DX Code of Conduit at dx-code.org

Wednesday, August 25, 2010

A SDR full blown layout V 1.1 - alternative and choices

From FT-9000 group - Re: Flex-500A

You are right Daniel, at all.

The 1K-FA it's really touchy when driven without ALC. I have some experience driving it also with the PROIII and the FT-950, with the respective ALC well connected and it works more easily for sure.

I could confirm and say as mine yours words and opinions from my side.

I gained some compromise setups band by band after all my intense usage with my poor dipoles in a restricted town shack with a condo roof for aerials. It's not easy and during contest, at night, ... those happenings that you could easily leaved me so disgruntled that I was ready just for the bed.

As I reached a good compromise, with the driver left a 50% and the main power output control done via the TX gain with heavy steps, I never tried, for example, a software utility called DDutil by K5FR. It could be, if you have any interest or just curiosity about, interesting to have a look here:

http://k5fr.com/ddutilwiki/index.php?title=Linear_Amplifier_Management

This is not to convince you to get back on FLEX, but just to say that yours problem isn't only yours. As somebody other have a managed to find out a solution.

Actually all the major players are using a SDR core well surrounded by more traditional layout for ours toys. All this is more perfectible over time and with some investment. Also I hope that the market would permit to gain for us the ability to use ours best traditional setups with all the needed and seamless useful new features. The time will tell all to us.

By now I would still continue to use my beloved FLEX, brought by me, and a guXX brought in by another operator for the multi single club station IQ1RY. Those vacuum tubes are steady as a rock and not touchy at all when driven even at full legal power for limited test purposes. At home I will stay more low, as US legal power will light neighborhoods lights and would illegal here. RTTY at 300/400W it's mine limit, 500W is the Italian hamradio operator limit for any class, not mobile.
You know Daniel it's kind alike the food: each one, name one, have it's way to be better cooked around the seasons of the year not only by it's nature and kind. Flavor are different each time.

Next time you will operate a FLEX I hope to be by myself operating a FT-9000 and my 1K-FA. It would be a real pleasure IMHO. I just have to find where are all mi mics, anyway one the right will suffice ... If I just could remember where I stored all of those. ;-)

73 de iw1ayd Salvo

on response of ---

Re: [FTdx-9000] Re: Flex-500A

Thank you so much for your input. I read all that you wrote with great interest.
I am happy to hear that the SDR platform has worked well for you. One of my
greatest problems with the Flex-5000A set up is the lack of an ALC circuit. I
have an SPE Expert linear which is VERY "touchy" in terms of input wattage. When
there is an ALC circuit, it works like a charm, but without it, I find the
linear constantly shutting down with one fault or another. I have tried several
approaches. I may try one of the other Flex radios again in the future, perhaps
the even the next generation. They are always updating and upgrading. I am sure
that there will be other Flex platforms which will be a bit less clumsy. Their
panoramic display is 2nd to none. Hopefully Yaesu and Icom will come up with a
panoramic display like the Flex.


Well, thanks so much for sharing your experience!


Daniel -- W3DCB

From TS590 yahoo group - Re: Extra Filter Slots?

Hello to all.

I would associate my opinion with the same from Peter - HB9PJT.
Clear points Peter. Also the last one it's clever: " We will see".

One more point that I would give for discussion is that, citing the FLEX software and hardware architecture, a big enhancement in the filter shape occurs with bigger buffers. I mean those buffers in the DSP processor, as far as I know. This seems quite independent from the DSP algorithms used. I just leave the standard one offered by default.
What I mean is that when I choose inside POWERSDR(tm) smallest buffers. As to decrease latency inside those buffers, let's say a 512 bytes buffer size. The filter shape degrades by any mean. RTTY signal nearby 100/300 Hz from a 230Hz filter leaks visibly on MMTTY waterfall. There is also the associated audible signal in the speaker. With bigger buffers, let's say 2048 byte, there is a perfect shaped effect with the same IF filter. No nearby signals at all.
In RTTY or any other digital modes that is not CW I don't think that those IF buffers need to be so tight.
And there are several other aspects to benefit the filter shape and the whole IF circuitry response, as peter wrote. I could associate also MMTTY BF processing after all.
I have no doubt that as it's done by me manually, with the simplest cause and effects process, it could be done flawless and better by the firmware. This last would be made, i hope, around a drill down processing of all the choices involved in. But there is also marketing.

Nonetheless, as this will be firmware controlled inside a 590 there is the possibility that something could go wrong. We have a lot of transceivers that looks really interesting but that are at least poor in real world performance. Not to say that some of those are good as door stop from perspective to perspective. By any means a single button action, a mode change to an AGC set, have several things to set and reset, code & data flow. I hope that there was no the wrong approach during the development of this so hoped and nice looking RTX. "We will see".

The famous and over speak K3 have had and could have several changes on it's firmware. But the habit about is that everything is well done at any time and for any season. More a mental habit and perception than a real issue as for everyone other builder on the market.

Lastly, SDR are the answer for today needs. But there are strengths ans weakness as usual. Different words than in the analog world but same final problems. Not speaking about the void of software and human interfaces. What we see is that all the commercial builders are, for the greater parts of the work done, approaching the soft way to SDR. But this is another story.

73 de iw1ayd Salvo

Tuesday, August 24, 2010

Just a great app SDR Console by HB9DRV Simon


Not just for the graphical integrated GUI idea but fo a lot more thing inside out.

A SDR full blown layout

Re: Flex-5000A

Hi all.
Not to be silly but just to trow in my 2 cents. Well, it's a bigger 44 cents coin at the end.
Sorry for the rather long message, I hope it's clearly readable due by my poor English.

Unfortunately I don't own any flavor of the FT-9000. It's quite out my target, well several of my targets. Nonetheless I am here and with other groups/mail to read the several interesting thing that come out every day.

My experience on FLEX is really positive. Mine activities are centered around RTTY, both Contesting and DX chasing. I like also to have real QSO kbd to kbd or simply air checks with this or any other digital mode. No ROS.
I have to admit that before of this radio I get almost acquainted with non SDR radios in half the time. The SDR took a lot of time just because there are so many switch, buttons and lists of values that I have had my time to get it running as smooth as I like. My latest try was with a FT-897 and is 300 Hz filter, just after reselling an FT-450. All went working in some hours, perhaps spreading the efforts over days. But nothing compared to the pure joy of the FLEX.

I am using a dedicated and protected PC with W7 32bit(!) for the POWERSDR(tm)radio application with a FLEX-3000. The software environment is completed by N1MM with MMTY or alternatively with MMSSTV, WSPR, WSJT. Last but not least Tmate help me a lot with the VFO fiddling.
I have had hard times running the HRD-DM780-LOGGER trio, virtual serial port emulation is quite a nightmare. Now it is solved with mini deluxe, a stretched application that emulate HRD and permit to DM780 and the LOGGER to run flawlessly. The HRD Logger is still the main logger, with HRD utils it help me a lot sorting all electronic QSL activities.
As a result s 26", 16:9, monitor is fully crowded mainly by the N1MM bits and pieces together with POWERSDR(tm).
This screen is my radio and my radio is on this screen.
Filtering and bandpass commanding is written inside CAT commands within the digital set of macro for the N1MM digital window: no radio touching is needed, all the operations are driven by the N1MM windows in focus. Quite the same thing went done before the advent of the FLEX, for filters and some other controls, with the PROIII/FT-950. Well, not all together, some in the N1MM and some in DM780, as needed by the different activities.

I have done several RTTY contests from my small pistol shack, helped by a 1K-FA, the Tmate USB device, a trackball instead of a mouse and a multiband V shaped rigid trapped dipole by Create.

I have also done, as with some new release the FLEX went more stable, some contesting from the biggest than mine shack of ours club, IQ1RY: astonishing!!!
This is a quite good DX site, not so far away and not so exotic, from here. The aerial and some other piece of the facility could be seen on and from ww.qrz.com

The FLEX 3000 is at least a good radio. I have have in view the band as I need. All is commanded via the N1MM macro buttons set. All is done almost via a mouse, seldom with the keyboard when not in a chit chat QSO. I just need to find a free spot, if there is one of course, to start a RUN session. Save the current QRG, look at the cluster when permitted or to the band in the panadapter to check out who is there to have a QSO more. He would be a multiplier, luckily, or simply another RUNner just to improve the rate.
By this time not SDR software is almost integrated by CAT and virtual and physical serial ports. 1K-FA frequency/bands control was done routing the virtual serial port CAT also inside a physical port, by the way.

During DX chasing it's almost the same. After getting the TX QRG of the DX stations it's just a matter of finding in a "view and click" a free spot UP to start calling. Having also the idea on whats going on even on my TX QRG and not only on the DX receiving QRG. This is similar to the double set of filters applied to the RX and TX QRGs VFO for YAESU. A different beast of what ICOM does.

Tight filters are really tight with bigger software buffers. I couldn't cry too much about my old and trusted couple of PROIII with theirs one shot TPF. Nor with the nice filters of the FT-950.

One thing that I have done is to retain the FSK style operations also in AFSK. The radio will be set dead on the RTTY tones, inside the RTTY filter, not clicking-clacking the mouse to get on the RTTY mouse inside the band pass. This is as FSK fixed radio must be used not AFSK ones. There is also a TX filter for RTTY, tight enough not to spread or spit everywhere unwanted tones. BTW I am running lower tones, mark is 1445 Hz. More pleasant to listen hour after hour, IMO.

Another thing that I was doing is to enforce the whole setup environment as to use N1MM any activity, QSO & Contesting. I have learned, over time, a good knowledge of the N1MM environment. Forcing myself to use it anyway since the start of this SDR adventure. I have not to say again that I was maxed out to do this by the first unfruitful bunch of approaches with virtual serial port and the SDR. Things that someone others seems not to have had. The change from DX and QSO activities to the Contest activity "desk", of that time, I simple switch over the related macro sets. Then I engrave it in the database reload information. Each one of these is cooked as needed and will come alone up in the deprecated case there is any application shutdown. Those sets are for the input window. Digital modes macro with filters and general RTTY sentences are the same whiting the digital mode windows.

Now I am quite a "satisfied customer", but still asking myself if there is the the real need to change anything out of the main antenna, something of a lack here in a dense urban area, to improve the general setup and, of course, the performances. I still look at mine other radios with pleasure and joy, PROIII, FT-950 and TS-870. I already get rid of several more new ICs and some older FTs (MP). I would also consider a Ft-9000 and is “smaller” brother the FT-5000.
I will need something to fill in the portable operation hole. But no way since now, there isn't the right piece of hardware out now. Owning FT-897/857, IC-703, and TS-480, for portable activities, doesn't help too much, also with 500 Hz or 300 Hz filters (remember RTTY!). I see to buy a portable PC to have smooth operations with the 3000 another nightmare that I wouldn't like to start by now. May be the 1500 should be a good starting point, but ... by now I have other priorities. I am in view also for the HPSDR project, the mercury standalone board. Actually it lacks the software facilities side, no CAT, no serials, in one one a good experiment but not to much a radio for the everyday use.

May be there are several aspects on witch I get so acquainted and that I even forgot to write all down here. ALC, RTTY TX filter and so on. There could be possible bunch of issues on some of these forgotten things but those last are quite working/solved here. At least I was lucky to find out the mix and max of related adjustments, for now.

73 de iw1ayd Salvo

Wednesday, August 4, 2010

All started here OpenHDSR


Well, the software side discussion is very interesting.
I should vote, speaking from the usability side, that I would prefer a
two folded model, Client - Server. Both to run on the same PC, as Piece
of Computer, or not, as the X-11. Yes sometimes they comeback.

It is a long message and I am not writing in my native language, thanks
in advance for yours patience.

What I mean thinking since now to a fully digital integrated model
containing the GUI down to the SDR HW, is that we need some strong
pivot points to be able to make the grow of ours child much more a model
build up than a blob up of him.

A premise. For some time to come we would need a strong interfacing
between ours currents variegated data application, even for the CW and
for phone. But after that first time, some "years" to come, we would
have such as N1MM application and all the things around that models well
integrated inside a single GUI. An not only it, let me envision the
whole plethora of applications that would have big advantages using the
SDR worlds to come: digital modes from RTTY, PSK, up to packet modes
framing to others. Embedded data/voice modes could have then their
times. The intrinsics of SDR could be a big advantage for all those
things, well resident inside ours small "audio" channels or even with
smallest bandwidth. BTW, also some bigger bandwidth modes now to come
could have some advantage by ours work, as it was before.

The future, right into and between this future and ours now-today we
should have an interim periods well extended or not. Today needs are
quite well depicted inside POWERSDR(TM). A strong user GUI, able to
spawn several homework, and not only, projects. A great CAT setup, able
to match ours current programs, logger and so on. But a monolithic
design, like POWERSDR(TM) is, isn't the beast approach to grow up to the
time and to more sophisticate hardware as we could have. The
expensivenes of this HW is another story.

A set of simple and high repetitive operationa could be run on
specialized hardware, this is part the radio in SDR. Perseus and other
commercial or now HW clearly demonstrated this. It could be done with a
FPGA, but it also could need some more power ... a 2004 QEX article
featuring a digital radio board with some integrated "mini"CPU could be
useful. Here will be the data manipulation and it will receive commands
and controls for this lowest level functions, not too much complex
commands and controls I think

A set of not so repetitive functions, all equals to the each one before
and after, more math intensive and user commands prone could reside in
a so called Server. This would have a specialized high speed low
arbitration headache/overhead "bus" with the SDR HW module. It also have
to drive, somewhat flawlessly, another "bus" to single or multiple GUI.
For X-11 this was the Client and here you have had the control for the
mouse. A well known set of casual movements and repetitive functions
mixed together. The order of the function wasn't casual but ordered as
by "sets of functions" instated by places, movements and so on geometry
recognition. Even at some meaningless looping states of theClient the
mouse was still responsive and working as to give the right "command and
control" sensation to the user.
The GUI or Server must have interfaces to the Server, a seam;es
communication interface for data, commands and controls. Multiple GUI
Server application both residing on the same user machine could
interact, limited by hardware capacity with the Client. The master data
processing, RF to numeric - SDR HW, would be still unaffected by this as
it resides on the "radio" and is working to the, let me say, 100 fold
the GUI ability to use this data. It's a matter of fact, nobody could
drive at the right speed. push it up and use it as it is needed.

The whole picture ... let me try to make a picture in words ... not so
easy but feasible with yours patience.

- Ours waves - the RF environment -

> HW BOX - damned good for RF, Sherwood and other teach here, and fast
for data I/O translation or transliteration - HBOX - What inside? A big
enough FPGA < <> fast lane, just 2 users, minimal control - on off start
stop check OK or NG - timing well knows - small set of commands and
controls, almost machine generated, on separated lanes (or much faster) >

> Client BOX - CBOX - damned good to communicate and pre/post process
data forth and back - CBOX dedicated computing platform - some sort of
general purpose CPU helped by more specialized data mangling machine
(integrated on the same die why not), user command tags translation in
atomic commands to himself and to the HW SDR < <> seamless standard lane, multiuser, slow data flow, delay
and congestion treatments, big commands and controls almost user
generated - timing well known >

Server BOX - SBOX general purpose computing box/platform - display, user
commands tags (just another set of outer world translation table, serial
interfaced to outer worlds (loggers, CAT, Pkt Clu., PA?, unatomized
commands and controls ) {great effort to translate those to CBOX "and to
HBOX", strict control over it, internal translation routine} interfaces
to plugin digital modules, any current or viable/future digital modes {a
different set of more atomized command tags, each module must be SBOX
aware, SBOX libraries - well know} - responsiveness timing well known

Well I don't clearly stated, by the language, by expressions or by
space, all my view, but that's all for now, I made enought noise. Since
we have to think a picture like this, call it Jeff or Jim, it would be
better to make it step by step from the start. Even making all
monolithically, but having well in mind that there will be a place or
two to folds and cuts. Just to be ready for the "next generation"
software we need and that even this will not be the last. There is a lot
more to write or state but I think this should be enough, i.e. not to
speak only of perils and frills about compilers. I don't hate compilers
but any of it don't gain enough point to make a match, it seems.

Sorry for the waste of bandwidth , but my hope is to be somewhat useful
and not only to clarify it by myself, HI. If I am wrong somebody would
correct me, as for example with X-11, I am working from memory. If it
mean nothing to you forget about.


73 de iw1ayd Salvo