Listen to Latest SolderSmoke Podcast

Monday, August 11, 2008

VK6DI Comments on QRSS Bandwidth, Modulation

David, VK6DI, is one of the Knights of QRSS. He sent in this very nice note about bandwidth and modulation methods in QRSS. Thanks David!
-------------------------
Yes, QRSS3 bandwidth is indeed 0.34 Hz. ON7YD has an excellent 'CW bandwidth' reference at

http://www.qsl.net/on7yd/136narro.htm#Bandwidth

Clearly 'hard keying' rather than 'soft keying' of a CW (or FSK) transmitter will increase the transmitted bandwidth. In a practical sense QRSS transmissions key the carrier at such infrequent intervals that an occasional key click every 3 seconds (for dots) or 9 seconds (for dashes) should be of little consequence QRM wise, and especially so when using QRPp. Ideally however it is preferable to control the carrier's rise and fall times during keying. Sometimes just simple R-C filtering is used for carrier wave-shaping, however 'Raised Cosine' rise and fall times such as used for PSK31 are optimal, albeit much more difficult to produce. 'Raised Cosine' shaping also seems to make good 'intuitive sense'. Hence some compromises as to carrier wave shaping are usually the order of the day.

A convenient way to become familiar with QRSS's bandwidth requirements is to run the program "Spectran" (by I2PHD & IK2CZL), and to then observe the preset QRSS parameters. Spectran is available for download at -

http://www.weaksignals.com

As with the program Argo, Spectran also has a series of predetermined settings for all standard QRSS speeds. Those settings have been optimized for best results when receiving QRSS. This is an extremely important factor for most beginners. Whilst it is true that the more sophisticated FFT programs offer tremendous flexibility of user settings, they also carry a proportionally higher risk of operator confusion.

To obtain further insights into QRSS bandwidth requirements try running Spectran and then select a 'Mode' via drop-down menu -



Now observe the "Show Controls" menu -



Note that 11050 / 32768 = 0.34 Hz FFT bin bandwidth. Each bin will take 32768 / 11050 or about 3 seconds to fill, and 3 seconds is the duration of a QRSS3 dot.

Similar calculations can be obtained for other QRSS speeds - QRSS10, QRSS30, etc. Note that both bin size and waterfall scrolling speed are factors that will determine the final visual S/N ratio. For that reason it is best to stick with easy to use programs such as Argo when beginning QRSS activities. Many seasoned QRSS operators use Argo with excellent results, as you can observe on many of the on-line grabbers.

QRSS10 is not considered usable on HF. It is impossible to keep even the most stable transmissions to within a few QRSS10 FFT bins at the receive end of the path due to continually varying ionospheric conditions that will disperse the signal. The final 'visual S/N ratio' is dependent upon capturing as much energy in as few FFT bins as is possible.

QRSS3 on HF over long paths often results in an ambiguous visual display following QSB. For example - a dash may appear as a series of dots. Is the character below an "O" or a "Z"?



Switching to QRSS6 and FSK-CW seems to partially resolve this problem, but not completely. Slower ID's will obviously allow more time to "visually integrate" the signal. That is, to make a 'decision', as to whether the portion just observed on-screen was a dot or a dash. The trade off (as always), is the rate of information transfer. QRSS6 transmissions are best received as QRSS3 in this instance, and look fine on the QRSS3 grabbers. FSK-CW transmission has additional advantages, apart from any perceived improvement in signal readability. FSK-CW is very easy to generate from an existing QRSS keyer, and cheap Red LED's when reverse biased seem to function adequately as Varactor diodes for FSK modulation purposes. A shift of about 5 Hz is all that is required. The best advantage with FSK-CW however, comes from not having to key the transmitter's carrier on and off. That helps with transmitter frequency stability when interstage isolation from the Crystal oscillator may be poor, as is often the case with simple transmitter designs.

There is certainly room for more experimentation with Visual Modes. The sheer variety of approaches to the modulation problem can be quite interesting to watch at times, but in the end the same old constraints remain. Dual freq CW (DFCW) is yet another option, and has the advantage over QRSS Morse that both dots and dashes can be the same length. So ID's are faster than for normal QRSS.

Ref - http://www.w1tag.com/Modes.htm#DFCW_ENTRY

One final factor not always appreciated by newcomers to the 'slow modes' is that the 'visual gain' advantage of QRSS over that of traditional speed Morse does not arise from the use of narrow receiver IF filters as might be expected. The advantage results from the narrow FFT bin size (Resolution) within Argo / Spectran - namely 0.34 Hz at QRSS3 speed. Narrow filters may be useful to exclude strong QRM that otherwise might impact the receiver's AGC, but they make no difference to the visual S/N ratio. (All else being equal.)

Unfortunately I think that it is fairly unlikely that I will see very many EU signals until conditions improve. So far this year I have caught only one EU signal. Conditions have been really poor these past 12 months or so. When the sunspots return I'm sure everybody will see many new EU and US stations. DL6JAN has previously made it down into VK with 5 mW, along with many other stations that were running 50 mW to 200 mW or so, and all with minimal antennas.

Ref - http://www.proehl-elektronik.de/qrss/dxrprt_e.html

Good luck with your QRSS experiments, I hope you will continue to have fun with the mode.

Regards,

David, VK6DI.

VK6DI Web page: http://tinyurl.com/lp2vv

VK6DI Captures: http://tinyurl.com/2lxyy5

No comments:

Post a Comment

Designer: Douglas Bowman | Dimodifikasi oleh Abdul Munir Original Posting Rounders 3 Column