SID detection with UKRAA VLF radio

[HN50]

My attempts at SID monitoring have been languishing for a while due to radio problems, though I have been getting to grips with Arduino and even had a go at building my own.

In order to move things along I went to the UKRAA stand at Astrofest and bought the VLF radio kit, their signal generator and the 15V power supply.

Last weekend I soldered the board.  It wasn’t too difficult following their instructions which are pretty clear.  The only fiddly bit was soldering on an SMD capacitor but they also give you a through-hole one as well.  Initial tests suggest it is working but I will have a better idea when it is all connected.

This weekend I have been working on the housing and hopefully tomorrow get the leads soldered.  From left to right will be led power, 15V in, 15V out, antenna in, then analogue -ve and +ve (can be 5V or 2.5V).  
 

I am building it as-is, so the first iteration will have the radio running off mains power and in turn the radio will power the Arduino.  That seems a bit backwards but I would just like to get logging data.  For the time being it will write to an SD card, but the Arduino MKR 1010 looks rather good and comes with WiFi and encryption.

That though, is a long way off..!  Anyway, this shows progress so far.

Edited February 15, 2020 by HN50
Space

[HN50]

I found a length of old extension cable so split it down into the individual wires with a view to using them in this radio.

Power in and out are wired up, as is the power LED.  Antenna BNC socket is connected too.  I had a bit of an accident with the black analogue out socket (broke it), so have ordered another and will sort next weekend.  Red one is soldered and screwed in place.

So after that I have just been using the multimeter to check voltages and connections, the voltage out is about 14.8V which will easily power the Arduino - the home brew one I made has a 5V switching regulator that can take up to 36V.

I had hoped to get it all done this weekend but I am pleased with progress.

Edited February 16, 2020 by HN50
Text

[HN50]

It had been a bit difficult to get much time lately to finish the innards, but a bit of work today and the radio is done.

I tested the output of the 5v output with my multimeter and it is measuring 0.43/0.42V.  I don’t know if that signifies something is wrong, but I also wonder if as I don’t have an antenna plugged in there will be very little signal in anyway.

The 12V regulator sits underneath the pencil line on the lid so I might cut several air holes.  But for the time being the radio is built.

Next I will have to try tuning it.

[HN50]

It has been a bit of a while since my last post but I finally got back to doing some more work on this.

The first task has been has been to tune the radio.  I have opted for Rhauderfehn (23.4kHz) as I understand there is a maintenance window each day when it gets turned off and which helps work out if you are tuned correctly.

I worked my way through the UKRAA instructions and it wasn’t as difficult as I expected, just a bit fiddly.  The instructions say the trimmer resistors have a maximum of 15 turns, but I found mine were 30.  I mailed UKRAA about this though and they got back really quickly.

Anyway the photo below is the outcome of this.  

It does not look much but that 1.23V took a few hours..!  I am using the 2.5V output so there is some headroom for the signal to vary when a flare happens.

So I think the radio is tuned.  

Next was to tune the antenna.  I am going to use the one I built in another thread but have assembled my own tuning unit (inspired by the UKRAA one).

My creation isn’t a thing of beauty but it gives a range between 135pF to 4.2nF.  The dip switches allow you to increase the capacitance and there is a variable capacitor for finer tuning. 

The white capacitors are 1nF, the blues 0.1nF

A few more things to solder but then I can get it connected to the antenna and start up the oscilloscope.  

Looks like tomorrow is taken care of!

[HN50]

I carried on with the antenna today for a few hours.  I found an old project box and screwed that to the antenna base and with a few wires connected up the tuner with bnc male socket.

 

Things were going well up to this point so I switched on the oscilloscope and started measuring.  Unfortunately the oscilloscope is 9V (JYEtech),  it worked for a while but then stopped at which point I realised the transformer was still on 12V.

So I may no longer have an oscilloscope(!).

To get round this I got out my LCR meter and measured the inductance of the antenna with tuner and extension cable added.  I then found an online resonant frequency calculator and put in 23.4kHz and it fired back 2.269nF.  I then used the LCR meter to set the tuner.  

I don’t know how accurate this is but I can tweak it if needs be once the antenna is connected and logging.

So the next task will be to get the Arduino measuring the output voltage from the radio.  

[HN50]

Today I set about getting the Arduino to read the voltage from the radio.  I have built a home brew version from an Instructables page (YABBAS - ‘Yet Another Bare Bones Arduino on Stripboard  ) and got it onto the breadboard.

I have set it to run using the internal voltage reference of 1.1V and added a voltage divider to drop the maximum output from the radio down to 1.1V as well.  My thinking (happy to be proven wrong) is that at 10 bit each increment will then be equivalent to approximately 1 mV rather than 5mV if I had been using the full 5V range.

I got this working and then wired up the SD card and RTC.  

It now goes into hibernation, gets awoken by an interrupt from the clock, resets the alarm to go off in a minute’s time, takes a voltage measurement, writes this and the time to the SD card and then goes back to sleep.  Which looks like this on the serial monitor;

Next step might be to get a veroboard circuit built for a more permanent set up.  Not least as my sausage fingers keep knocking out the jump leads.

[HN50]

I did a bit more work today and built a circuit board that the Arduino plugs into.  This gets rid of all the jump leads in the breadboard and gives slightly more secure attachment of wires.

I tested it for a few hours and it does indeed run in the same manner as the breadboard circuit; I didn’t want to set it running only to find it does not work.  I don’t have a case for it yet as it is a prototype and I may need to make further changes to the circuit.

Next I set about tuning the antenna.  I wrote a sketch that shows the voltage coming out the radio every second and began adjusting the capacitance in the tuner - it gradually began to increase and I spent a while trying to find the maximum.  

The signal is a little noisy (!!! sounds familiar from previous VLF experiments), however from a functional point of  view the Arduino is now logging data coming out the radio.  I am going to leave it in my roof for a few days to see what gets recorded and go from there.

Quite pleased with progress so far, not least as I have something beyond the breadboard stage!

Edited April 13, 2020 by HN50
Picture order

[HN50]

I have had the radio logging since Monday and was interested to see what had been happening, so today I went up into the roof and pulled the data out.  To my surprise/relief it has been logging data all that time.  The overall pattern looks like;

Hmmm, seems a bit noisy but there do appear to be patterns in there and not just a load of random noise.  The plot for 14/04/2020 looks like;

That abrupt drop in signal between 8 and 9 AM is very useful as that is the maintenance window when DHO38 gets switched off.  So that is visual confirmation I am tuned to the correct frequency.  There is a rise/dip/rise around dawn, but later in the day it all goes a bit pattern less. 

15/04/2020 and 16/04/2020 don't show the expected diurnal pattern (though I can see the maintenance interruption on 15th)

Finally 17th to ~13:00.  Again it has a similar dawn profile to 14th and also shows the maintenance window.

I have been in touch with UKRAA and the problem looks likes noise from appliances round the house swamping the signal from DHO38 - LED lights, computers etc.  So I have been turning stuff off at particular times to see if I can work out who the culprits are.  So my timings are;

17/04/2020

16:45 - Turned off everything upstairs at mains (LED lights/pc/phone chargers/toothbrush chargers)
17:00 - Turned off boiler for 15 minutes (not at mains)
17:15 - Unplugged my laptop for 15 minutes
17:30 - Turned on oven at 17:30
18:00 - Turned off mains alarm clock

I will then start turning stuff back on sequentially to see what happens...

Edited April 17, 2020 by HN50
Spacing

[HN50]

Over the course of yesterday evening I spent time turning off and on lights, chargers, computer and so on in order to try and replicate some of the (massive) interference and strange readings I had recorded in previous days. My thinking was that I could at least identify and perhaps mitigate them, so today I went and pulled out the data to analyse.

The full day is below.  Immediately I was a baffled (can't see the affects I had hoped for) but also pleased as it is starting to look like the type of plot I would like to see, plus I can still see the null during the maintenance window.

I went back to my log of what I did when and plotted it onto the data.  As before I was trying to replicate some of the anomalies I had seen on the 15th and 16th, but when I look at the plot I don't really see any obvious impact.  I have not run this via any statistical analysis but certainly by eyeballing it my tests have not had the effect I had wanted.

So... 

I have been turning things on and off again earlier in the day today in the hope that I will be able to see some impact, but I am beginning to wonder if something odd was happening to my setup on the 15th and 16th.  I am also leaving them on for longer. 

What I have pulled out for the 18th already looks similar.

I have remembered that the RTC has an inbuilt temperature sensor so I will tweak the Arduino code.  The instructions mention that temperature can affect tuning so it won't hurt to log it anyway.

Anyway, more things to think about!

 

[Tomatobro]

 I have been looking at the UKRAA SID detector so am following your posts with interest. looking forward to the next update

[HN50]

I checked the data again today and compared the signal to the times I turned things on and off.  As far as I can see they had no affect on the signal which I am pleased about however as can be seen in the chart below something else appears to have happened...

There is a huge jump in signal after I changed the SD card that lasted for several hours.  I don't quite know what has happened there but I left the radio running, unplugged the Arduino, changed the SD card before turning the Arduino back on.  That elevated reading kicked off again later and has carried on into today;

I can still see the 5 AM jump as well as the maintenance window 8-9, but the overall signal seems to be between 0.7 and 0.8 volts.  I am wondering if that jump in signal is due to the Arduino, I have been having a look around and I may need to add in a decoupling capacitor between the Arduino ground and the pin I am using for the voltage divider.  It makes me wonder if the funny readings on the 15th and 16th were also caused by this(?).

I decided to turn everything off at 11:48 for an hour or so and turned it back on at 13:15 but I am tempted to just leave it running for the week.  I am working again from tomorrow and won't have a chance to do any more today.  Next weekend I will build another veroboard chassis that the Arduino/SD reader/RTC/power leads plug in to and add a capacitor, if that works I will then retrofit the capacitor to the other board to keep them synchronized.

It also means with a second chassis I will be able to leave my production system running in the roof until I have tested changes and am ready to deploy them. 

Dear Lord, that sounds like I am back at work already..! 

 

[HN50]

18 minutes ago, Tomatobro said:

 I have been looking at the UKRAA SID detector so am following your posts with interest. looking forward to the next update

Hi,

Thanks for reading, our posts crossed! 

Dave

[Tomatobro]

I have constructed a "SID" antenna and have been using the sound card input as a first look at recording "SIDS" but Speclab just locked up and I lost my first days data. I think there is a BMP image somewhere on the hard drive so will post it here if its recoverable.

As I understand it we are looking for a sudden drop in signal strength with a slow recovery?

 

[HN50]

38 minutes ago, Tomatobro said:

I have constructed a "SID" antenna and have been using the sound card input as a first look at recording "SIDS" but Speclab just locked up and I lost my first days data. I think there is a BMP image somewhere on the hard drive so will post it here if its recoverable.

As I understand it we are looking for a sudden drop in signal strength with a slow recovery?

 

Hi,

That is annoying, have you been logging to a .csv as well? 

I think the change can be an abrupt decrease or increase, but both will be followed by a gradual recovery.

Dave

[Tomatobro]

Ran it again this morning and took a picture. I can see one of the stations turning on and off. The down trace is when I unplug the antenna.

 

[Tomatobro]

SID or Induction hob at 16:50 (UTC)?

[HN50]

On 20/04/2020 at 19:41, Tomatobro said:

SID or Induction hob at 16:50 (UTC)?

 

Hi,

Sorry, with one thing and another this week I didn't get round to replying.  I had a look on the NOAA website the other day and took a screenshot. They didn't flag any events for the same time, so it might be some local interference.

I resisted the temptation to pull the data out the SD card this last week and let the thing run continuously to see if I get any odd behaviour again.  Last night after 6 days I decided to look at the data.  The pattern for the week looks like;

Which looked promising.  I looked at what happened after I had switched off the radio for an hour and restarted it on the 19th...

The high voltage readings had dropped off afterwards, so I think the restart helped.  So I conclude it is probably better to turn everything off before I switch the SD card! 

Individually the days looked like;

What I am pleased about is how samey they look with a remarkably consistent daily plot.  Yes, there is some noise in the plots themselves but that does not seem to swamp the overall daily pattern. 

I had a look at the NOAA data for the same period (https://www.swpc.noaa.gov/) and it looks like there may have been an even on 23/04/2020 however it does not correspond with anything I can see in my plot for that time.  So I think if that is one I didn't detect it.  Having said that the plot puts it in the smallest 'A' class - if it it had been an 'X' class I would be more disappointed!

I also calculated an average daily curve for all my data as well.  The gradual accumulation of data means it is starting to look more like a normal daily plot as bad data is beginning to represent a smaller percentage of the total.

That plot does not show much, I just like it.  For the first few days of data the average daily curve looked nuts. 

I am still interested to see how running it all under battery power helps so the additional chassis I build today will have space for a TIP120 transistor to act as a switch.  I also have a pack of 12 NiMH AA cells that I will solder into a single pack but that will have to be another weekend.  Likewise, I don't forever want to be writing the data to an SD card so in spare moments I have been looking at the Arduino MKR WiFi 1010.  Further down the line I would like to write data to a database (nicely dovetails with my day job!) but for the moment I want to get to grips with the basics; I have found the jump from kitchen table prototype to reliable sensor is a big one.

Today's tasks;

• Add a decoupling capacitor to the voltage divider to see if that helps smooth things a little
• Build second chassis that is 1:1 copy of existing one
• Change code to log temperature measured by RTC as well

Also when I get it running again tonight I will make sure the radio and Arduino are out of the plane of the greatest sensitivity of the radio.  I don't know if that will help but it is worth a try.

Thanks for reading.

 

Edited April 26, 2020 by HN50
NOAA link added

[HN50]

And this is a GIF for the week.

[Tomatobro]

Thanks for posting your latest data.

Its interesting that out of 5 laptops of varying vintage only one has a sound card that seems to work. My setup is based on Jeff Lashleys book the radio sky and how to observe it.  I now have Spectrum lab writing a .txt file and are doing the post analysis in a spread sheet. All the "interesting" activity seems to be at night. I have to keep in mind that the book is now some ten years old so not sure how current the transmitting stations are etc.

If I go the UKRAA receiver route I would write some software for a PIC processor and do the averaging of the data before sending it too some software called MakerPlot for graphing. Take a look at Makerplot as its great software for this kind of thing as it receives RS232 commands and plots accordingly.

[HN50]

3 hours ago, Tomatobro said:

Thanks for posting your latest data.

Its interesting that out of 5 laptops of varying vintage...

Hi,

That MakerPlot looks interesting as you could hook up several Arduinos to feed in to it.  
I would like to go down the cloud route myself as I love the idea of building a sensor (battery or mains powered) that just fires its data off to an internet connection and all the processing and analysis happens online.  The Arduino store page for the MKR 1010 has links to several articles that discuss this plus it has some encryption built into the board from the outset.
 

How do you find SpectrumLab?  People seem to get good results but I found it a bit frustrating and looked at other means of detecting SIDs, which brought me to this method.

Dave

[HN50]

Today’s work;

I have built a 1:1 copy of the chassis so I can undertake further development and still leave my production set up logging in the roof.  They look and work exactly the same, which is good!

On checking through my components box I found I don’t have any 10nF capacitors (might have to check again as I don’t quite believe that) however I did find some discussion about the voltage divider and if your total resistance is 10k ohms or less you don’t need a decoupling capacitor.  So I recalculated and used a 5.6k and two 2.2k ohms in series to drop to 1.1V.

I also updated the code to log the temperature measured by the RTC.  I converted it to Kelvin as,  ummm, I could .

It is all running again in the roof so I will leave it to see what comes back later in the week.

Thanks for reading.

[Tomatobro]

Just out of interest I measured my loop antenna response to input frequency and the peak response is at 25.596 Khz. looks like I would need an ATU if I go the UKRAA receiver route. Blue is frequency steps from 20 to 40 Khz in 1 khz increments and red is ac millivolts change rms

Update. The antenna required 600pf to bring its peak to 23.4 khz. Its now located in the quietest place I could find, Spectrum lab is sorted. I am getting the early morning maintenance cut offs so just need a SID to verify the system.

Edited May 1, 2020 by Tomatobro
update

[HN50]

On 29/04/2020 at 11:21, Tomatobro said:

Just out of interest I measured my loop antenna response to input frequency and the peak response is at 25.596 Khz. looks like I would need an ATU if I go the UKRAA receiver route. Blue is frequency steps from 20 to 40 Khz in 1 khz increments and red is ac millivolts change rms

Update. The antenna required 600pf to bring its peak to 23.4 khz. Its now located in the quietest place I could find, Spectrum lab is sorted. I am getting the early morning maintenance cut offs so just need a SID to verify the system.

Hi,

Thanks for the post.  One thing I quite like about this work is that it generates data for analysis .  Out of interest what are you using to measure your loop antenna response?  Am I right thinking you have an A/C multimeter capable of going down to milivolts?

Having tuned the antenna to 23.4kHz are you finding that the other transmitters are now attenuated or have you got SpectrumLab just logging DHO38?

Dave

[Tomatobro]

I am fortunate that I have access to an expensive fluke multimeter and a frequency generator.

I will post some screen shots later on. I think that Spectrum lab offered a range of "standard" vlf frequencies and 23.4 is the only one I am currently logging ( and noise floor level) as all the others either give no signal or are seriously affected by local interference such as our induction hob and gas boiler.

I am reasonably confident that I am recording DHO38 because rotating the antenna gives me a peak signal when pointing towards the station in Germany and I am seeing the early morning maintenance shutdown sessions. I can see a problem trying to extract small signal changes from the data as the noise span covers some 3db (at least). I am taking 1 reading every 30 seconds as a plot. Perhaps averaging over (say) 5 readings might help but I have not looked at this yet.

Hopefully my signal maintenance dips will match up with your UKRAA receiver data

Regards

 

[HN50]

I was going to do some work on this yesterday however I found as the weather was unexpectedly better than I had thought it would be I spent it in the garden in the veg beds.  By that I mean I was working on them, not lying in them.

Today I worked out the bearing from my house to Rhauderfehn in Germany and it comes out at ~64 degrees, so once I start logging later I will re-orient the antenna.  At present the antenna is not in this orientation, however it does put (I think) a lot of noise from the house into the null zone.  So I am interested to see what the trade off will be - next Sunday I will pull out the data and have a look.

Earlier I cut the power and hopped into the roof to pull out the SD card.  Among other things I wanted to check was that the change to the voltage divider had not caused any problems, I don't know why that would be an issue but as I had made a hardware change I did want to confirm all is okay.

The general trend looks okay, I can see the diurnal pattern, the daily maintenance period is there, the readings taken have have not berserk and the Arduino has been logging the temperature recorded by the RTC as well.

By day it looks like the following;

And in GIF:

My setup is still returning the diurnal profile.  Some funny stuff happened 02/05/2020 between 0800 and 1000 but I wonder if that was due to work at the antenna(?).  Otherwise though the numbers logged look as expected, and I am gaining confidence that my setup is stable.

I had a look at the GOES data - there was an event on 01/05/2020 at about 14:30 however I still think that this lower/mid B class is below what the radio can detect.

For the first time I have temperature from the RTC being logged as well (red line).  I am interested in this as the tuning guide says that temperature increases can push the radio into oscillation (page 16, #12) and my radio is up in the roof where things can get quite warm in the summer (touching 40 Celsius!).  I don't think though at this stage I have enough data to give any meaningful results so I am going to use it for reference over the coming summer.

Anyway, in all I am pleased with what has been logged this week and am thinking about what I will do next.  I would still like to have the option to run from battery with power being supplied by the Arduino and the radio being turned on when needed.  I looked around and found the ready-built MOSFET switch unit using an IRF520 transistor.  Even more useful I found a website that as well as advertising the module for sale also gives a schematic of it;

https://protosupplies.com/product/irf520-n-ch-mosfet-module/

Turns out I have all the bits.  Handy! 

I am increasingly aware that needing to go up the ladder to get an SD card interrupts logging and, while fine for testing, isn't really conducive to real time SID monitoring.  Earlier this week a page arrived with an Arduino MKR wifi 1010 arrived.

I think I will try and work on the switching and initial MKR 1010 investigations in parallel.

Thanks for reading - data is attached.

Dave

VLF_T2_20200503.TXT

Edited May 3, 2020 by HN50
Added data