Saturday, May 22, 2021

Characterizing the RTL-SDR Blog (Version 3) for HF reception using the "direct" input.

An inexpensive option for SDR (Software Defined Radio) reception on the HF (low frequency) bands is a device sold by "RTL-SDR Blog" - the current iteration being Version 3.  Originally intended for digital VHF/UHF TV reception - and that of FM broadcast - the hardware is also capable of tuning much lower frequencies.

Figure 1:  An RTL-SDR Blog V3 USB receiver "dongle".
Unlike most other inexpensive RTL-SDR dongles, this has - via a single SMA port - the ability to operate in "direct" mode where RF below the VHF frequencies is passed straight to the A/D converter rather than via a down-converter, allowing reception from (theoretically) a few hundred kHz to around 30 MHz.

How does it do this?

The typical RTL-SDR Dongle actually consists of two tunable devices:

  • The Rafael R820T.  This is simply a frequency converter, capable of handling an input signal from somewhere below 60 MHz into the GHz range and converting it to what we'll call an "IF" (Intermediate Frequency) - which is actually somewhere in the 3.5-4.6 MHz range.  In addition to having a programmable oscillator and mixer for frequency conversion, his device has some built-in filtering that provides some protection to strong-ish off-frequency signals, and it has an AGC (Automatic Gain Control) that can adjust the level being output from it to prevent overload of the A/D converter as well as some front-end attenuation control to reduce the likelihood of overload on the input.
  • The Realtek RTL2832U.  The down-converted output of the R820 chip is passed to this device, which consists of two 8 bit A/D (Analog-to-Digital) converters that are clocked at 28.8 MHz, a USB interface, a (reported) microcontroller and a digital frequency "converter" that is also capable of being "tuned" to produce a quadrature "baseband" signal that is output onto the USB port - the rate being programmable from around 250 ksps to 2880 ksps.  This device does NOT have an AGC or gain/attenuation control.

VHF/UHF operation:

Briefly, reception on the VHF/UHF frequencies is done the following way:

  • A simple high pass filter/diplexer passes only the VHF/UHF signals (e.g. those above approximately 40 MHz) directly to the R820T.
  • The filtering of the R820T is programmed for the desired characteristics at the operating frequency.
  • The frequency converter in the R820T is offset from the input signal to provide an output (IF) signal in the 1-14 MHz range - more likely somewhere between 5 and 12 MHz.
  • The level of this output signal may be automatically controlled by the AGC system of the R820T to keep it within the optimal range of the A/D converter.  Similarly, input gain adjustment on the R820T can prevent it from being overloaded by strong signals.
  • The tuner within the RTL2832U is set to about the same frequency being output by the R820T (likely in the 5-12 MHz range) so that it is within the range of output ("baseband") sample rate, and output via the USB port.

HF operation:

  •  The signal to be received is applied via the RF antenna port and diverted from the R820T to a very simple low-pass filter/diplexer to an RF amplifier.  This diplexer's effect begins at approximately 22 MHz - See the discussion near table 1, below.
  • The output of from the RF amplifier is passed through additional filtering and applied to one of the two A/D inputs of the RTL2832U - typically the "Q" input.
  • The tuner within the RTL2832U is set near the frequency to be received so that it is within the range of the output ("baseband") sample rate, and output via the USB port.

There's a penalty to pay for simplicity:

For those familiar with receiver topology - digital or analog - several things the above description of the HF operation of the RTL-SDR dongle without proper measures (e.g. filtering, gain control) should give cause for concern - namely:

  • There is NO bandpass filtering at all.  Whatever is being intercepted by the antenna system will be input directly to the A/D converter via the preamplifier.
  • ANY signal applied to the antenna input - no matter the frequency - can contribute to overload.  Because the inputted RF goes into the A/D converter, a strong signal well away from where you are listening can cause overload.  For example, if you are listening to 14 MHz, a strong AM broadcast signal (e.g. mediumwave) could well be the culprit if you are experiencing overload.
  • There is no AGC in the HF ("direct") signal path.  Considering that the A/D converter is only 8 bits, this means that compared to even low-end shortwave receivers, the range of signal levels over which this device will operate is very narrow.  The lack of an "AGC" (automatic gain conttrol) means that there will likely not be enough gain when signals are very weak and overload of the A/D converter is likely if the signals are very strong.
  • The A/D converter's sample rate is 28.8 MHz.  What this means is that the Nyquist limit - the frequency above which the digitized output can no longer represent the input signal - is half this, or 14.4 MHz, uncomfortably close to the 20 meter amateur band, and entirely below the 17, 15, 12 and 10 meter bands - and this does not even consider the fact that the sample frequency is within the 10 meter band itself. 
What this means is that there are unsuppressed image responses all across the HF spectrum.  For any frequency you tune below 14.4 MHz, you can also hear any signal above 14.4 MHz, the frequency of which is calculated by subtracting its frequeny from the sample rate (e.g. 28.8 MHz).  For example, 15 meter signals will also appear, spectrally inverted (e.g. USB = LSB) 7.80-7.35 MHz.  Image frequencies appear in table 1 - along with MDS and clipping values - below.

In other words, if you are tuned to any HF frequency, you are actually "hearing" TWO frequencies at the same time unless you have specific filtering to prevent such image response.  See the right-hand column of Table 1 (below) for the frequency at which you will see an image.

The above put strict limits on the performance of the RTL-SDR dongle as an HF receiver and these realities must be considered when the configuring any system that might use them.

The usable dynamic range:

In theory, 8 bits of A/D sampling would indicate about 48dB of useful signal range, but the reality is more complicated than this.  Compared to the bandwidth of the narrow signals typically sought on HF, the overall sample rate of the A/D converter - and even the rate after the RTL2832U's converter has reduced the signal to the sample rate being sent to the USB port - are very much higher, effectively improving the bit depth via oversampling - and the fact that the HF spectrum is backgrounded with what amounts to white noise works to our benefit, helping to spread discrete, spectral artifacts that are the inevitable result of the imperfect signal aquisition.

All of this makes the actual, usable range a bit difficult to divine.  To this end, an RTL-SDR (V3) unit was put on the workbench, using the "HDSDR" program as a receiver, and its operation was analyzed by observing CW (unmodulated) signals on narrow (SSB) bandwidths.

Methodology:

Ideally, we would be able to determine the RF level at which the A/D convert clipped directly, but the HDSDR program does not provide a means to see the peak A/D level, requiring us to infer it by noting the level from a known-accurate signal generator at which the S-meter starts to decrease at the same rate that the signal level is reduced - and also by noting the disappearance from the waterfall many of spurious signal caused by overload.  Typically, this is about 3dB below the "maximum" S-meter reading.

For sensitivity, DL4YHF's "Spectrum Lab" program was used to measure the SINAD in a 500 Hz bandwidth set by the HDSDR program, the "minimum discernible signal" being equivalent, in this test, to 3dB S/N.

Table 1:  Measured signal levels for A/D clipping and MDS, along with corresponding image frequencies on HF.  Because the sample rate of the RTL-SDR is 28.8 MHz, ALL signals above half this frequency (e.g. 14.4 MHz) are, by definition, Nyquist images.
Frequency (MHz)
Clipping (dBm)
MDS (dBm)
Image  (MHz)
0.5
-17
-97
28.3
1.0
-24
-107
27.8
2.0
-27
-109
26.8
4.0
-27
-116
24.8
7.0
-29
-117
21.8
10.0
-30
-117
18.8
14.0
-30
-117
14.8
15.0
-30
-116
13.8
18.0
-30
-116
10.8
21.0
-29
-114
7.8
24.0
-25
-109
4.8
28.0
-14
-95
0.8
29.0
-16
-94
0.2
35.0
-16
-90
6.2
40.0
-2
-84
11.2

Discussion:

Table 1 tells us several things:

  1. As mentioned in the documentation found at the RTL-SDR Blog web site, signals below 2 MHz - and especially below 1 MHz - are rolled off by the "Bias Tee" blocking choke which has insufficient inductance at frequencies below the AM broadcast band.  To a degree, this effect can be mitigated by removal of the RF choke which will remove the capability to inject DC onto the cable.
  2. The RTL-SDR has a low-pass filter to diplex the HF and VHF and above frequencies to separate signal paths, and the effects of this filter are becoming evident above 15 meters (21 MHz).  This also means that by itself, the RTL-SDR becomes "deaf as a post" on the 12 and 10 meter bands.
  3. Note that at the very low frequencies, the sensitivity appears to be somewhat reduced.  The effects of high-pass roll-off - likely caused by coupling capacitors and the input bias inductor - appear to be evident through at least 7 MHz.
  4. There is clearly no image rejection at all, considering that the sensitivity through 21 MHz is comparable to that below 14 MHz.  For example, 15 MHz WWV will also appear at 13.8 MHz.
  5. Low-pass effects are evident by 24 MHz, most likely a result of the diplexer used to split the HF (direct) and VHF/UHF signal paths.  This limits sensitivity on the 12 and 10 meter amateur bands.
  6. With only 8 bits of quantization, additional noise will be generated due to the imprecise nature of the process.  This "noise" will show up as spurious signals and, less obviously, as a rise in the overall noise floor - depending on the nature of what is being digitized.  In short, the fewer the number of bits, the less likely it is that weak signals will coexist (and be audible) in the presence of strong signals.
  7. Across most of the HF spectrum, the RTL-SDR will overload signal of about -30dBm - which is approximately equal to an S-meter reading of "40 over S-9".  While this seems like a fairly strong signal, this power level represents the total amount of RF energy - no matter the frequency!
  8. Note that the power at which the A/D converter starts to overload is approximately -30dBm across much of the HF spectrum.  This represents the TOTAL amount of RF power required, at all frequencies combined, that will result in overload.

Points 7 and 8, above, should be considered very carefully in terms of its implications:

  • You cannot simply connect an RTL-SDR Dongle to even an "average" performing HF antenna and expect reasonable results as the total power from ALL signals reaching the receiver are likely to exceed the -30dBm signal level.
  • Particularly on the lower bands (80, 40, 30 meters) the signal levels of amateur and especially shortwave broadcast signals can, by themselves, exceed the -30dBm overload level - particularly in Europe and the eastern U.S. On some bands, the shortwave broadcast band adjacent to the amateur band (e.g. 41 and 40 meters) are too close to effectively filter out and it may be that an RTL-SDR is simply not usable on these bands when propagation favors reception on these frequencies. 
  • As seen in Table 1 (above) the RTL-SDR dongle becomes increasingly deaf on the higher HF bands (particularly 12 and 10 meters) making them unusable at these frequencies without additional amplification AND filtering.  What's worse is that the lower HF frequencies (e.g. below 10 MHz) are typically very noisy while the higher frequencies are quiet by comparison.  If you connect an antenna to the RTL-SDR dongle with no band-pass filtering and try to tune in, say, the 15 meter band, you will likely hear only noise (and signals) around 4.8 MHz, which will likely overwhelm any weak, 15 meter signals.
  • If you plan to use an RTL-SDR for HF reception and expect even mediocre performance, you should precede it with a band-pass filter for the frequency band of interest.  For the highest HF bands (15, 12 and 10 meters) the typical noise floor in a quiet location is around -120 dBm (in a 500 Hz bandwidth) - which is well below the noise floor of the RTL-SDR at these frequency meaning that a preamplifier (along with a bandpass filter) will be required for reasonable performance.
  • It's worth remembering that unlike an analog receive system, one cannot always use all 8 bits for digitization:  The signal input must be kept well below the "full scale" level to accommodate for random fluctuations that are ever-present on signals input from the antenna.  What this means is that the A/D must be under-driven overall and that fewer bits are actually being used most of the time.  In order to maintain suitable margin, it's typical to drive the A/D converter at between 1/4 and 1/2 full scale, meaning that for 8 bits of A/D conversion, 2-4 bits are typically being used.

Conclusions:

  • DO NOT simply connect an RTL-SDR to your HF antenna and expect it to work as well as even a low-end shortwave receiver:  If it doesn't get overloaded by local AM broadcast (mediumwave) signals, it will get overloaded by strong shortwave broadcast signals when conditions are favorable on certain bands.  This isn't to say that you won't hear anything if you do so, but know that normal signal levels present on even an "average" antenna will be enough to overload the RTL-SDR dongle.
  • ALWAYS precede an RTL-SDR with a band-pass filter that is specific for the frequency range of interest. For example, if you are interested in 40 meter reception (7.0-7.3 MHz for ITU region 2) your filter should pass only frequencies in this range, and this filtering will prevent unwanted reception of signals at the image frequency around 21.8 MHz.   Unfortunately, the use of a band-pass filter precludes reception outside its design range, but this is necessary considering the limited capability of the RTL-SDR dongle in terms of handling both strong and weak signals at the same time and its unfettered response to unwanted images.
  • In some cases - even with a mediocre antenna - the signals in the desired frequency range may exceed the signal handling capability of the RTL-SDR and cause overload.  As noted above, a single signal stronger than -30dBm can do this, but so could a number of signals whose total power can exceed this.  Typically, this overload is manifest as intermittent distortion across the entire receiver as signals fade in and out.  In such cases, it might be beneficial to attenuate the signals reaching the RTL-SDR as the degradation caused by overload is more "destructive" across the entire receive frequency range than too-little signal.
  • Because of image response and roll-off, the HF port of the RTL-SDR is really not well-suited for 12 and 10 meter (24-30 MHz) reception.

Comment about HF upconverters:

Specifically to address some of these issues, there are upconverters available for the RTL-SDR that will upconvert the HF spectrum to VHF (typically in the 100-130 MHz range) to allow the use of that signal path, making use of the Raphael R820T converter.  This converter has the advantage of having a degree of band-pass filtering and the ability to use AGC (automatic gain control) on the signal path.

This method can be useful, but there are several caveats:

  • If frequency stability is of importance, the addition of the upconverter introduces two additional frequency stability issues:  Drift of the upconverter itself, and the fact that any existing drift in the dongle itself will be multiplied because of its operation at the higher frequency.  This can be an issue in environments where the temperature is not stable and/or when a frequency sensitive mode like SSB or (especially) digital modes are used.
  • If multiple bands and receivers are to be used, there may be the temptation to upconvert the output of the upconverter to VHF and distribute this signal to the receivers.  While this may work in many cases, it's worth noting that if the entire HF spectrum is converted, the total signal power level can be significant, potentially overloading the upconverter itself, any RF amplifiers that might be used at VHF, and/or the front end of the RTL-SDR dongles themselves.
    • While a "fix" might normally be to filter out just the HF band(s) of interest, this can become impractical if the 40 meter band (7.0-7.3 MHz) is upconverted to, say, 137 MHz where it can become difficult to make an effective band-pass filter at that frequency.
    • It is possible to filter for a specific HF band before the upconversion, but this means that each RTL-SDR would require its very own upconverter.  While effective, the cost of an upconverter for each band may be prohibitive.

* * *

Successfully using an RTL-SDR on HF:

As mentioned earlier, one must precede the RTL-SDR with a band-pass filter to obtain reasonable performance on HF - or any other frequency range where very strong and very weak signals will be simultaneously present at the antenna input:  Remember that, especially in the "direct" mode, all signals applied to the RF input - even those MHz away from where the receiver is tuned - will count "against" you in terms of the total amount of RF power that may be applied to the A/D converter before it clips/overloads.

To see some RTL-SDR based HF receive systems in operation - and to be able to directly compare them with higher-performance receivers using 16 bit A/D converters, visit the Northern Utah WebSDR site at sdrutah.org (link).  This will take you to a "landing page" where you can select several receivers - specifically:
  • WebSDR #3:  This server uses RTL-SDRs for both the 80 and 40 meter bands.
  • WebSDR #1:  This server uses 16 bit sound cards and "softrock" receivers for reception of the 80 and 40 meter bands, split into 5 segments.

Both of these systems use the same antenna for 80 and 40 meters and it is possible to directly compare signals between the two in side-by-side windows.  Generally, the 8 bit RTL-SDRs used on WebSDR #3 hold their own compared to those on WebSDR #1, and this is possible only because the WebSDRs are preceded with both band-pass filtering and AGC as described in the link below.

* * *

Additional resources:

  • An article on using band-pass filtering and AGC (Automatic Gain Control) to improve the performance of an RTL-SDR when used for amateur band service may be found here:  http://www.sdrutah.org/info/rtl_sdr_agc.htmlSeveral "test" receivers are currently in operation at the Northern Utah WebSDR that demonstrate the efficacy of doing this.


 This page stolen from ka7oei.blogspot.com

[End]

Friday, April 30, 2021

Repair of a (lightning struck?) Yaesu FT-1000 MP Mark V

The background - and initial assessment:

A month or so ago an acquaintance of mine in the local amateur community asked me to take a look at his radio - a Yaesu FT-1000 MP Mark V.  While I don't routinely repair other people's radios, I decided to make an exception, as I've known this gentleman for decades.

The complaint was "The receiver is dead" - which is a very broad assessment, but it was an obvious starting point and upon putting this rather large, ponderous piece of equipment on my workbench - which, itself proved to be a challenge - I noted that it took about a 0dBm signal to attain an "S-9" reading:  Yes, the radio was deaf, to the tune of around 70dB!  Knowing that the antenna A/B relay and/or antenna tuner could be a problem with this particular radio, I tried different configurations - even the (differently-routed) rear "RX Antenna" jack - but no difference.

Figure 1:  A blast mark!
Getting it apart:

This radio's technology and construction is what I would refer to as "transitional" -  mostly reminiscent of 80s Japanese radios in that there is a mix of through-hole and surface mount, and the boards are mostly interconnected with a myriad of white wires with unmarked plugs on them that go everywhere in the case - with the occasional gray coaxial cable that use the odd board-mounted plug-in cables.  Fortunately - unlike some of those older radios - the boards are mostly double-sided epoxy rather than single-sided paper phenolic.

Of course, the main receive RF board was buried under the very large heat-sinked power amplifier assembly, so I first did a test on the "sub" receiver board, which was accessible:  That receiver seemed to be reasonably sensitive when I injected a signal directly into it, but there was no sign of RF getting to that receiver via the radio's rear coaxial cable connector.

The next, obvious step was to remove the PA board - which, like many things about working on this radio,  was a real pain:  Only four screws, but two of them were buried under a mass of white wires -  I pulled the amplifier module out from the radio with trepidation, knowing that I'd eventually have to get it back together.

A visual and an olfactory inspection of the receive board - before removing it - was done, and nothing was obviously amiss, so I removed a bunch more screws and cables - marking them as appropriate (you will hate life if you don't mark where they went as you remove them!) I was able to pull the board away and immediately saw the first indications of the problem:  A very obvious black mark on the aluminum chassis under the board and a corresponding area of char on the board itself (see Figure 1, above and Figure 2, below).

What might have happened:

Figure 2:  Carbonized crater in the RF board.
While it is difficult to determine what, exactly, happened to this radio, I suspect that it was likely a nearby lightning strike that did the initial damage, with peripheral damage occurring later.  In inspecting the radio and referring to the service manual, which I found online, I saw that capacitor C1278 on the "RF unit" - and its related circuit board trace, and a portion of the board itself - had been wholly transformed into the black mark on the aluminum chassis.  The fix for this was pretty simple:  Scrape away all of the now-conductive PC board carbonized residue and replace the vaporized surface mount 0.1uF capacitor with an axial-leaded unit to both replace both the component itself and the trace as seen in Figure 3.

Knowing that this was not likely to be the only problem, I half-installed the board again and checked the receiver:  Very slightly less terrible - but still deaf as a post, but with access to the board, I cranked up the signal generator to 0 dBm and started poking around with the oscilloscope to see where RF disappeared.

Figure 3:  New cap, carbon excised!
Fortunately, I was seeing RF at the input to the RF board - and even though many of the key probing points were on the (inaccessible!) bottom side, I found a few places where I could test on the top and discovered that the RF signal stopped at D1056, a PIN diode found at a low-pass filter stage in the RF path.  Pulling the board again I started probing with the multi-meter and discovered that not only was this PIN diode completely open, but R1157 and L1086 - both part of the DC path of this and (the still-working D1055) were open:  I had a suitable SMD replacement for R1157, but used a small, molded 1mH inductor for L1086 to accommodate the damage due to a blown-off pad and trace that hadn't been immediately obvious and I used a much more rugged PIN diode than the original to replace D1056.  While I was at it I did more probing around and found that R1150 - part of what appears to be a DC drain on the antenna circuit - was also burned open:  Because this resistor can only be "reached" via the 1 mH inductor in series with it, this meant that it was not likely RF that blew it up, but rather a transient with a strong DC component - or at least a low-frequency AC component - maybe lightning?

Initial re-testing:

Putting the board back in, temporarily, I re-checked the receiver sensitivity and found that it was "OK", in that I could touch a screwdriver and put it in the rear-panel RF connector and hear signals, but a quick check with the signal generator indicated that something was still amiss as it seemed to be off by about 15dB based on the specs in the alignment procedure.

At this point I decided to check the transmitter and to my gratification, I was able to get about 90 watts out of it.  My initial satisfaction was short-lived, as I soon realized two three things:

  • I should have been able to get at least 150 watts out of the transmitter.
  • The SWR indicator on the radio was showing a mismatch, with the tuner bypassed, into a known-good load.
  • I smelled epoxy smoke from the "RF Unit".

Figure 4:  Original TX isolation relay - welded!
At this point a felt a bit silly:  I should have checked to see that the unit had a functional TX/RX isolation relay - and a quick check with a signal generator and probing with the oscilloscope told me that it did not, the apparent culprit being  RL6414, a reed relay on the "Tuner-Main" unit that was clearly welded - a fact soon verified by multimeter.  The radio's power amplifier had dutifully cut back its power because it was seeing a mismatch caused by the receiver being connected at the same time as the dummy load.

At this point I again removed the RF unit and replaced D1056 (again!) - the apparent source of the smoke that I'd smelled and - without transmitting - restored the operation of the receiver.  In reviewing the service manual and online forums, I discovered that the failure of RL6414 was semi-common, and also that this particular relay - seen in Figure 4 - was difficult to source.  Nevertheless, I sent an email off to Yaesu Parts to find out.

A few days later I'd heard back from Yaesu:  This relay was available - but it would take 6-8 weeks.  In the meantime I'd tried to find an exact replacement elsewhere, but to no avail:  The original relay had a non-standard pin-out and was a brand that was simply not carried by U.S. parts suppliers - and I couldn't be sure if this particularly relay was still made!  In speaking with the owner of the radio I gave him two options:  Order the part and wait 6-8 weeks, or get a more common part and adapt it to fit:  He opted for the latter, so I placed an order with DigiKey, set the radio aside and waited.

It's worth noting that RL6414 is an SPDT relay with the RX signal path connected in the "Normally Open" position with the "Normally Closed" position grounded - that is, the relay must be energized for the receiver to be connected to the antenna.  If the radio is turned off, there is no direct path, so whatever "killed" the receive, must have either happened with the radio turned on, while it was receiving, or it was sufficiently energetic enough to weld the relay and, apparently, blow away the ground "N.C." contact:  Yet more evidence of a "high energy discharge" from lightning.

Resuming work:

Figure 5:  New relay on homebrew carrier board.
About a week later I again had some time to work on the radio and the relay from Digi-Key had arrived.  This relay - a Comus BFH-1C-12C (Digi-Key P/N:  1835-BFH-1C-12C-ND) was an actual RF reed relay with decent voltage and current ratings - but the pin-out was different from the original, requiring me to construct a simple "carrier board" to reroute some of the connections of the relay footprint to match the tuner board:  With plenty of headroom, the extra height of the combination was not an issue.

At this point, I will note again how difficult it is to work on this radio - highlighting, in this case, the antenna tuner.  This module consists of two boards, face-to-face, meaning that no components are accessible unless they are separated.  Unfortunately, there are several wires that appear to have been pulled through from the "lower" board (with the tuning capacitors) and then cut short - and none of the other wires were any longer than they absolutely needed to be.  Disassembly was pretty easy (especially with proper desoldering equipment) but re-soldering the short wires was an exercise of patience and the careful manipulation of small tweezers and screwdrivers to try to align all of the wires simultaneously without causing one or more of them to pop out of place!

Figure 6:  New TX isolation relay - looks right at home!
With the TX isolation relay (RL6414) once again functioning I found that the transmitter's output now exceeded 150 watts with no reflected power on the radio's meter - and no smoke!

Alignment - and more problems:

Not having part of the transmitter signal path in parallel with the receiver input helped the sensitivity a bit, but it still wasn't right so I did a bit more checking with the signal generator and scope, finding no-where that things were obviously amiss on the RF Unit - but I did notice that the secondary receiver was more sensitive than the main:  Since they share the same signal path that more or less ruled out an obvious problem with the RF unit when operated in the normal fashion so I decided to perform a realignment.

To my surprise, the I.F. stages in both receivers were quite far out of alignment - particularly on the main receiver - and after finishing this, the receivers "woke up" and met specifications:  I was surprised to see this on a modern receiver and am at a loss to explain it.  It seems unlikely that it detuned itself with age, so it was either not properly aligned at the factory, or someone else tried to "improve" it - but I'm suspecting the former.

In performing the alignment steps, I then "discovered" that the "VRF" circuit - an electronically-tuned preselector - was inoperative, about 25 dB of excess attenuation:  This problem was traced back to being caused by RL1005 having welded contacts, not allowing this preselector circuit to be fully switched into the signal path.  This not unexpected as this is in the same signal path as the vaporized trace and capacitor.  Fortunately, these exact relays were readily available from Digi-Key as well.

All's well that ends well:

The replacement of RL1005 turned out to be the last problem that was found and the radio easily met its specs when the job was done.  The reassembly of the radio was, as expected a bit of a challenge:  The inboard mounting flange of the PA unit shares the same channel as dozens of thin, white wires and there is practically no visibility as one tries to maneuver the screws into place while trying to avoid them getting knocked off the magnetic driver while, at the same time, trying to make sure that none of the dozens of small, white wires get pinched in the process - all the while trying to align to invisible holes!

In the end, the radio was handed off to the owner, with the recommendation of added lightning protection or, at least, disconnecting things when not in use - good advice for anyone!

This page stolen from ka7oei.blogspot.com

[End]


Wednesday, March 24, 2021

Using the QRP Labs "SoftRock" as a possible crystal replacement in "rock bound" VHF and UHF radios

As technology moves on, it is getting more difficult to find custom-made quartz crystals.

Why would one need a quartz crystal in these days of frequency synthesizers?  The can't really be beat for simplicity and stability in many applications. While it may be preferable in many cases to just use a cast-off synthesized radio, there are still some compelling reasons to use older "rock bound" radios used for linking and repeater where frequency agility is not important.

Many of these older radios have multi-section, narrow-band helical resonator front ends which provide protection above and beyond many modern radios and repeaters for rejection of off-frequency signals - often very important for linking and repeater sites:  Many users have replaced their old "rock-bound" repeaters with "store bought" Icom or Yaesu repeaters at busy sites only to discover that these new radios - with their intrinsically broad front ends - were being clobbered by signals that managed to get around their duplexers!

For an explanation as to why that happens, read THIS article about duplexers:  When "Band-Pass/Band-Reject duplexers really aren't band-pass"

Given that one might be using a rock-bound radio for a specific application, the recent demise of International Crystal Manufacturing (ICM) in the U.S. makes these crystals harder to obtain although there continue to exist other companies, such as "Quartslab" in the U.K., that can supply custom crystals.

What about retrofitting such a radio with a synthesizer?

The "ProgRock"

The "ProgRock" (short for Programmable "Rock" or crystal) is a device sold by QRP Labs in the U.K. (link) consisting of two parts:

  • An Si5351A-based frequency synthesizer
  • An Atmel-based microcontroller board that sets the frequency of the synthesizer.
The Si5351 chip, made by Silicon Labs, contains several on-board oscillators and dividers making it capable of producing frequencies ranging from high audio to somewhere "north" of 200 MHz with "part per billion" resolution over most of that range.  Capable of up to three independent frequency outputs, only two outputs are typically used owing to hardware limitation related to the number of internal synthesizer blocks and the fact that there is a bit of crosstalk between each of the three channels.

What this means is that while in theory one could use a single ProgRock for both receive and transmit simultaneously - as in a repeater - it may not be a good idea as low-level spurious signals from the "other" channel may cause low-level spurious responses:  While this may not be an issue for receive, it would be a bad thing for a transmitter - particularly when used for full-duplex repeater service where extremely low-level spurious signals may get into the receiver!

On the other hand, if the radio in question is half-duplex (e.g. not receiving and transmitting at the same time) one of the features of ProgRock, the ability to select multiple channel "banks", can be used to enable one channel with the receive frequency and then the other channel for the transmit frequency as needed.  By having just one synthesizer active, the production of low-level spurious signal is minimized.

Case study:  The GE Mastr II

The GE Mastr II is a crystal-controlled product line made from the mid-late 1970s into the early 1980s and aside from a few minor issues (e.g. the tendency for the receiver 's helical resonator castings to grow "hair" [e.g. metal dendrites] and make the receiver deaf - a problem easily remedied by the application of spray-on clear coat after cleaning and degreasing) it is a solid performer with very strong receive filtering and the ability to get many of the parts on the new and surplus market.  What we need, then, is to effectively simulate a crystal oscillator module (called by GE an "ICOM" - no relation to any other radio company!) to produce the desired frequency.

One potential advantage in this application is that the transmitter itself is modulated using phase modulation.  What this means is that the crystal itself (or any other type of frequency source) is unmodulated, with the audio added in the later stages using variable reactance techniques and this means that we do not need to figure out how to modulate our synthesizer directly.

Having an unmodulated signal source as our transmit frequency determining component is actually quite convenient as it provides the opportunity to "nail down" the transmit frequency with as much precision as you wish, allowing multiple transmitters to be "synchronized" to precisely the same frequency.  In contrast, transmitters with modulated crystals or PLLs have a tendency to "wander" a bit during modulation by amounts that may not be precisely matched between units and this can cause a bit of extra "clashing" of phase and frequency in overlap areas.

On the GE Mastr II receiver the crystal frequency is multiplied by a factor of 9 for high-band VHF (e.g. 2 meters) and 36 for UHF (70cm) typically using an IF (Intermediate Frequency) of 11.2 MHz in each case.  For transmit, the crystal frequency is multiplied by 12 for high-band VHF and 36 for UHF.

What this means is that for a hypothetical repeater on 146.940 MHz with a -600 kHz split, the receive local oscillator would be operating at (146.340 - 11.7) = 136.140 MHz with a crystal frequency at 1/9th of this, or 15.12666 MHz.  The transmitter would use a crystal frequency of (146.940 / 12) = 12.2450 MHz.

For a repeater operating on 449.000 MHz with a -5 MHz split, the receive local oscillator would operate at (449.000 - 11.2) = 432.80 MHz with a crystal at 1/27th of this, or 16.029629 MHz while the transmitter would use a crystal frequency of (449.000 / 36) = 12.472222 MHz.

We can tell from the above numbers that for a receive frequency in either the U.S. 2 meter or 70cm amateur bands we would need to produce a signal in the general area of 15-16 MHz for receive and around 12 MHz for transmit with the GE MastrII radios.

Synthesizing the receive local oscillator's crystal:

Let us first analyze the case of the receive local oscillator for our hypothetical 449.000- repeater.

Figure2:
Signal path of the receive crystal showing the multiplier stages (highlighted).
Note that the output of the crystal is sent almost immediately to a frequency
tripler stage which allows us to inject a frequency at 3x the original
crystal frequency at that point.
Click on the image for a larger version.
As calculated above, we would ultimately need to produce a 432.80 MHz local oscillator, nominally from a crystal operating at 1/27th of this frequency, or 16.029629 MHz, and because we can actually program the Si5351A, via the ProgRock, to Hertz-level resolution,  we have the potential of being within a few 10s of Hz of the actual frequency.

If one examines the signal path (see the block diagram to the right) from the output of the crystal oscillator unit you can see that one of the first stages after the crystal unit (the "ICOM") is a frequency tripler meaning that we don't need to start at 1/27th of the local oscillator frequency, but at 1/9th - three times that frequency, or 48.088888 MHz in the example above, effectively eliminating the need for the lower frequency that would get multiplied, anyway.  In so-doing we get a bit better frequency resolution and potentially lower spurious signal generation and better phase noise response - more on that later.

Synthesizing the transmit crystal:


Figure3:
The signal path of the transmit crystal showing the modulation and
multiplier stages, highlighted.
The first stage beyond the crystal is the modulator meaning that we must
synthesize a frequency that is the same as the original crystal.
Click on the image for a larger version.
 For the transmitter, things are a bit different as can be seen in the block diagram to the left.

Instead of going directly to a frequency multiplier as in the case of the receiver, the first stage beyond the oscillator unit is the phase modulator.   For various reasons, when phase-modulating a transmitter it is best to start at the lowest frequency practical so that the small amount of linear modulation that is possible at a given frequency is multiplied along with the signal.  What this means is that we are stuck at the 12.472222 MHz frequency that we calculated, above.

Modifying a GE "ICOM":

As mentioned above, the GE MastrII uses modules called ICOMs as the frequency-determining components with these modules containing the crystal, an oscillator circuit, a tuning capacitor and some additional electronic frequency-tuning circuitry - namely a varactor diode, and some additional circuitry to provide channel selection and transmit keying.  There are three types of ICOMs commonly found in VHF and UHF radios:
  • "2C" ICOM.  This is nominally rated for 2ppm frequency stability and is typically used at UHF.
  • "5C" ICOM.  This is nominally rated for 5ppm frequency stability and is usually used at VHF.  It can provide an external temperature-based frequency control voltage that is "shared" with "EC" type ICOMs.
  • Figure 4:
    A buffer/driver circuit added to an EC ICOM.  This
    circuit is constructed within the ICOM unit itself.
    Click on the image for a larger version.
  • "EC" ICOM.  This "Externally Compensated" ICOM uses the voltage provided by a 5C ICOM elsewhere in the radio for frequency stability.  Normally, one would never use an "EC" ICOM all by itself as it has no temperature compensation of its own and without a 5C driving it, its frequency tuning voltage may be indeterminate.
The most common type of ICOM by far is the "EC" version - and this version is all we will use.  In terms of modification, one need only remove the original crystal and the tuning capacitor and install the buffer/amplifier circuit shown in the figure to the right.

Figure 5:
The insides of the modified "EC" ICOM with the added circuitry
built "dead bug" inside the original ICOM enclosure.  The metal
can is removed for thispicture with the coaxial cable entering
through the hole that originally accessed the tuning capacitor.
Click on the image for a larger version.

This circuit, a simple common-emitter amplifier, assures that signal from the synthesizer is at a consistent level - and one that is capable of fully-driving the original oscillator circuit to full output.  While we can build a complete, new circuit for the ICOM, retaining its original oscillator circuit is a convenient means of keying the transmitter using the original PTT circuit in the MastrII radio.

While a bit cramped, there is enough room to build the simple circuit shown in the diagram above inside the original can.  When this is done, the input of the circuit and its connection to coaxial cable is arranged such that it passes through the hole originally used to access the tuning capacitor.  This circuit is insulated from the can using the "fish paper" that originally insulated the circuity and crystal from the can.

Putting it into practice:

To see if the synthesizer produced a transmit signal of acceptable quality, I decided to use a ProgRock in a 70cm transmitter.  Because of the higher multiplication factor of the UHF transmitter, this would be a good "worst case" test to determine if the spurious content of the transmitted signal was cause for concern, as well as analyze the same signal in terms of phase noise to see if the synthesis process caused additional degradation of spectral purity.

Because it was on-hand, I tested using a UHF GE Mastr II transceiver, using unshielded clip leads to connect directly to the "ICOM" socket, including a 1000pF blocking capacitor to protect the ProgRock's Si5351a chip.  The results were encouraging:  A very subtle increase in the background noise (hiss) on both receive and transmit (the TX and RX testing was done separately) but no other obvious issues.

Figure 6:
The as-built filter, using three 42IF129 10.7 MHz IF "cans".
Click on the image for a larger version.
In looking at the spectral purity of the transmitted signal, a slight amount of added noise was seen on a spectrum analyzer - well into the -90dBc range - within a few 10s of kHz of the carrier.  Farther out, in doing an "A/B" comparison with a crystal on the same frequency we saw a few low-level spurious signals, but doing some "hand waving"  (literally putting one's hand near the temporary clip-leaded wires) indicated that most of this was being picked up on the flying leads.  A visible spur was seen at +/-1.32 MHz, but this was immediately recognized as being from an AM radio station located about 1.5 miles (about 2.4km) away and had a "proper" installation been done with shielded cables, it's likely that this would not occur at all.


Making it better:

Without making things complicated, there is little that can be done about the low level hiss, largely attributed to phase noise of the synthesizer - but it's unlikely that this would even be noticeable in a typical installation, and if VHF were used instead, it may not be audible at all.  What would be prudent would be the use of a narrowband L/C bandpass filter  on the output of the synthesizer to further-filter its output and knock down farther-removed signals (and their mixing products) that might find their way into the transmitter - particularly those related to the 27 MHz reference oscillator.

Figure 7:
Schematic of the filter using 10.7 MHz IF transformers.
Click on the image for a larger version.


In the "old days" one option of filtering the oscillator output was with the use of tunable 10.7 MHz IF transformers, such as the "42IF129" that had been available from Mouser Electronics and other places as these could be tuned from below 7 MHz (with an added capacitor) to around 13 MHz with no modification - and even higher by removing the internal capacitor and using a smaller value external cap.  Unfortunately, obtaining any 10.7 MHz IF transformers these days is difficult - although several surplus and hobbyist parts suppliers still have a few on hand.

The lack of ready availability of this particular type of component means that we would need to make a similar unit ourselves - described below:

Figure 8:
An alternate version of the bandpass filter using more commonly-available components.
Click on the image for a larger version.

This filter is depicted in Figure 6, using readily-available components.  The frequency range can be tweaked as needed:  Changing the 82pF capacitors to around 100pF will allow tuning down to about 10 MHz while lowering them to 47pF will allow tuning around 14 MHz (20 meters).  This filter's insertion loss (about 6 dB) is comparable to that depicted in Figure 7, with 15-20dB of attenuation at +/- 1 MHz and at least 40dB attenuation +/- 3 MHz.

Implementation:

The ProgRock has three outputs available, but depending on your application you may need to deploy two of them, as noted above:

  • For half-duplex use, a single Prog-Rock could be used, one output for TX and the other for RX.  In this case one would use the available "channel banks" and program it such that for TX, only that specific output would be enabled.  For RX, one would use one of the channel bank bits to select a configuration in which the output of the TX is disabled and the output for the RX is enabled.
  • For full-duplex use, the use of TWO separate ProgRocks is recommended:  Even though it is able to output both frequencies simultaneously, the outputs of the synthesizer are susceptible to producing low-level spurious signals as these two outputs interact on the chip itself - and for this reason the isolation of two, individual units is suggested.

Mentioned above are the "channel banks":  There are three bits available for selecting up to 8 separate configurations.   If, on a half-duplex radio, one of these bits is used for selecting the TX and RX frequencies, two other bits remain available allowing up to 4 different frequencies to be selected.  Similarly, in cases where two separate ProgrRocks are used (one for TX, the other for RX) up to 8 separate frequency configurations is possible.

Finally, the ProgRock boards ship with an imprecise "computer" crystal that is not temperature stable.  The newer versions of the ProgRock have the additional footprint for a TCXO (Temperature Controlled Crystal Oscillator) that will hold the frequency to within 2.5ppm - more than adequate for 2 meter usage:  An article describing how to retrofit the unit with a 1ppm TCXO - suitable for UHF use - may be found here.

* * * * *

This page stolen from ka7oei.blogspot.com

[End]


Sunday, February 14, 2021

The appearance of the "Chinese Woodpecker" on the HF bands

Listening about on 40 meters this morning I heard a familiar sound - the "putt-putt-putt" of what sounded very much like the infamous "Russian Woodpecker" Over The Horizon Radar (OTHR) of the 1970s and 1980s - even having the same 10 Hz repetition rate that was common for the woodpecker.

Just by looking at the waterfall display, I could see that this signal was quite different:  Rather than taking 100s of kHz of bandwidth, this signal seemed to be fairly well contained within a bandwidth of a few 10s of kHz implying techniques unlike those of the signal from the 70s.

In other words:  It sounded like the bad old Russian Woodpecker, but it clearly was not.

Pulse versus chirp:

In the "old" days, a lot of RADAR systems simply blasted out a pulse of RF energy and then listened for the echo.  "Because physics", there is a 4th power distance relationship of reflected RF (e.g. doubling the distance causes a signal to decrease by a factor of 16) so tremendous radiated power levels were required to receive enough energy from the return pulse - which had to be intercepted over a wide bandwidth to get precise timing - from the object off which it had bounced.  Distance may be ascertained by timing the delay between the transmit and return pulse, often integrating this information over many pulses.

Audio clip of possible Chinese OTHR as heard on the KFS WebSDR system on
40 meters.  Significant backscatter is apparent in this recording, but the signal was
very much stronger and "cleaner" on remote receivers in Asia.
(Another type of signal - the "buzz-buzz" - was heard mid-recording and is not related.)

Many modern RADAR systems transmit a CW (continuous) signal that, instead of being pulsed, is swept in frequency.  Rather than relying solely on the time between the transmit and return pulse, one can measure the difference in frequency between the transmit signal and received (reflected) signal because the transmitter will have shifted frequency by the time the reflected signal arrives, and the greater the round-trip distance, the greater the frequency difference.  In other words, instead of timing the pulse directly - because there isn't one - the frequency difference, using a receiver that has a local oscillator that effectively tracks the transmitter's frequency, is what indicates distance.

Because the latter case uses a CW signal and a tracking receiver, one may use narrowband techniques (anything from a simple, narrow filter or an FFT with multiple "narrow" bins) on the receive end, potentially obtaining 10s of dB of processing gain.  In other words, to obtain the same Direction, Range and Distance information, far lower power may be required than with the old-fashioned pulse-type RADAR for comparable results.

It is likely because this "new" OTHR radar is chirp - that is, a swept-frequency transmitter - that its spectrum is far-better contained than the "bad old" Russian Woodpecker, and it's likely that the effective radiated power - while still quite high - is far lower.

A bit of sleuthing:

In the past, amateur radio operators could ascertain the location of the Russian Woodpecker only by obtaining antenna headings from multiple stations around the world and then compiling the data to determine a likely location of the transmitter.  These days, we have other methods at our disposal - and as readers of this blog will be aware, one of these is the KiwiSDR "TDOA" network.

In short, the TDOA network consists of a number of participating KiwiSDR receivers around the world that, when commanded to do so via the software extension's control panel, will record a GPS time-stamped audio file from the selected receivers and send them to a server that will analyze this data and determine the apparent location of the received signal.

This, I have done several times over the past several weeks, and today I had the opportunity to do it several times more - the results appearing in the maps below:



KiwiSDR TDOA results from the "Woodpecker" signal heard on 40 meters from remote receivers.

The above maps show the results of several TDOA sessions using several receivers scattered across the Pacific and Australasia - the lower image being derived from higher-quality data as propagation improved - and, possibly, as the transmitted beam was better-oriented in the direction of the receivers, reducing the appearance of backscatter.  Because these transmissions are (apparently) rapid frequency sweeps, they are nearly ideally suited for the type of analysis needed to determine the Time-Difference Of Arrival (TDOA) techniques employed - in other words, a RADAR in reverse.

Due to the vagaries of ionospheric propagation - and as should be apparent from the roughness of the numbers given - the absolute location of the transmitter is likely to be accurate within only about 100km at best for this sort of exercise - but something is clear:  It is probably not likely a "Russian Woodpecker", but more likely something akin to the " 啄木鸟   中国人 " - which Google tells me means something roughly akin to "Woodpecker from China".

Comment:

There is a known Chinese OTHR transmit site near Nanjing (approx. 32.05°N, 118.78E) but that location does not correlate with the results above.  It's very possible that the TDOA error is quite large, but the coordinates on the maps shown above were similar for several runs:  More TODA runs, over time, may help to resolve this uncertainty.

"Will my radio's noise blanker help?"

Probably not!

Unlike the bad old Russian Woodpecker that consisted of narrow (and very broadband) pulses, this appears to be a swept carrier, meaning that unlike the old, Russian variant, a pulse-type noise blanker is unlikely to work well at all:  Rather than the pulse being "everywhere" within a few 10s (or 100s) of kHz of the desired receive signal simultaneously in the case of the Russian Woodpecker - and other impulse noises like vehicle ignition, electric fences and lightning - this signal is only on one frequency at any given instant and the wideband amplitude detector comprising a standard impulse-type noise blanker will likely be ineffective.  This "new" signal sounds like a pulse only because it spends a small amount of time in an SSB receiver's passband during each sweep.

In theory, it should be possible to design a software-based filter that will remove this signal as it is very repeatable, but it's likely that no software-defined receivers in common use at this time (e.g. as of the original posting of this article) will have anything at all that will touch it!

* * * * *

This is not the first time that a Chinese OTHR has appeared on the ham bands:  For years, now, one can hear what sounds like a loud "buzz" that comes and goes as described on some of the links below.  The appearance of the "10 pps" version - possibly a refinement for longer-distance observation - seems to be comparatively new.

* * * * *

Related links about the Chinese OTHR:

* * * * *

This page stolen from ka7oei.blogspot.com

[End]



Sunday, February 7, 2021

RFI (Radio Frequency Interference) sleuthing with the TinySA - Part 1: The gear

Although focused on the use of the TinySA, the techniques and equipment described in this article may be applied to using any other type of Spectrum analyzer - or even a simple receiver - for the detection of interference/noise sources or general-purpose direction-finding.

Figure 1:
The Tiny SA, connected to an outdoor HF antenna during
daylight hours.  The displayed span is 2-22 MHz.
Click on the image for a larger version

About the TinySA:

The "TinySA" is a small spectrum analyzer with a 2.4" touch screen, costing approximately U.S. $50.  Capable of operating from below 100 kHz to as high as 950 MHz (the optimal range being between 100 kHz and 350 MHz) its minimum resolution bandwidth (RBW) is about 2.6 kHz.  This resolution bandwidth is too wide for precise analysis of the components of an SSB or NBFM signal, but it is useful for general RF surveying - including the measurement of harmonic and off-frequency spurious components of a transmitter in addition to the detection of low-level signal sources.

While it may seem silly to use a spectrum analyzer for the detection of signals, the TinySA, at about U.S. $50,  is only about 2-3 times the cost of the least expensive, battery-powered shortwave receiver that you'll find online - and it has the advantage of "seeing" a large swath of spectrum in a single view, allowing detection of signal sources that may otherwise go unnoticed.

The usable sensitivity of the TinySA in the range of 1 MHz to 350 MHz (e.g. the "Low" input) is approximately -152dB/Hz meaning that its sensitivity is slightly poorer - but roughly comparable to - "real" spectrum analyzers with expensive-sounding names:  This value is quite a bit poorer than a typical communications-grade receiver which will likely have a sensitivity better than -160dB/Hz - often much better.

The TinySA comes with a telescoping antenna, but being only about 12" (25cm) long it's usefulness extend below VHF frequencies (e.g. 50-100 MHz):  At HF, the combination of the 50 ohm input impedance of the analyzer and the phenomenally poor mismatch of the small telescoping whip results in uselessly-poor sensitivity, meaning that one must be nearly atop a signal source before it may even be seen on the analyzer.  Clearly, more help is needed here!

The TinySA is NOT the NanoVNA!

The TinySA is NOT electrically similar to a NanoVNA, despite physical similarities:  It has completely different circuitry and works like a "real" spectrum analyzer in that it's a proper, swept frequency, narrowband logarithmic receiver-detector.
 
While a NanoVNA does have a detector that tracks its built-in signal generator, it makes a terrible spectrum analyzer!

Where to get a TinySA?

The TinySA is widely available - but there are apparently many inferior clones out there (e.g. lack of shielding, poorer quality or omitted components, etc.)  

For the U.S. readers, I suggest R&L Electronics as a source (see the link HERE).  I have no connection to R&L and suggest them only because they are supplying the genuine article and they are an established equipment dealer (e.g. more likely to help if you have any problems with the unit), and they are offering it for a decent price.

* * * * *

Active antennas to the rescue:

Adhering to the theme of portability, a useful companion to the TinySA would be some sort of small, active antenna - and two of the most common types are the active (E-field) whip and the electrically-small loop.

The Active Whip:

A popular active whip is of the "PA0RDT" design - the so-called "Mini-Whip".  It is the intent that this antenna be mounted outside and in the clear with a quiet "local" ground - that is, well away from noise sources, grounded to a mass of metal (or a ground system) that is isolated from the (noisy!) shack ground with a common-mode choke or two - so using it as a "sense" antenna for sniffing RF sources isn't exactly what the designer had in mind.

Figure 2:
A homebrew E-field whip using the PA0RDT design.
This unit is built into a piece of 1/2" PVC irrigation pipe.
A BNC connector was chosen over an SMA for durability.
Click on the image for a larger version
The reason for this specification of a "clean" ground and being placed outside, in the clear is that this antenna - being a tiny fraction of a wavelength on the highest frequency for which it is to be used - is essentially a capacitor that couples into the "æther" (e.g. free space).  To accomplish this, the whip is just a Hi-Z probe that has no gain of its own, but rather a circuit that matches the impedance of the "antenna" portion (usually a piece of wire or a plate of copper on a circuit board) to 50 ohms typically using JFET and bipolar follower circuits, avoiding what would otherwise be tremendous losses due to mismatch - but this is only half of the antenna:  The feedline - and the gear to which it is connected - is the other half - and this is the basis of the recommendation of a "local" ground with the feedline being decoupled via a common mode choke to minimize the conduction of RF energy from a "noisy" shack ground.

As a "sense" antenna in this application, its purpose is largely to determine the presence of RF energy:  Sources of this energy may be located by noting the amplitude increase as one moves the antenna nearer to the potentially-noisy device - or near a conductor (e.g. power cord, cable, etc.) that is conducting or radiating this RF energy.  Because the antenna will be carried by the user, the "ground" portion of the antenna will be nothing more than the hand capacitance of the user holding the equipment and the connecting cables - and it is imperative that these not be a source of interference in their own right or connected to something that could be a source of such signals.

Figure 3:
Inside the E-field whip of Figure 2. The "antenna" is the
copper foil "plate" with the electronics having been built
on an SMD prototype board.  This version uses the
BF862 JFET as the source-follower and a BFU589GX as
 the emitter-follower.
Click on the image for a larger version.
Because of the vagaries of coupling energy into this "whip" - which include the inconsistencies mentioned above - it's difficult to make absolute signal level measurements, but relative amplitude measurements can be made by noting signal levels as it's moved about.  Because it lacks directionality and is relatively insensitive to polarization, proximity is indicated mostly by relative amplitude:  If one happens to be close to the suspected device or interference source, this is easy to discern as even small changes of distance to the source can markedly change amplitude, but if you are somewhat distant from it this will not be the case and the lack of a directional response can be a challenge as one may not know in which direction one should move to get closer.

A few more articles about the small whip - its construction and its operation:

dl1dbc.net/SAQ/Mwhip/Article_pa0rdt-Mini-Whip_English.pdf   - The diagram of this ubiquitous circuit is that in Figure 7.

472khz.org/media/pa0rdt-Mini-Whip-Manual.pdf

pa3fwm.nl/technotes/tn07.html

And an "improved" version:

pa3fwm.nl/projects/miniwhip/

 

Where to get a mini-whip?

If you choose not to build one, the so-called "Mini-Whip" and its clones may be found on EvilBay and Amazon - to mention but two places - typically for $25 U.S. or less.  All of these devices work reasonably well - the ones from Eastern Europe often performing slightly better and being more consistent in quality than those from Asia - and they come with a "power inserter" (a.k.a. "Bias Tee") - a device that injects DC power for the amplified antenna onto its coaxial cable.

These antennas are also easy to construct, the details being found - and the theory of operation - in the above links.

* * * * *

The Small Loop:

This name refers to an "electrically small" loop - that is, one with a circumference that is a small fraction of the wavelength of the highest frequency for which it to be used.  For LF, MF and HF purposes, we are referring to a receive-only loop that is typically 1-3 feet (25-75cm) in diameter.

Figure 4:
The completed shielded loop.  The gap in the shield is on
the top with the transformer and amplifier in the box at the
bottom.  The loop itself is 3/8" "Heliax" supported by a
frame made of 3/4" PVC irrigation pipe, using a "cross" in
the center and "tees" on the sides.
Click on the image for a larger version
Unlike the active whip, this antenna can be entirely "self contained" in the sense that the amount of energy impinging on the loop that is being conveyed to the receiving device and the cable itself has little effect on this level.  Additionally, this type of antenna exhibits a degree of directionality with sharp, deep nulls being present normal to the plane of the loop - which is another way of saying that if you were to hold the loop flat against your chest, the nulls would be directly in front of and behind you with very broad peaks to your left and right.

This loop, being electrically small for the intended wavelengths, has negative gain compared to a full-sized antenna, but even a chunk of coaxial cable connected directly to the loop with no attempt at matching will work "less badly" than a piece of wire or telescoping antenna - with no amplification or matching - of similar size.  Having said this, including a modest amount of gain in the loop's signal path is extremely helpful.

Having consistent "gain" lobes and nulls means that the amplitude readings from this type of antenna are more predictable and it is possible to divine the apparent bearing of the noise source by exclusion.  Because this type of antenna has two nulls (and two broad peaks) one must move about to resolve this ambiguity, noting the geometry of the direction of the these null and your surroundings to determine if it is very close or distant - and whether it is in front of or behind you!

Where do I get a loop like this?

Unfortunately, these loop antennas are nowhere near as ubiquitous as the "Mini Whip" - and those that are available tend to cost about an order of magnitude more than a mini-whip.  Fortunately, these types of antennas are not particularly difficult to construct - and such a loop is described below.

While this type of antenna may be constructed using just  loop of wire - and the techniques shown in Figure 7, below, will work fine with just a piece of wire, the preferred version is the "shielded loop" as it is somewhat less sensitive to very nearby E-field interference and its pattern of broad peaks and nulls can be more symmetrical and of better quality - as discussed here:

w8ji.com/magnetic_receiving_loops.htm

and

owenduffy.net/antenna/shieldedloop/index.htm

 * * * * *

Which type of antenna should I use?

If I had to choose just ONE of these types of antennas, I would pick the loop antenna due to its directional properties using the nulls.

Having said that, I would prefer to have both on hand:  The whip is much smaller and can easily be held near suspect devices and conductors to aid in detection/exclusion.

* * * * *

Constructing a shielded, symmetrical, untuned loop:

This portion of the article will described the "untuned, balanced, shielded" loop, and adding a simple amplifier to extend the noise floor of the TinySA.  There are many ways to construct this type of loop, but what follows is a rather simple and effective version that will tolerate component variations quite well.

In addition to locating sources of noise and interference, it provides reasonable performance for general listening on frequencies from below the AM broadcast band to the top of HF, and its ability to cast a null toward a noise source may prove to be useful.  While lacking the apparent "gain" response of the resonance peak of a tuned loop at a specific frequency, an untuned loop is easier to build and more convenient to use in a broadband application.

What to use for the loop itself?

It common to use "Heliax" (tm) coaxial cable (50 or 75 ohm is fine) to construct the loop itself - typically of the 1/2 inch or 3/8 inch variety as it is fairly rigid and can support itself fairly well, mechanically.  Aluminum-jacketed CATV (cable TV) "hardline" will work as well - although unlike the Heliax - which has a copper jacket - making connection to the outside shield can be a challenge.  

"I don't have any hardline/Heliax - can I use 'normal' coax?"

Barring the availability of "scraps" of Heliax cable, ordinary coaxial cable will also work quite well, but it must be mechanically supported as it may not be able to hold any shape -  The frame depicted in Figure 4, above, can do this.  If you are using "ordinary" coaxial cable, practically any coax will work, from RG-58 to RG-11 (the impedance doesn't really matter):  Even inexpensive RG-6 cable will work if one uses standard "F" connectors  - both on the cable itself and on the box containing the electronics - to make reliable connections to the aluminum shield found on most cables of this type.

Figure 5:
Details of the gap in the middle of the shield - the dielectric
foam being retained for additional support.  This gap
should be covered with heat-shrinkable tubing or electrical
tape to protect it from moisture.  Sliding a piece of rigid
tubing over the gap is recommended to immobilize the cable
at the gap in the shield.  A small piece of metal tubing may be
used for this as long as it's insulated from the shields.
Click on the image for a larger version.
For the loop described, I used a 66" (167cm) long scrap of 3/8" Heliax (FSJ2 type) which would correlate with a loop diameter of approximately 21" (53cm) - but the precise length and diameter is unimportant in this case:  While a larger loop (with a greater area inside the loop) will intercept more signal, a loop that is between 12" and 24" (25-50cm) diameter will work.

In the center of the loop, as depicted in Figure 5, there must be a small gap where the shield is opened, symmetrical about the loop's circumference:  Were a gap not present, this would simply be a loop of coax and a signal could not be excited on its center conductor with respect to the shield portion.  This type of loop has the advantage that an electrostatic field cannot easily excite the inner conductor, but the lack of the "shorted turn" provided by the gap in the shield means that the magnetic field portion of the signal is unimpeded.    (Hint:  Determine the center of the length of coaxial cable that you plan to use for your loop before you form it into a loop!)

Feeding the loop:

Figure 6:
Inside the box - the transformer and loop amplifier.
The transformer (T101 in Figure 7, below) may be seen in the
middle.  The circuitry is built "dead bug" on a piece of glass-
epoxy circuit board material bolted to the bottom of the box.
The shields of the halves of the loop are connected with
heavy wire to the circuit board - both for electrical bonding
and mechanical support - and wire (a.k.a. "zip") ties are used
to prevent the cable from being pulled out through the sides
of the box.  The RF out/DC in is via a BNC connector
mounted to the side of the box - this being preferable to
the bottom of the box to help protect it from damage in
handling.
Click on the image for a larger version.
The two open ends of the loop are brought together opposite the shield gap and it is there that the shields are connected to each other at a common signal "ground" - and this is typically done in a small enclosure:  While a metal box is preferred, it's not absolutely necessary.  The box shown in Figure 6 is a die-cast aluminum enclosure approximately 5-3/4" x 2-1/2" and 1-1/2" deep (about 122 x 65 x 40mm).

While one may simply connect a 50 ohm coax to either side of the center conductor of the loop - and connecting the shield of that cable to the shield of the loop itself - this will cause a bit of asymmetry.  Practically speaking, this isn't going to cause much of an issue in terms of directionality, but it is preferred that balance be maintained to provide the best symmetry in its response.  While there are many designs out there that include differential amplifiers, the use of a simple, ferrite transformer will be just fine for this - and most - applications.  Remember:  We are looking for noise sources!

For this, a small transformer must be constructed.  A good choice for this is a toroid - either an FT37-75 or FT50-75.  Experimentally I determined that either 2 or 3 turns on the primary and 3 or 4 turns on the secondary will provide good response across the HF spectrum (I used 3 turns on each, for a 1:1 turns ratio).  The gauge of wire is not critical and something in the range of 18-26 is suggested with a preference on the larger size.

Amplification:

Because of its small size, the gain of the loop will be very low compared to a full-sized antenna and to bring the signal level up to where weaker signals may be "seen" by the TinySA, a simple amplifier is used, depicted schematically in Figure 7, below.  This amplifier could be considered to be a general-purpose "gain block" as it is quite well-behaved in terms of stability and input and output impedance - and it is very forgiving in terms of component variation.

Circuit description:

Figure 7:
Schematic of the loop/amplifier and power inserter (a.k.a. "Bias Tee").
Click on the image for a larger version
 

The shielded loop is schematically depicted in Figure 7, above, along with the transformer T101, which I found to work reasonably well with a 1:1 turns ratio.  Using a VNA and with a bit of empirical testing, this configuration was optimized for low-mid HF (e.g. 3-15 MHz or so) but it seemed to work quite well through the AM broadcast band and below.  Like many similar loops - and the Mini-Whip - its performance will start to drop off at higher frequencies (above 20 MHz) and a bit of circuit redesign would be required to optimize for these higher frequencies.  As noted, the circumference of the loop must be a small fraction of the wavelength at the highest frequency at which it will be used in order to maintain the directionality of the nulls, so a loop of approximately 18" diameter will work well throughout HF, but not offer the desired properties on, say, 2 meters where its circumference would be approximately a quarter wave!

T101 does a reasonable job of maintaining the symmetry of the loop itself - although purists would insist on a somewhat different topology to eke out every bit.  As described, the symmetry is quite good and it is possible to completely null local 50 kW AM broadcasts stations.

Amplifier portion:

Q101 and associated components form a simple feedback-type RF amplifier.  This basic circuit is well-behaved and has reasonable input and output matching to 50 ohms.  Shown is the use of the common 2N3904 transistor which is perfectly acceptable for this sort of use - and similar devices, such as the 2N2222 and 2N4401 - work pretty well.  A "better" device would be the 2N5109, 2N3866 or similar RF amplifier which will offer a bit better performance in terms of intermodulation distortion and gain at higher frequencies, but they are more expensive and harder to find - and are likely overkill for a "sniffing" device.  This amplifier is not the penultimate in performance (e.g. IP3, P1dB, etc.) but it has very good performance and fairly low noise (6-8 dB noise figure) considering its simplicity.

To a degree, the gain of this stage may be adjusted by varying the value of feedback resistor R101 - resistances over the range of 330 to 680 ohms being useful, with higher gain (roughly 18 dB) being associated with higher resistance.  The purists will note changing devices or feedback resistance will alter the properties of the amplifier (e.g. input/output impedance, etc.) but one can generally ignore this in all but the most critical applications - such as matching to an impedance-sensitive filter network.

Although a battery could be placed within the loop's enclosure - in which case L101 could be eliminated and the V+ lead connected to the junction of C104/R104 - it is common to use a power inserter (a.k.a. "Bias Tee") that couples DC onto same coaxial cable that conducts receive RF from the loop, making one bias tee useful for multiple antennas!

How much RF noise does the TinySA produce?

Because it's a computer with a display, one might wonder how much RF noise the TinySA itself produces.
 
The designers of the TinySA appear to have been very careful about this - for example, critical components are shielded and they to have chosen to use linear voltage regulators instead of more-efficient switching-type regulators.  (I can't speak to the construction of the many clones out there!)
 
In testing, the TinySA was held up to the E-field whip and placed inside (and moved around) inside the circumference of the loop:  Through the range of 2-22 MHz, there are no obvious "spikes" or lumps of noise that appear on the display - at least above the ambient RF noise floor in my ham shack.
 
Clearly, one isn't likely to walk around with the TinySA held against the antenna - but the initial glance shows that you probably could get away with it!

The power for the amplifier is picked off the coaxial cable by decoupling choke L101.  The value of this choke is not critical - and anything above about 100uH will work fine to a bit below the AM broadcast band.  

Molded chokes may work, but they tend to have quite high internal resistance and the current consumption of the amplifier itself (50-80 milliamps) can cause a significant voltage drop.  An alternative is to use the same core as that used for T101 (you did get several FT50-75 toroids, didn't you?)

Power inserter (a.k.a. "Bias Tee")

The "power inserter" (a.k.a. "Bias Tee") is also included on the diagram and its job is to combine the RF and DC onto the same cable, using the same type of choke (L201) as on the amplifier portion.  Optional enhancements to this device would be a diode (D201) to protect against accidental reverse-polarity application to the antenna and an LED to show that power is turned on (LED201) with its current intentionally set low (e.g. a 10k limiting resistor for about 1mA) to minimize battery drain.  The addition of a fuse - preferably of the self-resetting thermal type (F201, with a current rating of 100-300 mA) is a good idea as well to prevent damage to the blocking choke, L201 should the output be accidentally shorted and to limit current into protection diode D201 should reverse polarity be applied.

If you happen to get a "Mini Whip" from one of the online sources, it will likely come with a power inserter/bias tee that is electrically very similar to that depicted above - and that device will work just fine with the loop/amplifier depicted in Figure 7.

* * * * * 

Using a shortwave receiver for RFI sensing and source identification:

Up to now we haven't mentioned one RFI-locating tool that may be already in your possession  - a portable shortwave receiver.

Figure 8:
A small selection of inexpensive portable shortwave
receivers.  Upper-right:  A DAK MR-101 - an inexpensive
receiver from the early-mid 90s.  Lower-right:  A Grundig
YB-400PE - the only one shown that is capable of CW/
SSB reception.  On the left:  A $15 receiver (labeled
LCJ 310) - one of many nearly-identical units available from.
Amazon capable of tuning from 5.9-21.85 MHz on SW and
on the AM and FM broadcast bands.  Of the receivers
pictured, only the Grundig and DAK have external
antenna connections (a 3.5mm "phono" connector) which
would be required to use the directionality of the loop antenna.
This connector would be needed to be added to the "LCJ"
receiver to take advantage of this antenna property.
Click on the image for a larger version.

While one can only "hear" a small bit of spectrum with a receiver rather than "see" a larger slice with an analyzer and increase the likelihood of detecting an otherwise-unnoticed interference source, being able to tune into an interference source and listen to it can give a clue as to the source of this signal and type of device.  If your portable shortwave receiver has an external antenna jack - especially one that automatically disconnects the built-in antenna when a connection is made - you can use the same antennas (whip, loop) described above and take advantage of any directional/nulling properties.

For example, a mains-frequency "hum" in the noise implies a switching power supply while a sharp "buzz" might indicate a triac light dimmer.  If your receiver has a BFO, one may be able to hear the harmonics of a switching supply and be able to specifically identify it by it's unique "sound" as it's powered on and off, being able to distinguish it from other devices.


A caveat with the use of a receiver:  Other than "missing" interference sources on frequencies other than that to which the receiver is tuned, a receiver will have an AGC which, by its nature, will adjust the internal gain to keep the volume constant.  When trying to locate a signal or noise source, this can work against you as it may be difficult to determine if the signal is moderately weak, strong, or very strong unless the AGC can be defeated and a manual RF gain control be operated - something that cannot usually be done with very inexpensive receivers.  Even if this is the case, the "sound" of the interference can still be useful in providing a clue as to the type of device that may be causing the interference.

* * * * *

Again, there are many possible ways to do this, but the gear described above has been proven to be useful.  The next installment of this two-part series will include details on how it has been used and what to expect when doing so.

* * * * *

This page stolen from ka7oei.blogspot.com

[End]