Call Search
     

New to Ham Radio?
My Profile

Community
Articles
Forums
News
Reviews
Friends Remembered
Strays
Survey Question

Operating
Contesting
DX Cluster Spots
Propagation

Resources
Calendar
Classifieds
Ham Exams
Ham Links
List Archives
News Articles
Product Reviews
QSL Managers

Site Info
eHam Help (FAQ)
Support the site
The eHam Team
Advertising Info
Vision Statement
About eHam.net

   Home   Help Search  
Pages: [1]   Go Down
  Print  
Author Topic: "Best" CW decoder - electronic or computer?  (Read 4645 times)
AD0AE
Member

Posts: 78




Ignore
« on: January 11, 2013, 06:46:26 AM »

Hi All-

OK, before I get too deep into this post, I already know what many of you will say,  YES I AGREE YOUR BRAIN IS THE BEST CW DECODER.

Now onto the real question.  What type of decoder has worked well for people?  Software programs?  Some sort of electronics box (MFJ, some sort of kit?)?  I am just curious.  How well do either of these methods work for decoding?

From my apartment, I have played with some of the software decoders and it has been a little bit disappointing.  I think in part because the SNR needs to be pretty high, which doesn't happen easily when I have an S9 noisefloor!  Is that correct? or am I totally off?

As for the "electronics" of decoding, how is that accomplished?  I mean to say that you must have to be able to distinguish between a dah and a dit, which is usually suppose to be 3 times longer.  That being said, it is usually some sort of discriminator circuit or a digital situation, where you can detect a dit as being "high"?  I am curious about that.

Anyway, I would be curious to hear what people have to say.  Having a read out may be handy, provided it doesn't cost an arm and a leg.

AD0AE
Logged
PA0WV
Member

Posts: 105




Ignore
« Reply #1 on: January 11, 2013, 07:45:57 AM »

I designed and build the VeronCWdecoder, it is published in CQ-QSO and a copy can be found on my website.

I measure the length of each mark and space with a microcontroller. The input of the controller is the output of a tone decoder PLL IC according to the design of IK?OIL a published in QST.  Because it has uncertain bounced on off as output, there is a simple FIR averaging post detection filter in the controller. The output switches the sidetone, so you know what he tries to decode.

The original development was for an amateur with brain defects due to a traffic accident, however looking at  the large number of guys that build it there are a lot handicapped hams, I suppose so.

There are 5 wav sound files on my website and the two loudest  4 and 5 are detected solid, the lower S/N not.

The technique used is to determine from an element ringbuffer with at least 12 elements (mark or spaces)  the average mark duration. After done that the elements are classified in dots and dashes. And then the elements are decoded. I forgot about the details. Look at the article. flow charts are included. When you can't read it because it is in Dutch use a translation website of yahoo, to get a  kind of English you may understand.

It is speed independent.

There is also a fistmeter on my website, that takes the same technique but uses a longer buffer for 16 characters or so, and the fist is calculated as the average modulus of the distortion of the ideal element length. (the average of the buffer)

Wim
Logged
N3QE
Member

Posts: 2137




Ignore
« Reply #2 on: January 11, 2013, 12:43:48 PM »

Not as a "decoder" but as a "mind-blowing tool for an operator", you want CWSkimmer. http://www.dxatlas.com/cwskimmer/
Logged
VA7CPC
Member

Posts: 2372




Ignore
« Reply #3 on: January 13, 2013, 01:59:23 PM »

Hi All-

. . . Now onto the real question.  What type of decoder has worked well for people?  Software programs?  Some sort of electronics box (MFJ, some sort of kit?)?  I am just curious.  How well do either of these methods work for decoding?

From my apartment, I have played with some of the software decoders and it has been a little bit disappointing.  I think in part because the SNR needs to be pretty high, which doesn't happen easily when I have an S9 noisefloor!  Is that correct? or am I totally off?

As for the "electronics" of decoding, how is that accomplished?  I mean to say that you must have to be able to distinguish between a dah and a dit, which is usually suppose to be 3 times longer.  That being said, it is usually some sort of discriminator circuit or a digital situation, where you can detect a dit as being "high"?  I am curious about that.


1.  You are correct -- the decoders only work well when the S/N ratio is fairly high.

2.  I have used CW Get, and recommend it.  The "trial version" is free, and works fine, but nags you occasionally.  I think registration is $30.  I've also used "fldigi", which is free.

3.  CWGet works just as you suspect.  There's a "threshold level" -- signals below that level are considered "off".   The signal _above_ that threshold is "on".   There's an algorithm that tries to match the "on" periods with dits and dah's.  If the incoming signal is "nicely sent" -- standard spacing, 3:1 dah:dit ratio, known speed, it works pretty well.

Filter bandwidth is adjustable, and you can listen to the "post-filter" signal if you want to.

You'll learn more by dowloading the software, and playing with it for an hour or two, than you will from this Forum.  "Free" is cheap enough.  You don't need a soundcard interface, just use the rig's loudspeaker and the computer's microphone.

.               charles
Logged
ZENKI
Member

Posts: 916




Ignore
« Reply #4 on: January 25, 2013, 04:44:51 AM »

Skimmer is incredible. Skimmer while it decodes CW it would not be the best for daily use since the decode is only on 1 line. I just notice when I use Skimmer that it seems to do a half reasonable job.
I dont know if CW Skimmer can use a larger window like CWGET. The thing that I find is that it does not really matter if the accuracy is 100%, if you can see 4 or 5 lines including garbage on the screen
you can work out what is being said if you like using a decoder. Anything with a small display cant give this ESP ability.

The CW decoder on the K3 is very temperamental and  because of the limited screen size you cant see the general QSO details because it scrolls into nowhere.
At our club they have beginners station and it has the K1El K42  decoder/keyboard. The only thing that annoys me  about it is that it uses a fixed pitch something like 600HZ.
I think its the best LCD decoder that I have come across in terms of accuracy.

I dont  need to use them but i like seeing  what the potential of the technology is. I think the program with the most potential would be skimmer if it was turned into  general
CW  QSO package rather than a skimmer  semi only application.

I think noise is the worst enemy of these decoders and like DSP noise reduction on HF, unless we go digital CW decoding will always be a hit and miss affair, thats why its best using your head.

You might email the author of CW SKIMMER and see if a general decoder package is available. Skimmer in a contest is scary the way it decodes so many callsigns.


Logged
K0CBA
Member

Posts: 296




Ignore
« Reply #5 on: February 08, 2013, 06:57:36 PM »

............ "YES I AGREE YOUR BRAIN IS THE BEST CW DECODER."

Wellllll, there you have it.
Logged
N6YWU
Member

Posts: 16




Ignore
« Reply #6 on: February 10, 2013, 04:21:26 PM »

I wrote some Morse Decoder apps for the iPhone, iPad and Mac OS X that are available from Apple's App stores.  Search for apps by HotPaw.  (Sorry, no trial versions.)  I just sent Apple some bug fix updates that hopefully should be in the store next week.  The apps do some statistical signal processing to adapt to varying WPM, "fist" weights, background QRM levels and such, as well as the usual DSP audio filtering and envelope estimation.

In QRM and fading, I can still decode better by ear below 25 WPM.  But my apps still work well past 40+ WPM in QRQ mode.
« Last Edit: February 10, 2013, 04:25:09 PM by N6YWU » Logged
GILGSN
Member

Posts: 201




Ignore
« Reply #7 on: February 11, 2013, 11:32:36 AM »

Quote
Search for apps by HotPaw

I bought that app a while ago  Smiley It works.. I would put my iPod on top of my K1 and uses it as backup while I learned the code..

Here is a suggestion for another app: make a keyer app.. Two big buttons side-by-side, dits and dahs..

Gil.
Logged
DJ1YFK
Member

Posts: 188


WWW

Ignore
« Reply #8 on: February 11, 2013, 11:59:42 AM »

Here is a suggestion for another app: make a keyer app.. Two big buttons side-by-side, dits and dahs..

There's iDitDahText by KB1OOO:

http://iditdahtext.com/iDitDahText.html

73
Fabian
Logged

N6YWU
Member

Posts: 16




Ignore
« Reply #9 on: February 11, 2013, 05:04:24 PM »

Quote
Search for apps by HotPaw

Here is a suggestion for another app: make a keyer app.. Two big buttons side-by-side, dits and dahs..

Done!  In the iOS App Store.  And Free.  Search for MorseKey.  Supports both straight key and 1 Iambic keying mode.

I use it to help test my Morse Decoder app.
Logged
Pages: [1]   Go Up
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.11 | SMF © 2006-2009, Simple Machines LLC Valid XHTML 1.0! Valid CSS!