[Elecraft] [KPA500] Rear power switch thread

David Herring dave at ah6td.com
Thu Dec 1 18:58:26 EST 2011 

Olli,

Oh, OK, I see where you are coming from.

The final answer would have to come from Elecraft itself, but I'll hazard a guess that it is probably performance and longevity.

Writing to EEPROM is a relatively slow process and may ( ? ) represent a performance hit within the K3 architecture, the effects of which can be minimized by choosing when to write and when not to write to EEPROM.

EEPROMs have a finite number of writes in their designed lifespan.  You want these things to last a very long time.

So I imagine the firmware developer caches settings in volatile memory (for performance) and writes to EEPROM in well timed batches (for performance and longevity) as a way of addressing these two issues.  

Again, just a guess and it'd sure be nice if Elecraft could weigh in on this and stop the guessing (though I get that this is unlikely the most important thing they have to do right now). ;-)

73 & Aloha,

Dave
AH6TD


On Dec 1, 2011, at 8:27 AM, Oliver Dröse wrote:

> Dave,
> 
> thanks for taking the time trying to explain it. I know what actually 
> happens and why one should first power of the radio before cutting the 
> lines. Nevertheless Elecraft is not the only one using encoders but still 
> they are the only ones with a requirement for that procedure. This has 
> nothing to do with being an SDR.
> 
> Lets stay at one defined example, the VFO. Everybody is using an encoder 
> here, whatever brand it is. Still when I cut power off my Icom (and all 
> other rigs I know of) it always comes back to the last used frequency & 
> mode. If I just cut power from my K3 it never comes back to the same 
> frequency! It usually is one in the same band but not the last used one.
> 
> So how do others do it? I imagine they simply write the current frequency 
> into memory after probably a few milliseconds of no VFO move. Elecraft could 
> do the same (actually they are doing it when powering down the rig).
> 
> So my question is simple: Why don't they do it? What are the reasons behind? 
> Not enough time while working down the operations code? A limitation of the 
> PICs used as the central brain (the other brands use bigger/faster 
> processors)?
> 
> Don't get me wrong, I really like my K3 & KPA500 and have no problems at all 
> with how they work. Nevertheless I am curios to know the reasons. Might be 
> 'cause I'm into (only a little) programming myself and want to understand 
> other guys logics and learn something, too. ;-))
> 
> 73, Olli - DH8BQA
> http://www.dh8bqa.de
> 
> 
> 
> 
> ----- Original Message ----- 
> From: "David Herring" <dave at ah6td.com>
> To: "Oliver Dröse" <droese at necg.de>
> Cc: <elecraft at mailman.qth.net>
> Sent: Thursday, December 01, 2011 12:07 AM
> Subject: Re: [Elecraft] [KPA500] Rear power switch thread
> 
> 
> Olli,
> 
> This is a bit of an oversimplification, but I think it will convey the point 
> good enough to address your very valid question without getting into what 
> could otherwise be a very long answer.
> 
> On most other rigs on the planet, the settings are dictated by the position 
> of the knobs and switches.  Pull the plug on them and whatever computing 
> faculties are present in the rig will simply reboot their software and 
> continue on as normal the next time power is applied.
> 
> The K3 is more software defined than that.  The settings are NOT necessarily 
> dictated by the position of the knobs and switches.  The settings are 
> defined by what's been stored in memory, which is a result of the last 
> change in knob and switch position, not necessarily where the knob or switch 
> is now.
> 
> When you power down the K3 using the front panel power button, this sends an 
> instruction to the computer to save all this information in an orderly and 
> proper manner for next time, and then power off.
> 
> When you pull the plug, the computer in the K3 doesn't get the opportunity 
> to save this stuff out in an orderly fashion.  99% of the time it's no 
> problem.  But pull the plug while the computer happens to be doing a routine 
> save of this information, so that not all gets saved or saved properly (as 
> an example), and you could wind up with a collection of settings that are 
> inconsistent or self-contradicting.  At that point, when you power on next 
> time the computer hasn't got a hope of figuring out what to do.  Your rig 
> has just become hosed.  You'll need to reset everything back to a generic 
> state and start over.
> 
> I don't have first hand knowledge of other SDR rigs, but I imagine Elecraft 
> is not the only one with this "feature."  (I don't call it a problem because 
> it isn't...it's exactly how the architecture is designed to work)
> 
> I hope that helps.
> 
> Dave
> AH6TD
> 
> 
> 
> On Nov 30, 2011, at 12:35 PM, Oliver Dröse wrote:
> 
>> Hi Ron,
>> 
>> the arguments are told over and over again ("it is like a computer, etc.")
>> for the K3 and KPA500. Still I do not see WHY it is possible with every
>> other transceiver and PA on earth to simply shut down power (disconnecting
>> from mains) without first switching it off and without any possible
>> problems? They use the same "computer technique" inside ...
>> 
>> Just curious. ;-))
>> 
>> 73, Olli - DH8BQA
>> 
>> 
>> 
>> ----- Original Message ----- 
>> From: "Ron D'Eau Claire" <ron at cobi.biz>
>> To: <elecraft at mailman.qth.net>
>> Sent: Wednesday, November 30, 2011 4:42 PM
>> Subject: Re: [Elecraft] [KPA500] Rear power switch thread
>> 
>> 
>>> Just to be clear - the Elecraft equipment should be turned off using the
>>> front panel switch *before* it is disconnected from the mains supply so
>>> the
>>> logic system can do a proper shut-down. You can get away with "pulling 
>>> the
>>> plug" much of the time, but sooner or later you'll catch the logic in the
>>> process do doing something critical - writing data to memory, etc. - and
>>> will end up with scrambled data. That can be corrected by reloading the
>>> firmware, but it's a hassle easily avoided by powering down with the 
>>> front
>>> panel switch first.
>>> 
>>> Ron AC7AC
>>> 
>>> 
>>> ______________________________________________________________
>>> Elecraft mailing list
>>> Home: http://mailman.qth.net/mailman/listinfo/elecraft
>>> Help: http://mailman.qth.net/mmfaq.htm
>>> Post: mailto:Elecraft at mailman.qth.net
>>> 
>>> This list hosted by: http://www.qsl.net
>>> Please help support this email list: http://www.qsl.net/donate.html
>>> 
>>> 
>>> -----
>>> eMail ist virenfrei.
>>> Von AVG uberpruft - www.avg.de
>>> Version: 10.0.1411 / Virendatenbank: 2092/4046 - Ausgabedatum: 29.11.2011
>>> 
>> 
>> ______________________________________________________________
>> Elecraft mailing list
>> Home: http://mailman.qth.net/mailman/listinfo/elecraft
>> Help: http://mailman.qth.net/mmfaq.htm
>> Post: mailto:Elecraft at mailman.qth.net
>> 
>> This list hosted by: http://www.qsl.net
>> Please help support this email list: http://www.qsl.net/donate.html
> 
> 
> 
> -----
> eMail ist virenfrei.
> Von AVG überprüft - www.avg.de
> Version: 10.0.1411 / Virendatenbank: 2102/4048 - Ausgabedatum: 30.11.2011
> 
> 
> ______________________________________________________________
> Elecraft mailing list
> Home: http://mailman.qth.net/mailman/listinfo/elecraft
> Help: http://mailman.qth.net/mmfaq.htm
> Post: mailto:Elecraft at mailman.qth.net
> 
> This list hosted by: http://www.qsl.net
> Please help support this email list: http://www.qsl.net/donate.html

More information about the Elecraft mailing list