KE7IZL
Member
First, here's a bit of explanation about the main piece of hardware in my setup. The Picoscope 2204A is the cheapest computer-based oscilloscope I could find, but it's still got some really good specs. Most of there scopes have a certain sample rate that can be used in the official software, but can go higher when using the SDK to write your own program. However the 2204A (the cheapest scope) can't use any higher sample rate with the SDK than when using the official software, at least according to the specs. However, I have pushed it beyond these official specs using the SDK. It seems to go up to 6.25 MSPS (million samples per second) when using the SDK, even though the specs say it should top out at 1 MSPS.
That's when I realized that this scope would for sure be useful for SDR work. At 6.25 MSPS, the nyquist frequency limit is 3.125 MHz, immediately giving me access to the VLF, LF, MF, and the lower end of HF radio bands, all without a tuner. I originally bought it as a test unit, to see the kind of quality of devices that the Picoscope company made. With the stated maximum sample rate of 1 MSPS, giving me only a 500kHz bandwidth, I wouldn't even be able to receive the AM broadcast band, but I could still get a sense of quality of these devices, so I could see if spending more money later for one with a higher sample rate would be a justified spending of money. However, after discovering that the bandwidth of this scope is over 6 times the official spec, I decided to use this scope as is for my SDR. If it only had a 1 MSPS sample rate, I would have needed a tuner to shift even the AM broadcast band down into the device's bandwidth, but with a 6.25 MSPS sample rate, I can receive 3 full bands (VLF, LF, and MF) as well as the first 125kHz of the HF band (from 3MHz to 3.125MHz) without using any tuner at all. No need to solder together a DBM (double balanced mixer) and an oscillator so as to be able to shift the desired spectrum into the Picoscope's bandwidth.
But now how do you actually use it as an SDR? Simple, connect a long wire antenna to channel A on the Picoscope, and then run the software I made and click Record to start recording, then Stop to stop recording. It saves it to a standard WAV file, which then you can load into any freeware SDR software (HDSDR I found works best for this), and play it. While it is playing, the SDR software behaves as if it was receiving the signal in real-time, and you can select demodulation (AM, FM, USB, LSB), AF/IF bandwidth (or AF and separately IF bandwidth if using FM demodulation), and click on the spectrogram display to select the frequency to tune to within the 3.125MHz bandwidth.
Here's what I've discovered in the different bands:
VLF (very low frequency, 0Hz to 30kHz) mostly has AC hum, which is EMI (electromagnetic interference) at 60Hz and its harmonics. This AC hum goes up to about 10kHz. Above 10kHz it's mostly quiet except for one or two weak signals that appear to be narrow-band EMI. At about 24.8kHz there appears to be an actual intentional signal, and by listening to it and seeing its spectrogram, it appears to be an MSK (minimum shift keying) modulated signal.
LF (low frequency, 30kHz to 300kHz) has both narrow and wide band EMI (most of which are very strong, including EMI generated by CFLs, compact fluorescent lamps), particularly in the lower part of this band (30kHz to 100kHz). Above this there are some signals that appear to be harmonics of EMI signals who's main signal is in the lower part of the band, but the strength of these harmonics taper off pretty quickly. There's a few very weak signals in the uppermost part of the band that seem they may be intentionally transmitted signals, but I'm not sure. They may just be narrow band EMI.
MF (medium frequency, 300kHz to 3MHz) contains the AM broadcast band (530kHz to 1.7MHz). The MF band also seems to contain a few constant carrier signals that are not modulated at all (though some do waver in frequency slightly, but this doesn't appear to be intentional FM modulation), which I assume are narrow band EMI. Well above the AM broadcast portion of the MF band (in fact, over 2MHz), there are a couple FSK modulated signals, which I assume are RTTY (radio teletype). However, they aren't in any ham radio band, and are much wider than ham radio RTTY (about 800Hz between the 2 tones, instead of 170Hz), so I assume these are professional communications. There's also several signals constant carrier signals without modulation and have no frequency wavering, that fade in and out over a period of several seconds, which suggests they are arriving by ionospheric bounce and are fading due to conditions in the ionosphere, which means they are very far away, and likely an intentional transmission (though there appears to be no information being transmitted) as EMI is usually too weak to be received at the distances involved in ionospheric bounce. It takes a very strong transmission to be received at the distances associated with ionospheric bounce (well over a watt of power actually transmitted), while most EMI escaping from a device is limited in strength (due to shielding of electronic devices, as required by the FCC).
LHF (lower end of high frequency, 3MHz to 3.125MHz) just contains the same type of stuff as the upper end of MF, which is unmodulated carriers fading in and out over a matter of seconds. Note that LHF is not an official RF band designation, though HF (3MHz to 30MHz) is official. I'm just using it to show that while my reception does reach the HF band, it is limited to only the lowest part of the band.
And here is the WAV file for download, hosted on Dropbox.
VLF+LF+MF+LHF.wav
Here's the file specs
Size: 677 megabytes
Duration: 56.9 seconds
Sample rate: 6.25 million samples per second
Sample size: 16 bits
Actual precision: 8 bits*
Number of channels: 1
* The analog-to-digital converter in the Picoscope 2204A is an 8-bit ADC (values from -127 to 127, with -128 meaning over-voltage warning), but the kernel-mode driver for it automatically scales the values into 16-bits (values -65535 to 65535, with -65536 meaning over-voltage warning) before sending it to the user-mode application (what you write using the Picoscope SDK combined with your favorite programming language). Therefore, there are only 255 possible valid values (plus an over-voltage warning), even though they occupy 16 bits.
This is a really big file (nearly 700MB) so it will take a long time to download. Also, this is a single channel WAV file, not an IQ WAV file, so the SDR software you use must be able to process non-IQ data. Don't use SDR# for this, as it can only handle IQ WAV files. Use HDSDR instead. To make sure HDSDR is configured right, click the "Options" button (or press F7 on the keyboard) to get to the options popup menu, and then go to the "Input Channel Mode for RX" submenu, and click "Left channel only". To load the file, click the play button (green triangle) to get a file select dialog, and just browse to the WAV file and double-click it. Make sure to click the Loop Playback button as well (looks like an infinity symbol) so it will start over when it reaches the end. This way you can explore the entire spectrum, without having to click the play button when it reaches the end while you are browsing around the signal. Note that that dark cyan progress bar is actually clickable, and you can use it to set the playback point in case you want to rewind to a specific moment and play it from there. All the rest of the functionality of HDSDR is usable as if it were a live signal, including recording of the audio signal, meaning if you find a signal that's interesting and demodulate it, you can record that to a WAV file for playback later.
That's when I realized that this scope would for sure be useful for SDR work. At 6.25 MSPS, the nyquist frequency limit is 3.125 MHz, immediately giving me access to the VLF, LF, MF, and the lower end of HF radio bands, all without a tuner. I originally bought it as a test unit, to see the kind of quality of devices that the Picoscope company made. With the stated maximum sample rate of 1 MSPS, giving me only a 500kHz bandwidth, I wouldn't even be able to receive the AM broadcast band, but I could still get a sense of quality of these devices, so I could see if spending more money later for one with a higher sample rate would be a justified spending of money. However, after discovering that the bandwidth of this scope is over 6 times the official spec, I decided to use this scope as is for my SDR. If it only had a 1 MSPS sample rate, I would have needed a tuner to shift even the AM broadcast band down into the device's bandwidth, but with a 6.25 MSPS sample rate, I can receive 3 full bands (VLF, LF, and MF) as well as the first 125kHz of the HF band (from 3MHz to 3.125MHz) without using any tuner at all. No need to solder together a DBM (double balanced mixer) and an oscillator so as to be able to shift the desired spectrum into the Picoscope's bandwidth.
But now how do you actually use it as an SDR? Simple, connect a long wire antenna to channel A on the Picoscope, and then run the software I made and click Record to start recording, then Stop to stop recording. It saves it to a standard WAV file, which then you can load into any freeware SDR software (HDSDR I found works best for this), and play it. While it is playing, the SDR software behaves as if it was receiving the signal in real-time, and you can select demodulation (AM, FM, USB, LSB), AF/IF bandwidth (or AF and separately IF bandwidth if using FM demodulation), and click on the spectrogram display to select the frequency to tune to within the 3.125MHz bandwidth.
Here's what I've discovered in the different bands:
VLF (very low frequency, 0Hz to 30kHz) mostly has AC hum, which is EMI (electromagnetic interference) at 60Hz and its harmonics. This AC hum goes up to about 10kHz. Above 10kHz it's mostly quiet except for one or two weak signals that appear to be narrow-band EMI. At about 24.8kHz there appears to be an actual intentional signal, and by listening to it and seeing its spectrogram, it appears to be an MSK (minimum shift keying) modulated signal.
LF (low frequency, 30kHz to 300kHz) has both narrow and wide band EMI (most of which are very strong, including EMI generated by CFLs, compact fluorescent lamps), particularly in the lower part of this band (30kHz to 100kHz). Above this there are some signals that appear to be harmonics of EMI signals who's main signal is in the lower part of the band, but the strength of these harmonics taper off pretty quickly. There's a few very weak signals in the uppermost part of the band that seem they may be intentionally transmitted signals, but I'm not sure. They may just be narrow band EMI.
MF (medium frequency, 300kHz to 3MHz) contains the AM broadcast band (530kHz to 1.7MHz). The MF band also seems to contain a few constant carrier signals that are not modulated at all (though some do waver in frequency slightly, but this doesn't appear to be intentional FM modulation), which I assume are narrow band EMI. Well above the AM broadcast portion of the MF band (in fact, over 2MHz), there are a couple FSK modulated signals, which I assume are RTTY (radio teletype). However, they aren't in any ham radio band, and are much wider than ham radio RTTY (about 800Hz between the 2 tones, instead of 170Hz), so I assume these are professional communications. There's also several signals constant carrier signals without modulation and have no frequency wavering, that fade in and out over a period of several seconds, which suggests they are arriving by ionospheric bounce and are fading due to conditions in the ionosphere, which means they are very far away, and likely an intentional transmission (though there appears to be no information being transmitted) as EMI is usually too weak to be received at the distances involved in ionospheric bounce. It takes a very strong transmission to be received at the distances associated with ionospheric bounce (well over a watt of power actually transmitted), while most EMI escaping from a device is limited in strength (due to shielding of electronic devices, as required by the FCC).
LHF (lower end of high frequency, 3MHz to 3.125MHz) just contains the same type of stuff as the upper end of MF, which is unmodulated carriers fading in and out over a matter of seconds. Note that LHF is not an official RF band designation, though HF (3MHz to 30MHz) is official. I'm just using it to show that while my reception does reach the HF band, it is limited to only the lowest part of the band.
And here is the WAV file for download, hosted on Dropbox.
VLF+LF+MF+LHF.wav
Here's the file specs
Size: 677 megabytes
Duration: 56.9 seconds
Sample rate: 6.25 million samples per second
Sample size: 16 bits
Actual precision: 8 bits*
Number of channels: 1
* The analog-to-digital converter in the Picoscope 2204A is an 8-bit ADC (values from -127 to 127, with -128 meaning over-voltage warning), but the kernel-mode driver for it automatically scales the values into 16-bits (values -65535 to 65535, with -65536 meaning over-voltage warning) before sending it to the user-mode application (what you write using the Picoscope SDK combined with your favorite programming language). Therefore, there are only 255 possible valid values (plus an over-voltage warning), even though they occupy 16 bits.
This is a really big file (nearly 700MB) so it will take a long time to download. Also, this is a single channel WAV file, not an IQ WAV file, so the SDR software you use must be able to process non-IQ data. Don't use SDR# for this, as it can only handle IQ WAV files. Use HDSDR instead. To make sure HDSDR is configured right, click the "Options" button (or press F7 on the keyboard) to get to the options popup menu, and then go to the "Input Channel Mode for RX" submenu, and click "Left channel only". To load the file, click the play button (green triangle) to get a file select dialog, and just browse to the WAV file and double-click it. Make sure to click the Loop Playback button as well (looks like an infinity symbol) so it will start over when it reaches the end. This way you can explore the entire spectrum, without having to click the play button when it reaches the end while you are browsing around the signal. Note that that dark cyan progress bar is actually clickable, and you can use it to set the playback point in case you want to rewind to a specific moment and play it from there. All the rest of the functionality of HDSDR is usable as if it were a live signal, including recording of the audio signal, meaning if you find a signal that's interesting and demodulate it, you can record that to a WAV file for playback later.
