+++++++++++++++++++++++++++++++++++++++++++ Date: Sat, 10 Mar 2001 07:23:37 -0600 From: Tom Hammond =?iso-8859-1?Q?N=D8SS?= Subject: Re: [Elecraft] K1 int. keyer Steve: >I am however, experiencing some aniexty with the use of the internal >keyer. I find with the use of either A or B iambic mode, the keyer allows >very little forgiveness in terms of dot or dash timing. I have been using >a MFJ keyer for about 5 years and send flawlessly. Has anyone else >experienced this problem? > >ex: I tried to load a CQ and call sign msg into memory and after about 20 >tries I gave up. The Internal keyer on either the K1 or the K2 is pretty unforgiving with regard to loading messages for playback later. However, I've found the keyer itself to be MUCH more forgiving... maybe not quite as forgiving as my CMOS Super Keyer III however. Here's a reply to a similar message wihch was received over a year ago (07/31/99). Maybe this'll help a bit: The K2 message recorder currently 'forces' a 2-word (instead of a 1-word) space when you pause between words (e.g. if you allow more than a 3-baud space between words, even if you're sending continuously). Additionally, it forces a 3-baud character space if you leave more than a 1-baud space between elements of a single character (as it should). NOTE: Recent firmware versions (at least for the K2, and I _think_ for the K1 as well, have FIXED the 2-word-space when loading messages problem. If you do not have a current version of the firmware, drop a note to support at elecraft.com and ask Gary whether your firmware version has the 'fixed' message loading algorithm. The 2-word space is a problem, about which Wayne and I have spoken. Hopefully he can fix this glitch in the next iteration of the firmware. The key to recording smooth CW with the K2 is to stay just a bit AHEAD of the keyer when you are recording WORDS, and to be just a bit AHEAD of the keyer when sending groups of words with spaces inbetween... e.g., don't allow the keyer to come to a complete (2-word space) rest when you're recording words, be ready to start sending the next word AT the time you know it should appear in the message. Also, play with the speed at which you record... some speeds may 'push' you just enough that your timing suffers just to the point that you leave too much spacing between both character elements AND between words. Either of these problems will result in added (unwanted) spaces within words and between letters. DON'T SEND OVER YOUR HEAD WHEN RECORDING. 73 - Tom Hammond N0SS FT ++++++++++++++++++++++++++++++++++++++++++++++++++++++ Date: Wed, 14 Mar 2001 17:17:39 -0500 From: "Brian B. Riley, N1BQ" Subject: RE: [Elecraft] K1 - can't get word spaces into programmed messages I have been having a problem with 'recording' messages on my K-1 (s/n #234). They seem to record accurately, that is, I am not seeing the problem Jake is reporting, but, the response of the keyer is erratically sluggish. Most of the time it responds just as it does inthe 'live on the air' fashion, but occasionally it is 'behind,' that is, i key a dit or a dah on the Bencher and there is a finite delay in the sound response, not very long, but enough to throw off my timing. The pause/delay is reproduced exactly in the 'recording.' I have firmware Rev 1.06 ... it almost seems to me like the CPU s busy doing something else, spending maybe too much time in an interupt sewrvice routine, or something. Has anyone else seen this ??? 72 de brian, n1bq +++++++++++++++++++++++++++++++++++++++++++++++++ Date: Wed, 14 Mar 2001 15:32:13 -0800 From: Wayne Burdick Subject: Re: [Elecraft] K1 - can't get word spaces into programmed messages Hi Brian, Message record "quantizes" your dots and dashes as you enter them, which for some operators can cause a lagging sensation, as we've discovered. We're planning to improve this in the next firmware revision. 73, Wayne N6KR "Brian B. Riley, N1BQ" wrote: > > I have been having a problem with 'recording' messages on my K-1 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++ Date: Mon, 02 Jul 2001 15:14:50 -0700 From: Wayne Burdick Subject: [Elecraft] Re: K1 Memory Bug ? A few K1 users have reported "sluggish" response when recording CW messages. This is due to a slight delay as data is stored in EEPROM. If you have trouble recording, try temporarily setting reducing the keyer speed. We will be improving this in the next K1 firmware release. 73, Wayne N6KR ++++++++++++++++++++++++++++++ Date: Mon, 02 Jul 2001 17:39:49 -0500 From: Tom Hammond =?iso-8859-1?Q?N=D8SS?= Subject: Re: [Elecraft] Re: K1 Memory Bug ? Wayne wrote: >A few K1 users have reported "sluggish" response when recording CW >messages. This is due to a slight delay as data is stored in EEPROM. If >you have trouble recording, try temporarily setting reducing the keyer >speed. In addition to slowing things down (try about 3-5 WPM below the speed you would normally use for loading a message), try to send just a wee bit 'ahead' of the timing. This helps to ensure that an extra baud doesn't sneak in between elements of a character (creating two characters instead of one) or between characters( creating two words instead of one). This same problem was present in the K2 firmware before v2.0x. 73 - Tom Hammond N0SS ++++++++++++++++++++++++++++++++ Date: Mon, 20 Aug 2001 16:16:11 -0700 From: Wayne Burdick Subject: [Elecraft] Re: K2 keyer memories CW message users: To conserve space in nonvolatile memory, the K2's CW message record function "quantizes" incoming CW to the greatest extent possible (short of doing any form of mathematical compression). A single "1" or "0" bit stored in memory represents a dot-length element or dot-length space, respectively. In the case of letter and word spaces, this "interpretation" on the part of the firmware can alter the way stored messages sound when played back. This isn't an issue with code sent in real-time, because the firmware doesn't have to make decisions about letter or word spaces. The operator is free to leave as much or as little space between letters or words as they chose, including small fractions of an element space. A few operators have mentioned that they've found letter spacing to be too long on message playback, which is one possible result of quantization. The way to avoid this problem is to keep your letter spacing close to the standard (3 dot-lengths, or even slightly shorter). In other words, pack the letters closer together. I've observed that during QSOs, many if not most operators send CW with somewhat longer than standard 3-element letter spaces. (It happens most often when one is thinking about what to send next.) So it may be a little unnatural to send code close-packed. One thing that may help is to write out the message on paper before recording it. All versions of the K2 firmware do quantization in exactly the same way, with only very minor (millisecond) timing differences. I just compared 2.01F to earlier versions, and at least for the way I send, I couldn't find any significant difference between them. In all cases, sending the code "close-packed" resulted in code with perfect spacing on playback. We could conceivably allow for more natural-sounding (i.e., "personalized") CW playback, as well as permit a wider range of input spacing, by increasing the quantization level to 4, 8, or even 16 bits per element. Some keyers do this. Unfortunately the K2 doesn't have a large amount of nonvolatile memory available, so this would require cutting way back on the number of CW message buffers and/or their size, which was a tradeoff we felt we shouldn't make. Please let me know whether the suggested message recording technique helps. 73, Wayne N6KR ++++++++++++++++++++++++