Issue #24: Cool Devices to Learn CW

To add for those not familiar with the "MMX System'
PreppComm DMX-40 - Morse Code Decoding QRP Ham Radio - YouTube
this has a few screen shots of the decoding in action.
Josh also talks of the use as a decoder with a rig already at hand.

I've covered this device elsewhere on the web and also here on the Zed (Better keyboard CW for Technicians | Page 3 | QRZ Forums #32)
where I noted:
"The higher frequency audio feed caught my attention. For many ham rigs, the 1300Hz would be right at the edge of most RX passbands/filters & ditto for many DSP filers."

This high audio frequency input is what sets this decoder system apart from others on the market - IMO. This 'feature' all but eliminates the use of a classic CW filter - unless you have a robust RIT function in your rig. I can see how this would work to aid in processing the CW signal to be decoded. This argues for the 'use widest BW setting' instruction.

As for the claim to decode at or below the noise floor - also credible, as very small, standalone PSK31 systems, like the NUE-PSK modem routinely does this. I upgraded my NUE-PSK modem with the CW software and setting the input to a higher audio frequency does seem to help - much to my surprise.
Much of the settings are quite 'fiddly' - trying to match the rigs RX to the modem, but it does work fairly well. Understanding the RIT function of your rig is key here.
**

The device appears to be well made as seen on this teardown video - PrepComm MMX TearDown - Keyboard CW? - YouTube
**

For the ARS CW purist, the system is a non-starter.

For a 'prepper' that just wants a tool to communicate with friends/family, it would seem to be a ready market.... It is most certainly cheaper than a satphone setup.

As always - YMMV.

 

Eric-
Thanks for dropping in on the discussion.
Any chance you would share a schematic of the MMX?
**

The "AI decoder" meme was covered, quite extensively - here:
The rise of the robot AI CW decoders | QRZ Forums

 

And like you mentioned, when you move to a new signal, you need to press the space bar so it restarts it's algorithm to determine how best to decode the signal. Another point is it's perfect pass band for decoding is 1300hz, to get thru the noise better. That means if your using it on the side of say a G90 set to 700hz CW ; you have to use the RIT on the G90 to get that green dot working at 1300hz.

 

If a Schematic would be published, his market would be stolen in China with clones that are of low quality just like all the other small QRP transmitters.

 

Yes, I don't know why we hams expect this from for-profit companies. Makes no sense to give away the candy store. I was impressed with one Chinese company (Venus) who actually published the schematic of THEIR little QRP radio, which one of our QRZ members helped write up a manual for. That's impressive. A quick glance however and it looks to me like Venus took one of the MTR radios and just reverse engineered it. Dave

 

You looked wrong. Sorry, Dave... To there was no reverse engineering from Venus. One example - the bfo injection frequency is generated by the DDS chip, directly. The Mt Topper has a tunable bfo oscillator.
The input voltage is wider for the SW3B, and the PA amplifier handles SWR much better than Steve’s rig. The tuning rates are far different and so on.
**The software is completely different between the rigs.**

Also, I didn't help write the manual, I wrote the entire manual - based on my earlier manual for the MFJ-920o/HS1A. Hate to sound peckish, but as a professional technical writer, it is a matter of taking pride in my own personal effort.

FWIW, Steve Weber has published the schematic for the Mt Topper, and other rigs, none of which have shown up as a ChiCom clone. Why? **They are unable to copy the software**.

The now defunct DCW1 qrp rig made it to the preproduction prototype stage....but it failed to go into production because they could not successfully steal the software. I sent the only two DCW1 units to make it out of China to @W7UUU , who still has them.
The NUE PSK was cloned, using the open-source software in the clone...and now the Chinese company is gone and the NUE PSK modem is no longer...they are a true victim of Chinese IP theft.
***

Both the SW3B and the Mt Topper are single conversion, multi band rigs using the NE602/SA 612 chip.
Just like the earlier KX1 of Elecraft fame. IMO, the failed DCW1 was likely going to try to replicate the KX1, but failed in the software....

To my earlier point — the folks that sell the MMX would - IMO - face no threat in publishing the radio schematic, as the **software** is the real value of their product. I’ve seen ZERO Chinese clones that have successfully hijacked proprietary software. Only open-source software...a la the (tr)uSDX saga...and the McHF, the NUE PSK and on and on.

As I've noted before, the target market, so-called preppers, usually lack the skills, training and from my conversations there, frankly, the interest to do their own maintenance or repairs. But without the schematic, neither maintenance nor repair is going to be possible for the average non-hambone owner.

This is why, in my postings to these so-called prepper sites, I extoll the virtues of ARS HF kit radios that provide complete technical data. If nothing else, these prepper type would at least have a chance on the repair of a malfunctioning rig.

/rant

 

Thanks for the clarification Don. I think a part of this is the assumption (so often true) that the theft of ideas and inventions is rampant there. My apologies to Venus, Dave

 

Thank you. Accusations of IP theft are a big deal...and yea, the term Chinese Copy is valid, from the many cloned products sold here.

 

Have carefully read and digested the previous comments on both Morserino and MMX----as always, exposure of new products generates multiple examples of personal experience. That said, I will submit my relatively long experience with my MMX20/40/80. I was first licensed in 1953 as WN7iut---, I simply had and have a major preference for CW. I did a lot of loner practice in early years and was very fortunate to have an elmer (SK) who was an ex-navy and RCA commercial telegrapher---he did 50 wpm with ease! He was a very competent teacher! I followed his suggestions and got my license. I have NEVER become a high speed CW hotshot! I can send and follow near 100% at 20-25 wpm. I bought the MMX20/40/80 and can tell you it is very friendly to the shaky 94 year old fist that is me. The MMX DOES provide a genuine incentive to hearing and associating the sound and rhythm that is Morse code. The device has been a joy as I am devoted to QRP and the MMX at around 4 watts is simply fun to report to a contact that is 2000 mile away. I have no experience with the Morserino but the MMX is a much needed boot to retain and magnify the fact that ancient CW is simply unique, irreplaceable and sticks in my mind that allowed me to make 77 critical CW message transfers for ten days in the 1964 Anchorage, Alaska devastating earthquake. I recommend the MMX and CW! 73 from Bob W7iut.

 

Thanks Bob. Great comments, much appreciated. Dave, W7DGJ

 

Well, I would not expect you to use our products - they were not designed for you! We did not build this for people who already know Morse code! If you already have your brain trained, well, then... that is what you should use.

Yes, band noise can make you "not hear." If there is band noise and a fading signal, your ears will stop hearing before the decoder stops hearing. That is simply because the decoder does not hear 99% of the noise, which is eliminated in the decoding process. 99% is probably the wrong number, maybe 95%, dunno. But the decoder separates noise from signal, not perfectly, but better than you can when the S/N is really bad - to the point where the signal has faded out and you can no longer hear it. If you think this can't happen, you don't understand your ears. I know it happens to hams all the time, and they think the signal has completely faded out (because they can't hear it) but it is often still there and if it is, the decoder can decode it, assuming it is not below its hearing lower limit.

Anyway, it is only an interesting point, and it is amazing when it happens to you (even though I invented it, when I see that happen, t blows my mind). It is an infrequent event, not the main event, which is reliable decoding with noise (and you can hear the signal).

 

The "noise doesn't matter" idea certainly is proven out in the work that Joe did in developing our digital modes! Dave, W7DGJ

 

I'm sure some people...those who want to use Morse code, but not necessarily learn it, will like your radio. After having taught the code for years, however, I am skeptical it's a good way to learn it.

HOWEVER...I would be the last person to oppose anything that causes more people to become interested in the code in any shape, form, or fashion! Good luck!

 

Your last line is the right attitude to have on these devices, Rod. At least you're open, still a skeptic, but open for more people to join the CW section of the ham bands. I love the effect that the two big morse classes have had on the hobby (Long Island and CW Acad.) and anticipate that there are just some users who will ALWAYS gravitate to a technology method as opposed to the practice-practice method of our Dads and Elmers. By having all this available, and the discussion here and elsewhere, we can keep waving the CW banner. Dave

 

One thing that might be of interest is a transceiver I designed for digital modes communication that also has a built-in CW encoder and decoder:

http://www.github.com/profdc9/RFBitBanger

It is built with the minimum of specialized parts that would likely be available in a hobbyist junkbox. It is a prepper radio in the sense that if there is a supply chain disruption, this radio relies on the fewest specialized parts that might be difficult to get to repair the radio.

I wrote the CW decoder I use. It's based on a narrowband FIR filter at 667 Hz. The decoder builds a moving-average histogram of mark and space delays and uses these to determine the current cutoffs between dit/dah and interelement / interword delays. It accumulates elements and then does a best fit of the elements to a particular symbol. There is a delay between receiving and decoding to allow the histogram to fill and provide a sufficient sample set to distinguish the elements. A keyboard can also be plugged into the radio if you want to send that way rather than use a straight or iambic key. There are also other features such as built-in digital modes (a new digital-mode that I designed) as well as SSB support in the same manner as the uSDX.

I am giving a presentation on it at QSO Today on the 9th-10th of September.

73,
Dan
KW4TI