K2 frequencyh stability
Locked
 
|
Hello all,
I notice that on K2 #03547 rev B (with KPA100, KDSP2, KSB2, K160RX and thermistor installed at time of build) the master frequency drifts as a function of temperature. From an ambient temp of 65F to max rig temp when the KPA100 fan kicks in the receiver may drift by 800 Hz. The range is usually more narrow given the typical temp shift experienced. My solution has been to calibrate C22 against WWV such that as the rig warms up, the master frequency set by C22 drifts towards zero-beat with WWV. My question is this: how much drift is typical? Adam, N1KO _______________________________________________ Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
|
|
You don't mention what you mean by "master frequency" nor how you measured
the drift, but I believe you've misunderstood the purpose of the 4MHz oscillator. It is used to program the lookup tables for frequency generation, but during normal operation it's exact frequency does not matter. So, you should set C22 as accurately as you can and immediately run CAL PLL and CAL FIL and then ignore any drift. If you measure the actual performance of the radio, you should see 100 Hz drift maximum. I've never heard anyone say what the drift of the 4MHz oscillator might be, but since it doesn't matter - who cares? Or did I understand you wrong? Dan / WG4S / K2 #2456 <snip> the master frequency drifts as a function of temperature. From an ambient temp of 65F to max rig temp when the KPA100 fan kicks in the receiver may drift by 800 Hz. </snip> _______________________________________________ Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
|
|
In reply to this post by adamkern
Adam,
I assume you are referring to drift of the 4 MHz master oscillator and not to a drift in the PLL reference oscillator. Drift in the 4 MHz reference is of little to no consequence in normal operation of the K2. It is only important at the time CAL PLL and CAL FIL are performed. The K2 does not use the 4 MHz reference to count the frequency on a real-time basis. The values obtained during the CAL PLL and CAL FIL process are retained in EEPROM and those values are used for control and display of the K2 frequency, so the 4 MHz oscillator may drift many kHz and all will still be well with the K2. The 4 MHz oscillator does serve as the clock signal for the microprocessor, but it's accuracy is not important for that use. If you wish to achieve good dial calibration on the K2, warm the K2 up to operating temperature before calibrating the 4 MHz reference oscillator and running CAL PLL and CAL FIL. Instructions for calibrating the K2 dial to an accuracy of 10 to 20 Hz can be found on my website http://w3fpr.qrpradio.com . 73, Don W3FPR > -----Original Message----- > From: [hidden email] > [mailto:[hidden email]]On Behalf Of > [hidden email] > Sent: Sunday, December 24, 2006 8:48 AM > To: [hidden email] > Subject: [Elecraft] K2 frequencyh stability > > > Hello all, > > I notice that on K2 #03547 rev B (with KPA100, KDSP2, KSB2, > K160RX and thermistor installed at time of build) the master > frequency drifts as a function of temperature. From an > ambient temp of 65F to max rig temp when the KPA100 fan > kicks in the receiver may drift by 800 Hz. The range is > usually more narrow given the typical temp shift experienced. > > My solution has been to calibrate C22 against WWV such that > as the rig warms up, the master frequency set by C22 drifts > towards zero-beat with WWV. > > My question is this: how much drift is typical? > > Adam, N1KO > _______________________________________________ > Elecraft mailing list > Post to: [hidden email] > You must be a subscriber to post to the list. > Subscriber Info (Addr. Change, sub, unsub etc.): > http://mailman.qth.net/mailman/listinfo/elecraft > > Help: http://mailman.qth.net/subscribers.htm > Elecraft web page: http://www.elecraft.com > -- > No virus found in this incoming message. > Checked by AVG Free Edition. > Version: 7.5.432 / Virus Database: 268.15.26/597 - Release Date: > 12/21/2006 6:45 PM > No virus found in this outgoing message. Checked by AVG Free Edition. Version: 7.5.432 / Virus Database: 268.15.26/597 - Release Date: 12/21/2006 6:45 PM _______________________________________________ Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
|
|
In reply to this post by adamkern
For a point of reference for N1KO's observed 800 cycle
drift as his K2 warms up with use, I looked up the specs for one of my non-TCXOd radios (IC-765) & overall the radio can drift +/- 350 cycles from 0 to 50C & after 1 minute from switch on, +/- 200 cycles in the first hour. Bung in a 0.5 ppm TCXO & the 0-50C drift goes down to +/- 100 cycles. Again, my 765 does not have one. Unlike some other rigs also in use here, the 765 does not use this reference to derive everything it needs inside (the BFOs are simple crystal oscillators). Both the 765 & my K2 with the all the drift mods were tweaked on an HP5350B & then left to soak. The 5350B is a relatively modern 20 Gc counter & after 30 minutes from applying mains is in +/- 1 ppm range by itself, even better with fitted option reference or external references handy at the time (ultimately a R+S rubidium house reference, model which escapes me now) - essentially down to it's own +/- one least significant digit accuracy spec. Drift observed was therefore essentially just that of the radios. I have yet to get the K2 to indicate the frequency it is actually on (say <200 cycles) & stay that way based on gut feeling from side-by-side operation of the two rigs over a number of years of contesting. And the 765's walkabout is glaringly obvious when compared to a TS-950S without TXCO (but uses that common reference for everything, I believe - can't find manual at the moment). With the K2, there is roughly +/- 500 cycle window that must be considered when near band edges or trying to find some DX station somebody else has spotted. Depending on what one's operations are like, the accuracy of the K2's frequency readout & amount of drift observed may or may not be of concern. It is unrealistic to expect performance in these areas like some radios, as it was clearly not intended originally & with mods, is still probably a bit shy of what might have been possible if was a design goal - though it certainly beats any "backpack" type radio I have had or used. 73, VR2BrettGraham _______________________________________________ Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
|
|
I believe the N1KO data was that he observed a drift of 80 Hz, not 800 Hz.
A drift that large would indicate something seriously wrong in one of the oscillators. The current K2 design is not as good as a TCXO, but it is a lot better than an unstabilized crystal oscillator. It should be possible to reduce the drift to significantly less than 100 Hz over a temperature range of 10-50C with adjustment of resistor RA in the temperature compensation circuit. -John KI6WX ----- Original Message ----- From: "VR2BrettGraham" <[hidden email]> > For a point of reference for N1KO's observed 800 cycle > drift as his K2 warms up with use, I looked up the specs > for one of my non-TCXOd radios (IC-765) & overall the radio > can drift +/- 350 cycles from 0 to 50C & after 1 minute > from switch on, +/- 200 cycles in the first hour. > _______________________________________________ Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
|
|
In reply to this post by adamkern
KI6WX added:
>I believe the N1KO data was that he observed a drift of 80 Hz, not >800 Hz. A drift that large would indicate something seriously wrong >in one of the oscillators. The current K2 design is not as good as >a TCXO, but it is a lot better than an unstabilized crystal >oscillator. It should be possible to reduce the drift to >significantly less than 100 Hz over a temperature range of 10-50C >with adjustment of resistor RA in the temperature compensation circuit. No, N1KO's post clearly said 800 cycles. Some hours later that was then revised to 80. By then I was out backpacking, I think. Since then, I see that we have been told that another manufacturer could not have sold a product to amateurs in the quantity they appear to have with both frequency accuracy & stability of less than +/- 0.5 ppm (-10 to 50C) due to use of single reference for everything inside. I also could not be using another product this very moment that uses only crystals & is spec'd for stability of +/- 5 ppm (-10 to 50C). From cold, 20C ambient, 45 minute drift of ~12 cycles - though could very well be the beacon it was tuned to drifts & my computer probably is not really a good frequency counter to begin with. Invalid points of reference, apparently (pun not intended ;^). Usual disclaimer - one's operations may or may not suffer from some number of dozens of cycles of drift due to temperature, likewise LSD (or next-to-LSD) uncertainty may not be a problem if you do not need to be absolutely certain you are inside the band. Now time for another hike. 73, VR2BrettGraham _______________________________________________ Elecraft mailing list Post to: [hidden email] You must be a subscriber to post to the list. Subscriber Info (Addr. Change, sub, unsub etc.): http://mailman.qth.net/mailman/listinfo/elecraft Help: http://mailman.qth.net/subscribers.htm Elecraft web page: http://www.elecraft.com |
«
Return to Elecraft
|
1 view|%1 views
| Free forum by Nabble | Edit this page |