Good that you could measure the numbers there, Ed. I don't know the numbers for any of my rigs, but I just did a quickie listening test and the K3 definitely gives me audible band noise conditions at 35. I am new to the K3 and don't really know how to get the optimum from it yet - it probably does as well as the official Elecraft claim: "With the internal keyer running at an indicated 50 WPM, you can hear band activity between dits while using the narrowest filters. " (From the Elecraft website K3 FAQ).
The K3 lets me dial in a specific speed, so it was easy to set it to 35 for tests. Not so the ICOM 756PRO3, which I also tested. I tried it at what I believe is at least 30 WPM, and I could hear band noise on it also. The ICOM uses reed relays. I had an IC-7000 in operation for a while a few weeks ago and satisfied myself that I could hear band conditions between dits with it also, but didn't record any WPM numbers. It was probably in the 30 WPM range. W0QE has made keying time measurements on a number of modern rigs (he was interested in switchover timing for amplifier QSK use) at http://www.bnk.com/w0qe/amplifier_timing.html . Some of his data may be usable to estimate possible QSK speed limits per your calculations, but the receiver recovery time is not included in his report. I hope someone with a better "lab" than mine can put some of these rigs to the test as you request and actually record and publish the resulting audio. That would help popularize full QSK, I think - when you have heard/used the real thing, you want nothing less. Thanks for the work you have done to improve PowerSDR. Jerry W4UK At 01:11 PM 5/14/2008, Ed Russell wrote: >Jerry has brought up an important point. We CW ops believe true QSK >means hearing between dots at high speed. I thought so myself until I >began looking closely at the performance of box-based rigs. > >There is an easy calculation that shows why true QSK is difficult: > >R/T + T/R - ES = Band Sound > >where > >R/T is time it takes to mute receiver and begin transmitting >T/R is time it takes to stop transmitting and fully activate >receiving >ES is element spacing > > >ES is 30ms at 40wpm >ES is 40ms at 30wpm >ES is 60ms at 20wpm > >Hit the key. It takes about 10-15ms to get a waveform started. >Release the key. It takes about 15-20ms minimum for T/R turnaround; >most rigs take 30 or more. That means in the best possible case R/T >plus T/R turnaround takes 25ms. > >I measured the response of my 950SDX: >R/T = 13ms >T/R = 19ms > >That means no Band Sound at 40wpm. I measure 10ms of Band Sound at >30wpm, but subjectively that isn't enough to call useful. > >The 950SDX is an old radio, you might say. But if you look at the >ARRL reviews the numbers are equal to or better than newer rigs. So, >what is the rig that can really do QSK? The K3? Maybe, but it's a DSP >rig too. > >Send me a tape :) > >73 Ed W2RF > >On 14 May 2008 at 11:59, Jerry Flanders wrote: > > > Thanks, Ed. BTW, what you describe as QSK is usually referred to as > > "semi break in", or "Semi QSK" by CW ops. There is a world of difference. > > > > For the guys and gals out there who may not understand the > > distinction, CW ops refer to QSK as the ability to actually hear > > activity on the band between dits at a reasonable speed. Modern > > radios that actually do true QSK can do it at 30-50 WPM. > > > > I look forward to purchasing another 5000a (first was sent back for > > poor CW performance) if and when new software actually allows it to > > perform to this standard. I am pretty sure the 5000a hardware will > > allow this, but the following quote from Frank Brickel's April 26 > > post in the then-current "CW QSK ability" thread eliminated my hope > > that the current PowerSDR could _ever_ provide QSK: "I think this > > line of discussion has run its course of usefulness. Development > > effort is being focused on doing it right, not patching the current > > system, which is basically unfixable in a systematic and > maintainable way. " > > > > If I understand Frank correctly, we will not see what CW ops call QSK > > in PowerSDR/Windows. Ever. > > > > Jerry W4UK > > > > At 07:34 AM 5/14/2008, Ed Russell wrote: > > >Hi Jerry, > > > > > >It's a 5k. I don't think they apply to the 1k without a little > > >coding. > > > > > >73 Ed W2RF > > > > > >On 13 May 2008 at 22:14, Jerry Flanders wrote: > > > > > > > What radio are you using, Ed? > > > > > > > > Jerry W4UK > > > > > > > > At 07:29 PM 5/13/2008, Ed Russell wrote: > > > > >I thought I was finished with PwrSDR QSK a couple of days ago, but > > > > >then I had a few more thoughts... > > > > > > > > > >There is a new version in the SVN branch w2rf\bin\release > > > > > > > > > >My favorite settings are DSP 4096/256 at 96Khz/512. What are yours? > > > > > > > > > >I honestly think that these latest changes take QSK about as far as > > > > >it can go with the current architecture. Hopefully, someone will > > > > >prove me wrong :) > > > > > > > > > >73 Ed W2RF > > _______________________________________________ FlexRadio Systems Mailing List FlexRadio@flex-radio.biz http://mail.flex-radio.biz/mailman/listinfo/flexradio_flex-radio.biz Archives: http://www.mail-archive.com/flexradio%40flex-radio.biz/ Knowledge Base: http://kb.flex-radio.com/ Homepage: http://www.flex-radio.com/