Receiving weak signals with the FT-991 #hamradio

In this blog I would like to spend some time on the use of a number of features of this transceiver from Yaesu which can be quite useful to receive those weak signals that otherwise remain hidden in the noise. Rather than listing the functions of all the buttons as in the manual I want to discuss here an operating practice for receiving those noise QRP stations.

Once you have found a weak station first try to tune it such that the signal becomes maximal, use the fast setting or the channel dial/multi knob only for quickly browsing through the frequency band. Next turn fast off, and tune so that the weak station sounds best. The clarifier button in combination with the clarifier knob can help to adjust the final frequency offset that makes SSB modulation sound like Donald Duck rather than a human voice. (Added: study the difference between clarifier RX and just clarifier, also, eventually turn the clarifier back to zero when you are done with the station that needed it and return DXing with other stations, see comments by PA2QXT at the end of this blog). But now the noise.

Listening to noise is problematic, actually, you will usually fall asleep if you do this late in the evening. Most people give up when they hear a weak voice that is hidden in the noise. There are several types of noise, natural and man made noise. In radio language QRN is for natural noise, QRM is for the man-made noise. Sometimes QSB is used to indicate signal variation like fading originating from changes in the ionosphere. Other forms of QRN are lightning strikes which are clearly audible at all frequencies. You can’t do much against QRN or QSB except to wait when it goes away. High Kp indices with corresponding Ap averages are typical for QSB, this is because the magnetic field is changing rapidly so that signal reports rapidly vary when you listen to a station. You can’t do much here, maybe the advise should be to not give up because sporadic long distance contacts may be possible under these conditions.

The repeating ticking noise that you sometimes hear at many frequencies are usually caused by something in your neighborhood, oftentimes the noise blanker helps to filter the ticks away, so for this we have the NB button. Only use it when the ticks are there, otherwise leave it off. Reason for leaving NB off is that audio may be distorted which is otherwise ok, see the comments from PA2QXT with this blog.

If there is noise of neighboring transmitters or local oscillators then we speak about RFI which stands for radio frequency interference. For this I describe an experiment with a home-built Arduino on youtube. All oscillators in unshielded casings around us may cause some form of RFI. Locate such devices with your SDR like the airspy or tools like the MFJ 805 and change the situation. In particular, make sure that the local noise generators are grounded and that they are within a metal case, or otherwise shielded within at the point where the RFI is generated. It may also be that cable connectors are the problem or that unshielded cables are used. Homeplug PLC systems are really bad, in particular the older ones. If they exist in your neighborhood then ask your neighbors to change to for instance WIFI or a cabled network.

Back to the buttons on the transceiver. My experience is that the NB button usually does not really help against RFI and that you need a better ‘weapon’. PA2QXT reports that the NB does help against the noise of switching power supplies, but also warns that once the NB is on that audio may become distorted because of the NB. The noise of switching power supplies may be suppressed with the NOTCH and the CONT filters that are helpful when your are listening to a station close to a RFI source. With the CONT filter you can suppress the splatter of a station near the frequency where you are listening, the CONT filter is a wide band adjustment to your reception window. The NOTCH filter is good for suppressing a very narrow part of the spectrum within the range you are listening to a weak station caused by an RFI source that comes with a narrow band width. Switching power supplies may appear everywhere in the spectrum usually at every 50 or 100 kHz, sometimes slightly drifting around. If power supplies are really noisy then maybe it is also time to replace them, or alternatively if you can’t do anything against it, to go on a camping trip away from all the QRM. PI4CQ is a club in the Netherlands that does this.

The computer and the screen that you probably want to use near your transceiver is a well known source of RFI because it is full with local oscillators that pollute the spectrum. This used to be the case in particular in the early days, nowadays you can simply ground the metal case of the computer, it takes most of the effect away is my experience. It could also mean that you are sometimes better of without a computer near the antenna or the transceiver. Hard to say, but, rather easy to find out.

There is also noise caused by the receiver itself, if it is the case then we speak about a birdy. If you get a new radio then first learn where all the birdies are, for this you can remove the antenna and just quickly tune through all the bands. My former FRG-7700 receiver had quite a few birdies, in particular between 25 and 30 MHz. The new FT-991 hardly has any. Memorize the birdies so that you don’t waste your time on them. If there are many birdies, or if the problem gets worse over time, then maybe your radio needs maintenance. The NBCONT and NOTCH filters will probably not help against birdies.

Once you have experimented with the NB, NOTCH and CONT then you could continue with the more difficult step, which is to make the signal more audible. To improve the signal to noise ratio you need to experiment with the WIDTH filter, the AGC, DNR and the attenuation knob called RF gain on the transceiver. Normally I first experiment with the AGC response setting and the WIDTH, change them until you’ve found something that works best for your situation. The AGC is the gain feedback, the input into the AGC system is obtained after demodulation and what you adjust is the response time for adjusting the gain at the HF front end amplifier and the intermediate frequency amplifiers of the receiver. Sometimes a fast response time works best, for instance when there is QRN from lightning, but more often a slow response time works better, it is hard to say. Do something that you like best is my advise. The AGC response times can be modified in the menu, for operation you simply choose between off, fast, mid,  slow and auto. AGC off is hardly used because chances are that you enter in the range where non-linearity of the amplifiers makes this cracking sound. AGC at auto is often a good choice although some radio-amateurs like AGC slow better for weak stations.

Next comes the DNR filter which is a digital noise reduction filter that you find on the FT-991. It is a very effective tool and I found that it works best in combination with the RF gain dial on the mid right of the FT-991. What I initially do is to leave the DNR off and turn the RF gain clockwise to maximum gain, and then play with the preamplifiers (IPO/AMP1 and AMP2) until you’ve found the point where the signal sounds best. For signal strength reports you should (I believe) report the signal strength on the S meter that you get with an IPO, the AMP1 and the AMP2 just add a number of dBs per amplifier, this is fine, but AMP1 and AMP2 also increase the noise that comes with the signal. Once you turn the pre-amplifiers on you add a number of dBs that show up in the S meter. The signal strength to report is then the difference between the no-modulation value and the modulated value. All the 59’s that you hear all day are probably exaggerated, my opinion is that the pre-amp added dBs should not go in the signal strength report, but I doubt whether this practice is also carried out by most radio amateurs. We hear you, the signal is ok, so it is a 59. No, this is not the way to do it.

Suppose that AMP1 worked best without RF attenuation on the 20 meter band. A weak station may then still be difficult to hear, but now turn the  DNR to like 1 or 2, that is, hit the square on the display and use the multi-knob to change the DNR filter value. How does this sound? Probably it does sound a bit better but there is still noise, also, it may sound as if someone is speaking through a tube. Next turn up the RF gain attenuation knob (it turns counterclockwise to attenuate more). At some point all the noise disappears, but the voice still remains. The audio gain (AF) can now be turned up, at some point it becomes optimal is my experience. There are more filters on the FT-991, there is also a DNF filter and sometimes this helps to remove some hiss in the received signal if it still persists. I don’t often use to DNF filter, but apparently it has a purpose sometimes. The DNF option could be a last step in this process.

If you ever decide to do PSK, WSPR or any other digital mode then all these filters, including a speech processor (PROC button) or microphone equalizer (MIC-EQ button), should not be used. In this case your computer has to process the signals, not the transceiver. The best results for digital modes were obtained with a pre-amp setting at AMP2 on the FT-991. For the transmission of PSK signals I describe in this article what you should do with the optimal audio input gain.

With the above described operating practice you will see that even a poor signal to noise ratio stations can still be received pretty well. It can take some time to carry it out, but you will see that it really works. Take some time to understand fundamentally how your transceiver really works, the manuals do not help here, and the exam material for your license only partly helped here because there is no common interface for radio receivers.

In particular the noise makes you tired when listening to those weak QRP stations.If you ever have trouble sleeping, then put up a headset and listen to white noise, chances are that you sleep right away. In any case I do recommend a headset because it suppresses the local dBs that you hear. Also, the dynamic range of a headset is oftentimes much better than the computer audio box that I otherwise listen to. When you live with a spouse and children as I do then a headset is certainly recommended.

If I had some experience with different transceivers then I could write a similar manual as in this blog, the filters probably have different names. Previously I owned a FRG-7700, bought it around 1980 and sold it (with a caveat emptor) last year because it was becoming deaf on the amateur bands. Digital filters did not exist in the 80’s, also waterfall displays and the ability to easily store and recall frequencies. The FRG-7700 had a real time clock and I was quite happy with the local oscillator stability, the selectivity of the roofing filters and the birdies were somewhat of an issue, but otherwise it was a really nice receiver. Quickly memorizing a few frequencies is easy on the Ft-991, for this there is the store/recall button, press it shortly and you recall a stored frequency, press it longer and it stores a frequency. This is a first in last out (FILO) type of memory and I find it convenient.

I hope that this blog assists you in your ham operations. Good luck DXing.

Last update: 29-5-2016, PA1EJO Ernst





Yesterday evening we had a lot of noise caused by thunderstorms with lightning passing south of us over Belgium and Limburg. It caused a lot of noise on 20 and 40m, today we have something similar, but it is further south below Paris.

Update 19:37 on 28-may-2016, this is the current lightningmap that I found here Every discharge results in the QSN and QSB.


Dipole or Windom antenna?

A standard dipole has 1/4 wavelength at each side relative to the center feed point. The standing wave ratio relative to a 50 Ohm transmission line that we can expect for such a dipole depends on the height above the ground, and it is found to change between 1.4 and 2.0, the first optimum is obtained around 11.5 meter above the ground. All calculations assumed an antenna length of 10.35 meter and a frequency of 14.2 MHz.

Oftentimes it is inconvenient to feed the dipole from the center, one would need two masts and the coax cable attaching in the center. More practical is to put the feed point closer to one side, this idea leads to the so-called windom antenna. I assumed a version that had 1/6 of a wavelength on one side, for instance the side running to your house, and another 1/3 running into the garden to a pole. This saves at least one mast to worry about, the antenna was discussed by Loren G Windom, W8GZ  written in the September, 1929  issue of the QST Magazine.

The windom antenna, image source

The half-wavelength dipole has a theoretical impedance of 73 Ohm that results in a SWR of 1.46 for a 50 Ohm transmission line. The antenna tuning unit in the transceiver can usually match this, as long as you stay under 3.0 when the FT-991 is used. But as said before, the actual value one obtains for the SWR depends on the height of the dipole above the ground which we assumed to be an “average ground” in the NEC-2 software.

For the windom antenna discussed before you find a greater impedance that also depends on the height, with NEC-2 I found a minimum value of 1.7 at 12 meter above the ground. Point is however that the windom antenna has a larger range of SWR values when we vary the height, I found with NEC SWRs between 1.7 and 2.5 which is still acceptable for most transceivers.

With the help of my autotuner (an YT-1200 from LDG) one can tune over a much wider range, namely from 4 Ohm to 1000 Ohm, but the maximum power it can handle is limited. A balun is always required for both the dipole and the off-centered dipole (the windom antenna), a 1 to 4 ratio is sufficient but it is already an overkill. A 1 to 2 balun would be good enough, possibly a 1 to 1 balun would also do the job in combination with the autotuner or the ATU in the transceiver.

Attached is the predicted plot of the SWRs of both antenna’s as a function of the height above to the ground, the values were estimated with the NEC-2 software. There is no reason to assume that SWR’s larger than 2.5 will occur for the windom antenna discussed here.

SWR values for a center fed 1/2 wavelength dipole (blue) and a windom antenna (red). Although larger SWR ranges are found the ATU in the transceiver would be able to handle both antenna’s probably without the need of an external auto-tuner. A 1:1 or 1:2 balun would be good enough for the wind0m antenna.

PI2ASD knuffelmiddag

De beheerders van de Amsterdamse repeaters waren zo vriendelijk een open middag te organiseren op 15-mei, ik hoorde de aankondiging de dag ervoor voorbij komen op pi2nos en besloot er naar toe te gaan. Hieronder een fotoverslag:

Met enkele radio amateurs tussen de antenne’s op het dak van het world fashion center in Amsterdam
Andy pd2emc is een van de repeater beheerders

De UHF repeater bestaat uit een ontvanger en een zender, en deze staat op de bovenste plank, op de onderste plank vind je de ontvanger en de zender van de VHF repeater.

De repeaters die geknuffeld moesten worden

Wat je zelf als amateur nooit zou doen is bij een repeater pure noodzaak, er moeten met bandpass filters zijn om de verschillende frequenties van elkaar te scheiden. Je kunt op verschillende manier bandpass filters maken, op onderstaande foto zie je cavity passband filters. [link] [theory]

VHF bandpass filters
UHF band pass filters

Staande boven op het world fashion center wist ik met 5 Watt de ontvanger in Breda open de krijgen volgens pe4jw die zo vriendelijk was een ontvangstrapportje te doen, het was er wel erg windering. De dipool in het midden van onderstaande panorama foto is een antenne van een lokale omroep in Amsterdam. De ontvangst van pi2nos was daardoor vlak naast deze antenne eigenlijk vrij slecht.

Panorama op het dak

Onderstaand panorama de andere kant uit, dit is richting de kust dacht ik

Ook een panorama op het dak, het stormde er

Totaal verwaaid moesten we toen weer naar binnen, even bijkomen.

Koek en zopie, een gezellig onderling QSO
Omgevingskaartje, station Lelylaan ligt dichtbij

Last update: 10-jun-2016

C4FM op pi1dft

Op PI1DFT draait een C4FM repeater, de frequentie is 438.025 MHz met een split van -7600 kHz, je moet hem instellen in voice wave mode (digital vw). Eerst zat ik op de uitgang te zenden, maar daarna kreeg ik een glasheldere verbinding met Ton PD4DFT en zag ik ook PD4X langskomen. Callsigns worden via het scherm op je transceiver zichtbaar, de audio kwaliteit is prima, beter dan bij D-Star volgens mij, kraakvrij in ieder geval. Wat ik nog niet geprobeerd heb is de group monitoring mode, GM, waarin je calls kunt bijhouden.

c4fm via pi1dft

Via kun je zien wat er rond de digitale modes speelt. C4FM is een nieuwe ontwikkeling en PI1DFT op 438.025 MHz is een test opstelling, een overzicht van C4FM repeaters is hier te vinden.