are included, and are generally OK, except that the position of the USB port printed on the acrylic base panel doesn't correspond to where it is on the BeagleBone (in my version at least) and the photos are lousy and lack detail. Top tip: the bottom acrylic sheet (and yes, you will need a hairdryer and some serious patience to get the protective film off) goes on the UNDERSIDE of the metal side panels and not INSIDE, or you simply won't have enough clearance to get the paired boards underneath the slide-in top sheet. You'll also have to separate the Beagle and Kiwi to bolt it all up and then get the enclosure push-pins in place which hold it all together (remarkably well). Be sure to push the GPIO pins fully home between the boards too, otherwise clearance will once again be an issue when assembling.
I've made all that sound like a bit of a nightmare, but in reality it's really straightforward and fun if you take your time - especially if you've done this kind of thing before.
Next Time...
What happens when all of this goes to hell in a handcart, how to fix it, and how to finally get yourself some much needed alone-time with your new KiwiSDR.
Berni M0XYF
