++++++++++++++++++ See also K2 VFO Drifts During Transmit ++++++++++++++++++ Reply-To: "Don Wilhelm" From: "Don Wilhelm" To: "Ron ZL1TW" , "Elecraft" Subject: Re: [Elecraft] Freq creep. Date: Mon, 27 May 2002 20:56:23 -0400 Ron and all, The latest firmware contains a fix for the 'creeping VFO', so for those who do not want to install Builder's Alert 16, upgrading the firmware is an alternative. (this latest version is the one that supports the KPA100 - version 2.02e). 73, Don Wilhelm - Wake Forest, NC W3FPR home page: http://www.qsl.net/w3fpr/ QRP-L # 485 K2 SN 0020 mailto: w3fpr at arrl.net ++++++++++++++++++ To: elecraft at mailman.qth.net Date: Wed, 20 Feb 2002 18:07:55 +0000 (GMT) From: ehramm at dk3uz.shnet.org (Edmund H. Ramm) Subject: [Elecraft] K2 frequency runaway A few minutes ago as of typing this, it happened again: While I was transmitting (not in split mode) the displayed (and transmitted) frequency slowly crawled upwards. It stopped crawling as soon as I stopped, and got moving again when I resumed keying. Curiously enough the frequency stayed put when I used the paddle instead of the manual key! This reminded me of a piece someone else posted here about the encoder sitting right on a transition, which appears to have been the case here. After I moved the VFO knob a tiny fraction the frequency crawl stopped regardless of how much I cared to hammer the trusty old Junker M.T. And I couldn't reproduce the phenomenon by turning the VFO knob back by what felt to be the same amount. 73, Eddi (K2 #2018) ._._. -- email: dk3uz AT darc DOT de - AMPRNET: dk3uz at db0hht.ampr.org Linux/m68k, the best U**x ever to hit an Atari! ++++++++++++++++++++ Date: Wed, 20 Feb 2002 11:40:59 -0800 From: Wayne Burdick Organization: Elecraft To: "Edmund H. Ramm" Cc: elecraft at mailman.qth.net Subject: [Elecraft] K2 CW TX frequency change: possible fix "Edmund H. Ramm" wrote: > After I moved the VFO knob a tiny fraction the frequency crawl stopped > regardless of how much I cared to hammer the trusty old Junker M.T. And > I couldn't reproduce the phenomenon by turning the VFO knob back by what > felt to be the same amount. Hi Eddi, Your posting was very helpful. I have been working very hard trying to duplicate this problem on my lab K2--no luck so far. One of my theories has been that our off/on cycling of the encoder's ground pin could be compromising its built-in hysteresis. We turn the encoder off/on in receive mode to reduce RX mode current drain. But if you were right on an edge (which is very difficult to find), you could see the encoder tick in one direction or the other due to the slight change in supply voltage or load that might occur during transmit. Your observation about moving the VFO slightly stopping the up-tick lends credence to this theory. And this may also be why I can't duplicate the problem: my lab K2 is running experimental firmware that leaves the encoder turned on at all times! I would like to ask you and others who have noticed this problem to try a very simple modification to the K2 front panel board. This mod will keep the encoder powered on at all times, increasing current drain slightly in receive mode. If encoder on/off cycling is causing the problem, this will cure it. Here are the modification instructions: 1. Turn the K2 off. Also set the internal battery switch to OFF if applicable. 2. Remove the top cover (disconnect all option cables). 3. Remove the screws from the front panel module, and unplug it from the RF board. (NOTE: There's no need to remove the front panel knobs or separate the board from the sheet metal.) 4. On the back of the front panel PC board, locate pin 14 of U3 (TPIC6B595 or similar). (This pin goes to pin 4 of the shaft encoder, Z1.) 5. Install a short jumper between pins 11 and 14 of U3. This will permanently enable the shaft encoder. Pin 14 is an open-drain output, so it can be shorted to ground without causing any damage to the IC. 6. Plug the front panel module back in and secure it with four screws. 7. Power up the K2 and make sure the encoder and other front panel functions are working normally. 8. Re-install any options, then the K2 top cover. Please let me know right away if you see the problem continue to occur despite this modification. If the mod *does* fix it, we'll make it permanent, either in hardware or firmware. One possibility is to only power-cycle the encoder in OPT=BATT mode, and only after a long period of VFO inactivity. But the saved current drain would be minimal. Thanks, Wayne N6KR ++++++++++++++++++ Reply-To: "Don Wilhelm" From: "Don Wilhelm" To: , "Edmund H. Ramm" Subject: Re: [Elecraft] K2 frequency runaway Date: Wed, 20 Feb 2002 21:27:54 -0500 Eddi, The one thing that comes to my mind is that you may not have the knob snugged against the felt washer quite enough. The clue I am concentrating on is that it happened when using the straight key, but not with the paddles. Try 'drumming' on the desk with your fingers a bit and see if you can get the frequency to change with that kind of mechanical vibration. If you do, move the knob a wee bit closer to the panel so the felt washer gives a bit more frictional drag. 73, Don Wilhelm - Wake Forest, NC W3FPR home page: http://www.qsl.net/w3fpr/ QRP-L # 485 K2 SN 0020 mailto: w3fpr at arrl.net ----- Original Message ----- > A few minutes ago as of typing this, it happened again: While I was > transmitting (not in split mode) the displayed (and transmitted) frequency > slowly crawled upwards. It stopped crawling as soon as I stopped, and got > moving again when I resumed keying. Curiously enough the frequency stayed > put when I used the paddle instead of the manual key! > ++++++++++++++++++ Date: Wed, 20 Feb 2002 17:02:55 -1000 From: John Buck To: Don Wilhelm Cc: elecraft at mailman.qth.net, "Edmund H. Ramm" Subject: Re: [Elecraft] K2 frequency runaway I posted to this thread once already, but believe that the friction issue is being under-rated. The symptoms being described correspond exactly with the problems I was having when I has a too light or no friction setup on the knob. Because the knob is slightly unbalanced, it will depend on the exact setting of the knob position. It can be highly intermittent or start all by itself even without vibration. It shows up because the shaft friction is actually very low. This is good. So if you have this problem, and are using very light friction, set in some more friction and see if the problem disappears. I did not believe this was the cause until I tried it. You cannot see the knob moving unless the lighting is right and you look very carefully. My knob still has very light friction but not zero. And I have not had the problem since. Aloha, John KH7T ++++++++++++++++++++ Reply-To: From: "Ron D' Eau Claire" To: Subject: RE: [Elecraft] K2 frequency runaway Date: Wed, 20 Feb 2002 20:31:35 -0800 John, KH7T, wrote: > The symptoms being described correspond exactly with the problems I was > having when I has a too light or no friction setup on the knob. You may be right in some cases, John. But I can tell you that it does not cover every case of this symptom. I used to have the knob rather loose and experienced that 'creep'. It's now solid and I still get the problem at very odd intervals. When it strikes I have managed to make the K2 move almost 100 kHz with no visible movement of the knob. At the 10 Hz setting I use that would be quite visible if the knob were actually moving. Knob creep may be the cause for some ops, but it is not the only cause of this "drift". Hopefully Wayne has the solution in the simple mod he came up with. Only time will tell, since it happens to me about once in a month or two of regular operating. Ron AC7AC K2 # 1289 ++++++++++++++++++ Date: Thu, 21 Feb 2002 11:33:25 -0800 From: Wayne Burdick Organization: Elecraft To: Elecraft Subject: [Elecraft] special request to K2 owners A few postings recently have mentioned an increase or decrease in operating frequency that occurs very infrequently during CW keying of the K2. In some cases this may be due to inadequate friction between the VFO knob and the felt washer, but in other cases it could be a side effect of the way we cycle power on/off to the optical shaft encoder. We'll need help from K2 owners to determine if this encoder power cycling really is the culprit. So I'd like to ask a favor of everyone who may have seen the frequency shift occur: please make the simple modification I posted yesterday. We're especially anxious to have those who get on the air frequently make the mod, since the problem crops up so rarely. The more trouble-free on-air time we observe with the mod in place, the more confident we'll be that we've found the problem. It should take only 5 or 10 minutes to make the mod. You don't even have to remove the front panel controls; just unplug the module. One short jumper wire is soldered on the back of the front panel PC board. If you make the mod but still see some sort of frequency shift symptom, please send me (not the list) the following information: - the exact operating frequency (down to 10 Hz if possible) - the VFO tuning RATE selection (and whether the VFO was locked) - the nature of the observed shift (up/down, amount, regular or sporadic) - whether or not you had SPLIT or RIT turned on - whether you were using internal or external keying - CW menu settings (STL, STP, T-R, IAB, INP) - whether moving the VFO slightly restores normal operation Our goal, as ever, is to continually improve the K2's reliability. Thanks in advance for your help. 73, Wayne N6KR +++++++++++++++++ To: elecraft at mailman.qth.net Date: Thu, 21 Feb 2002 14:30:08 -0700 From: Gary L Surrency Subject: [Elecraft] K2 knob 'feel' I echo the other comment made about standing the K2 with its knobs facing straight up, and allowing the weight of the VFO knob determine how far it is on the encoder shaft, This usually results in about the right amount of felt washer friction to keep the knob from free wheeling. Just be sure the knob is not so loose that it can creep or rotate on its own. Then we can separate the loose knob problems from the real VFO freq. changing that some are seeing under certain conditions. Gary Surrency AB7MY Date: Mon, 25 Feb 2002 14:37:43 -0800 From: Wayne Burdick Organization: Elecraft To: elecraft at mailman.qth.net Subject: [Elecraft] K2 Frequency shift on transmit: FIXED K2 BUILDER ALERT: Correcting K2 Frequency Shift on Transmit APPLICABILITY This builder alert applies to ALL K2s, and all revisions of K2 firmware. It affects ALL modes of operation (CW, SSB, data), with or without SPLIT enabled. BACKGROUND Some K2 operators have reported seeing the operating frequency shift up or down one VFO step at a time (10, 50, or 1000 Hz) as the rig is keyed. It may move continuously in one direction as long as keying continues. This symptom has only been reported in CW mode, probably because CW key closures happen much more frequently than mic PTT closures. However, such a shift can also occur in SSB modes when PTT is being used. We have now identified the cause of the problem: on/off cycling of power to the optical shaft encoder. If the encoder is positioned such that it is on the border between two counts, this power cycling can cause the encoder to lose its hysteresis (analog position memory). When this happens in conjunction with keying of the DOT or PTT line, the small amount of noise injected into the encoder's logic circuitry can cause the encoder to switch states. This make the firmware think the VFO knob has been moved. The reason why the condition is so rarely seen is that it is extremely difficult to position the encoder on one of these count boundaries. Fortunately there is an easy fix for this in hardware. We recommend that all K2 owners make the simple modification described below. A single short jumper must be added to the Front Panel board. MODIFICATION INSTRUCTIONS NOTE: These instructions will be posted on our web site soon, and we'll be adding a photo showing exactly where to install the short jumper. If you're anxious to make the mod right away, just make sure you have correctly identified U3 pins 11 and 14. 1. Turn the K2 off. Also set the internal battery switch to OFF if applicable. 2. Remove the top cover (disconnect all option cables). 3. Remove the four screws from the front panel module, as well as the two screws holding the control board to the front panel board. Then unplug the front panel module from the RF board. (There's no need to remove the front panel knobs or separate the board from the sheet metal.) 4. On the back of the front panel PC board, locate pin 14 of U3 (the 20-pin I.C. near the shaft encoder, Z1). 5. Install a short jumper between pins 11 and 14 of U3. (This will permanently enable the shaft encoder. Pin 14 is an open-drain output, so it can be shorted to ground without causing any damage to the IC.) 6. Plug the front panel module back in and secure it with four screws. 7. Secure the front panel board to the control board with two screws. 8. Power up the K2 and make sure the encoder and other front panel functions are working normally. 9. Re-install any options and the top cover. TECHNICAL NOTE Power to the optical encoder was cycled on/off to save a few milliamps of receive-mode current drain between encoder reads. However, encoder reads are done at a fast rate anyway, so the current saved was only about 10-15 mA (varying somewhat from unit to unit). ++++++++++++++++++++ Date: Mon, 29 Apr 2002 13:36:25 -0700 From: Eric Swartz WA6HHQ - Elecraft Organization: Elecraft To: rondec at easystreet.com Cc: Glenn Maclean , elecraft at mailman.qth.net Subject: Re: [Elecraft] VFO Drifts during Transmit See: http://www.elecraft.com/Apps/Alert16.html on our builder's resource page. 73, Eric =========== Ron D' Eau Claire wrote: > > >...I was having a CW QSO on 20 meters this afternoon. My VFO started drifting > > un commanded during transmit. > > > > Glenn WA7SPY > > Hi, Glen: > > Wayne posted a fix via e-mail on the reflector some weeks ago. It is a minor > change that disables the power switching circuit so the encoder stays active > at all times. > > The fix might be on the Elecraft web page (www.elecraft.com) under > "Builder's Resources". ++++++++++++++++++ ++++++++++++++++++