Radio W4KAZ

Thanks for stopping by the virtual KazShack. Feel free to comment - I often approve them.

Linux Tinkering

The latest version of Ubuntu(v10.4) came up as an option in the updates.  Went ahead and let it install and the upgrade was very smooth.  One moment of head scratching due to a query window regarding options for the upgrade to the “Grub” boot loader, but everything seems to have worked.  The transition was mostly seamless.

The upgrade process did not seem to make any changes to the LAMPP install, which was a good thing.  Shoulda backed up the config files in there, but didn’t – so its good that the install went so smoothly.

Been tinkering with bash scripting, which for the most part is easy to pick up. I’m still a Linux mega-noob, but I can see why programmers like it.  I wish Linux had been at this level 12 years ago.  The folk working on the Ubuntu distro have done a lot of good things to improve the ease of use and ease of install issues – a commendable acheivement. Still having issues with networking to the windows boxes, which is certainly due to the dumb user.[Linux can access the WinXP shares, but not Win7, and no Win flavor can access Linux.]

I can see becoming accustomed to Linux for daily tasks, except for the few pesky windows apps which have no serious alternatives on the Linux platform.  Like a fully functioned contest logging program – at least that I have found.

N4AF has been tinkering with the port of TR to windows, TR4W.  Not sure that’s the first choice for me.  I know that SD is supposed to work in Linux under WINE, but SD may be a bit too bare bones for the things I like in a contest logger.  Time to start experimenting more seriously with Linux.

TurnKey Linux Appliances

This is worth exploring, and here is the direct link.  One day….

Windows 7 GodMode

Leave it to the RedmondGeeks to create a useful tool but leave it undocumented rather than make it easily available.  And a big thanks to NumberOneSon for showing me the trick.

There’s a feature for Windows 7 users called GodMode, which is simply a tool/folder that has a lot of the more useful system administration tasks grouped together in one place.  [As opposed to navigating five screens to get to them.]

All that is required to use the feature is to create a folder then re-name it.  See the link for the details or just goog up the word “godmode” for yourself.  No use re-inventing the wheel here.

[hey!  I didn’t name it…]

Installing Writelog Under Windows 7 UAC

OBSOLETE:  Versions after version 11 are much better.  Disregard if using a version of Writelog after 2011. [w4kaz, 20150101]

ENVIRONMENT: The following applies to an install of Writelog on a Windows 7 64 bit platform with User Account Control enabled.  Probably works for any Windows 7 version.  It may also apply to Vista, but that version of Windows has not found its way into my hands, so experiment at your own speed.  All installs were done under the administrators account, and testing of the application done in a limited user account.  No special permissions were granted to the limited user, nor to any of the directories.

BACKGROUND: The new BlogBox is not used down in the KazShack dungeon, but  is the day-to-day computer.  After a contest, I move the log file up to the main computer and use an install of WriteLog on that box to spew out the Cabrillo, and ADIF backup, and the reports.  For my own nefarious reasons, I chose to set the new BlogBox up with an administrative user, and do all of the day-to-day activities within a limited user account.

WriteLog is still backwards compatible with older versions of windows, and runs well even on systems with limited resources.  That is something that is a useful feature, as it allows a wide  choice of hardware platforms to be pressed into service.  Plus, I just damn well like WriteLog better than anything else.

But since Writelog was designed before the day of user accounts there are some adjustments that need to be made to get it working in Windows 7.

THE INSTALL: One approach used by many is to disable User Account Control[UAC] and run the system as the administrator.  That’s a judgment call.  Diametrically opposed to the goal here…but used with success by many.

Another approach that seemed to work is to install it under an account with  administrator privileges.  But that also falls short of the goal, which is to get it going under a limited user account.  When installed as the administrator, the user accounts were able to run the program, but not able to save the configuration settings.

The approach that seemed to work is to install WriteLog into its own directory[I named it c:\writelog_install_home].

The install went without a hitch.  The real trick is simply to find where everything is right after the install.  The important part is locating the “writelog.ini” configuration file.  For whatever reason the RedmondGeeks in Windows 7 [and maybe Vista] have an environment variable [“appdata”] that is used for hiding certain bits of data under UAC.  The term “hiding” is used deliberately since the directory referenced by the environment variable is indeed hidden.

Finding STUFF:

  • start a command prompt window, and type “set” with no other parameters.  that will display all of the environment variables. The pertinent one is “appdata”
  • In the file explorer, under “organize” –> “Folder and search options” –> “View” -> “Hidden files and folder”  check “Show files, folders and drives”
  • In Windows 7 user info/program data is generally is stored in “C:\users\xxxxxx”, where xxxxxx is whatever your user name might be.
  • Writelog creates a directory in the “c:\users\xxxxxx\documents” directory[i.e., “My Documents” under the logged on user account] for its data files, wav files, contest “ini” files, etc.

The basic install is all pretty easy once you locate the files.  The critical file is the configuration file, “writelog.ini”. In this sort of install there are actually two copies of writelog.ini.  One copy is in C:\windows.  That copy is an abbreviated version, which has what I expect are the bare minimum bits of info required by WriteLog to run.  [NOTE:After installing the program several times it is possible this copy in C:\windows could just be cruft left in place from a previous install.] It seems likely that “writelog.ini” is used to initialize the program.  Probably best to ignore that copy of Writelog.ini, and leave it undisturbed.  You will need to have admin privileges to edit it.

The second copy is stored under the c:\users\xxxxxx\appdata directory in the sub directory \VirtualStore\Windows.  The full path in my install was “c:\users\w4kaz\appdata\VirtualStore\Windows\writelog.ini”.  THIS is the file used to store config settings for the logged on user, and it is here that customizations should be added.

The user copy is created for each user individually and uniquely. That copy is the version that can be customized as required for the particular situation.  A lot of WriteLog users have custom versions of their config file for different situations.  Rather than maintain numerous copies of the file, it might also be appropriate to define a separate user for each situation.  Then all of the copying/management of config files could be avoided by simply logging on to the appropriate user.

That’s enough to get the program functional as far as opening logs, exporting reports and files.  Testing connectivity is to peripherals is more difficult, since the shack is not in the same location and there are no USB dongles yet in use in the KazShack.

The defaulted directories were as follows, after the user had opened the program, and done a “save config”:

Directory=C:\Program Files (x86)\WriteLog\



Wave files and data files are defaulted to the user’s documents folder.  Customize those as needed.

The only real curiosity is that the user writelog.ini install directory was not updated in the final install to reflect the actual install directory.  Its probably best to update that entry to reflect the actual installation directory.

Installed in this manner I have not yet run into anything that required administrator privileges, or that Writelog be “run as administrator”.  But admittedly, I have not yet tested  keying CW, .wav file audio, or rig control.  The BlogBox is not the computer used for contest logging, nor do I use any USB dongles at this point.  In the end, admin privileges may indeed be required for full featured usage that bangs away on the com ports, but in a limited use it is not required.  Best guess is that LPT keying probably is more complicated, if possible at all, and that a Winkeyer would probably work easily via USB.

CAVEAT: After I was satisfied the main program was functional, I went through the Writelog directory and executed each of the utility programs.  The  “tuning indicator and audio snapshot” program received an error window trying to edit the system registry.  That program is for use with RTTY, a mode I am not using, so it is not clear to me that the program has actually failed.  After acknowledging the error, the program appears to load and be ready for action.

UPGRADES: For these tests, Writelog v10.70c was used for the full install.  Upgrades 10.71 through 10.75 were then applied via the administrator, with no problems encountered.  The program ran for both the administrator and the limited user accounts.

And dats da fax, jack…..

New BlogBox

Surfing over to the Dell Outlet last week resulted in a moment of weakness.  There were several bargain buys on Inspiron 537’s, 545’s, and 546’s.  I settled on a nice Inspiron 545s, boasting 4gig of ram, 64 bit install of Windows 7, and a Dual Core Intel CPU.  No extra “flufware” was installed on the system by the folks in Austin.  The 545s is a low profile slimline system, so there not much room to add junk into it later, but at the Outlet price it solved an itch that has been begging to be scratched for a couple of years.

New Hotness.

After almost 10 years, the old Dimension 4300 is really dragging on normal day-to-day usages.  So it is well past time where it should be put to pasture.  It runs Ubuntu Linux well enough, but I have a lot of windows applications I use frequently, so XP has been king.  The 4300 has a 1.6Ghz P4, but it is memory constrained at 512Mb max.  PC133 memory(its that old).  XP runs a lot of things well enough despite the system constraints, but is hard pressed to run the modern antivirus/firewall packages and be able to run a modern internet browser with all of todays’ superfluous and gratuitous  visual content.  Bloated OS, bloated security, bloated browsers, bloated web content.  Itty bitty memory.  Not conducive to a good user experience.

Old-n Busted.

So the Dimension 4300 is going to be a great file server, Linux experimentation platform, and all around backup contest logging box. Just in case the really old and busted CPU in the shack croaks…

It is hardly a shock to find that the New Hotness is pretty damn nice.  Everything is relative.  Windows 7 is taking some mental adjustments.  Ten years of XP have worn big habitual ruts in the gray matter.  But there are a few nice surprises under the hood of Win7. Anybody using/administrating Windows 7 should goog up “GodMode”, a nifty but undocumented control panel quick-list.  The MS-geeks should have made it a documented feature.  Also, the whole system is just blazing fast in comparison to the old.  The on-board graphics are quite sharp, and seem fast enough for daily use.  Probably not a video gamer’s box, but certainly fine for general usage.  Ten years from now – who knows?

One quirk that took a moment to adjust to was finding the directory being used for storing program data.  Many XP applications defaulted to storing program data in the program’s directory.  Under Win7, the application’s program directory is protected.  To get around that, data files used by an older application are shunted into a directory under the current user. A bit of head scratching and cussing later, the location of that directory was found to be defined by the environment variable “appdata”.  After a bit more head scratching it became clear the “appdata” folder was indeed under the user account, but it is a hidden directory.  To get to it directly you can hit ‘Start’ and type %appdata% into the run/search box.  Or you can just enable the file explorer to see hidden files.

Grrrr.  Not sure why it ever made sense to RedmondGeeks to hide the application data……

So far I’ve only run into a couple of minor problems installing software on the 64 bit OS.  I decided to define a separate admin account, and set up user accounts for everybody as users without admin privileges.  General applications can be installed by running the install as an administrator.  This worked well for most applications.  Two glaring exceptions to that generalization: Security software, and Google’s browser.  The security software is understandable – that is best installed by the administrator directly.  But there’s no legit reason Google’s browser should be such a pain in the ass.

And pain in the ass it is.  Running the install of Goggle’s browser as a user with “run as administrator” resulted in no visible installation.  The install program runs, then ends without any messages/ warnings/ errors.  Backing off on that, the install was run directly from the admin account.  In that instance the install succeeds – almost.  After that installation, the browser was available under the administrators log-in, but not to any users.  The install did not offer an option to choose users.  Somewhat less than satisfactory.

So for the moment, Google has been kicked from the New Hotness.  Banished.  Shunned. Deposed. Rejected.

What with IE/Firefox/Opera/Safari all working properly, its not like there are no options.  On the browser front, it has been nice to be able to get back to using Firefox regularly.  Firefox had become really bloated since  I first used it – it is a real performance pig on the old-n-busted 4300.  Having adopted Firefox early on, it was really disappointing to see it become fatter and slower than IE. Over the last 18 months Opera has been the preferred browser.  Opera has been an off-and-on affair over the years, since it has in many cases been the most innovative of the browsers.  In previous incarnations rendering of web pages was not always as reliable as the others, but it has always been the fastest of the group. Since the more recent editions of v9.xx and now v10, it is both fast and consistent.

The Firefox performance issues on a low resource system seem to be a script related problem, although the sites with a ton of images are always slow too.  All of that probably relates to the memory constraints, system paging, thrashing, and the intrusive nature of modern antivirus applications.  Running without the AV software speeds it up in some cases, but the hardware limitation is a bigger problem.  The 4300 box at idle uses almost half the available memory in that system.  The New Hotness zips right along.  Sweet.

The New Hotness can support up to 8gig of ram.  I expect to stuff it to the limit to allow room for tinkering with virtualization(VirtualPC, Sun’s VirtualBox, etc.), and maybe a bit of low end graphics card upgrade, if a decent low profile card is available at a bargain price.

Set for another decade – maybe.

2009 ARRL 160m

Got a chance to work on the matching network for the inverted-L, adding in series capacitance, a choke, and balun. This gave me a good match at about 1840, and the SWR was decent from 1815 up to about 1860. The results seemed to be an improvement in the chances of being heard by most stations. That’s with just 100w. Decent results, given the low power and inefficient antenna.

Got in three really nice runs. The first was disrupted by another station sliding in. That was annoying, because the 10 minute rate was up over 100. After I moved, I settled into a nice 40 minute run that finally pooped out of its own.

There was only a limited amount of time available, so the periods from 2300 local “until” were chosen from both evenings. There seemed a whole lot less activity on Saturday evening. Logged about 3 hours and 130 Q’s Friday night but only 2 hours and 70 Q’s Saturday.

Found KH6ZM working a big pile-up on Friday night. Listened to that for a while, but I only heard him work a couple of east coast stations. The west coast and mid-west were pretty thick, so I didn’t waste time. Tuning up the band I found HI3 calling with no pile-up. Also worked a G3 and a few Caribbean stations.

The K9AY made the difference on several Q’s. The noise wasn’t bad, but There were a couple of relatively weak stations that really peaked on the K9AY but were in the noise floor and not copiable on the inverted-L. Armchair copy on the K9AY. Interesting. Many others were easier copy on the xmit antenna. Also interesting.

The matching network is another great application for a relay box, so I can have a good match at selected parts of the 160m band. With low power, the 3KV panasonic capacitors seemed up to the task, and I saw no signs of the SWR shifting when running, so I presume they are able to handle the 100w level. With several in parallel they seemed to handle the current. Given the bandwidths I am seeing, resonance points at 1815, 1840, 1865, 1900, and 1930 should serve the purpose.


  • Elecraft K2, 100w, into inverted-L and four direction K9AY rx array

The Good:

  • The matching network changes worked.
  • The inverted-L improvements seemed to help the station to be heard, but there’s no easy way to quantify that.
  • Rain shield added to cover feedpoint held up to nasty WX.

The Bad:

  • Not enough time available to operate.
  • Missed the start of the contest.

The Ugly:

  • None!


Call: W4KAZ
Class: Single Op LP
Operating Time (hrs): 5

Total: QSOs = 200 Sections = 50 Countries = 5 Total Score = 22,825

    RX Antenna 160m/80m Band Splitter/Switch

    edited and amended11/02/2009, kaz

    Here is a small project that will work along with the K9AY RX antenna, and solve a minor SO2R problem in the KazShack.

    Currently theK9AY feedline comes into the KazShack directly to the RX input for one of the transceivers. I wanted to have a way to share the antenna between the two radios without connecting the RX inputs of each rig directly to one another(RF isolation), or manually swapping the feedline between radios.

    There are some comments in various places about using the W3LPL RX bandpass filter design to split the bands to multiple destinations. The NCJ article “Distributing Receive Antennas” by K3NA and W2VJN is a very handy and well explained reference.

    This was also desirable here in the KazShack, which sitsin the shadow of the 50KW WPTF on 680kc broadcast transmitter. Rolling the W3LPL filters was done using some T-50-3 toroids and NP0 and high accuracy monolithic ceramic capacitors from the parts bin. The filter is built dead bug style. Each band is on opposite sides of a single piece of copper clad board. Basically, the input is fed to each of the filter banks, and the 160m and 80m bands come out separately, each isolated from the other.

    The coils for each band are identical within the band(i.e., L1=L2=L3), so after winding each I used the MFJ-259 to resonate each coil to the same frequency using the same capacitor. After soldering everything together, a quick test with the antenna analyzer into a dummy load showed each section to show minimum SWR right where I wanted it. No other tuning was required. Almost too easy.

    Left alone at that point, I could feed either radio from either band, but there needs to be a switch of some sort to eliminate the need for swapping coax feeds during the heat of the action. This appeared to be another ideal application of the small signal relays that were also on hand. Using a single DPDT relay, the filter outputs can be switched between the radios with the flip of a switch. A toggle switch mounted on a remote panel is used for convenience . Simple but effective.

    The remote panel is a smallsection cut from 1’5 inch(about 38mm) aluminum angle stock. I pre-drilled pilot holes for future use, and installed the switch for the RX antenna splitter, as well as a control for a planned 40m remote antenna switch. The “panel” is then attached to the inside edge of my home brewed fold out station cabinet. The cabinet is filling up fast – not much room for any more equipment in there.

    Using the switch is going to make swapping the low bands from one radio to the other a snap. Literally as easy as flipping a switch. The band pass filters will also help isolate the radios from one another in the SO2R environment, as well as reducing the broadcast band harmonics.

    There is a bit of signal loss in the filters, but probably not enough to be significant while operating. Hopefully the much lower noise levels on the RX antenna will offset these slight losses. I have not felt a need for an external RX preamplifier before now, but now I am looking at the ARR 1-30. It would be nice to boost the RX signals to parity with the noise on the TX antenna. That might reduce the amount of volume control “riding” needed when looking for the best RX on a contact when toggling between the RX and TX antennas.

    Yet another fun little project. It isn’t as much satisfaction as growing an entire rig from scratch, but it is always fun to put a useful bit of home brewed kit into action.

    PHOTOS (Full set of photos on external page)

    160m/80m Antenna splitter


    PNG image of schematic for W4KAZ's "W3LPL RX band pass filters" built as an antenna splitter and switch.

    PNG image of schematic for W4KAZ’s version of the “W3LPL RX band pass filters” built as an antenna splitter and switch.

    Why Didn’t I Think Of That….

    Jeff, KE9V, has a blog post on an outstandingly good idea for a QSL card display.

    In a nutshell, scan the cards you really like, then load the images into one of the new digital photo frames.


    W4KAZ SO2R Collection – Engineer The Possible -SO2R part 2

    At the bottom of this page is an accumulation of some of the SO2R resource materials I used in developing my own custom SO2R solution. My first SO2R post hashes out the thought process involved in choosing the homebrew methodology for hacking together a workable SO2R set up via home brew components.

    The big issues for customizing an SO2R capability appear to me to be philosophical. It is possible to wear two sets of headphones and manually switch everything, praying you don’t transmit from radio A into radio B. That is a bit TOO minimal, even for me. A minimum SO2R set up for my purpose came down to an audio switch box, a switch for the CW, microphone, and PTT, as well as band pass filters and filter switching(manual and/or automatic).  And something like the Array Solutions SixPak to “keep ’em separated”[cue “The Offspring”].

    It turns out none of those components are out of reach for home-brewing – if you are willing to compromise. The audio and radio input functions can coexist, but they could also easily be separated into two discrete units. One unit to handle headphone audio and the second for CW/MIC/PTT switching and route band data(if used). Likewise, the filters, switching, and band decoder for automatic antenna/filter switching can all be discrete units.

    The crucial decision is whether to use USB(New Hotness) versus serial/parallel(old-n-Busted) interfacing. It didn’t take very long to determine that Old-N-Busted was going to be much easier to twist up in the KazShack. YMMV, and it is a VERY subjective decision. For my own uses it is just simpler to use the parallel interface, even if it requires milking the life from a few old computers running un-supported OS’s. But I suffer no illusions that “simpler” equates to “better”. That’s a subjective call, and will depend upon the circumstances and resources available.

    Building a custom SO2R set-up grew out of my interest in a project by Jim, K4QPL, as well as my interest in filters, both band pass filters and coaxial stub notch filters. Being able to scout a second band will be fun, and it isn’t a great leap from a bit of extra S&P to full two radio operation. I don’t expect to be very proficient at it, but running at low power into mediocre antennas is not terribly productive either. So a full integration of the second radio into the station set-up might be fun.

    All of those considerations lead to researching the topic. Others have done a good job of documenting certain things via the internet, so I’m just aggregating a few of the links I found useful. Some are ideas I have incorporated, like the band pass filters. The filters merit their own separate treatment. Some of the other SO2R links discuss ideas that seem to have merit, but did not apply to my situation. Some are just good reading.

    The first set of links are station specific information, posted by folks describing their own SO2R set-ups. My own customized designs will be referenced first, simply because I can. But just so it is clear, my own design is an amalgamation of the work of others, including K4QPL, KK1L, and others.

    Many thanks to K4QPL. Jim sparked my initial interest in this project via a club program about his own SO2R project, and answered several philosophical questions that led me to my own research and experiments.

    The next set of links point to reference materials or other sites aggregating similar links, or some of the commercial sources. Note: There is a lot of duplication and cross linking. K8ND’s site has a good round up of the commercial sources from here in North America.

    Hardcopy Reference:

    2004 ARRL Handbook, Chapter 22.47, “A Computer Controlled Radio Switch Box”

    Last Amended 9/20/2009, w4kaz

    BPFF – The Guess-timated Scale and actual Guess-timates – Part 6

    Part 6 of the W4KAZ filter project series discusses the actual measured S-meter calibration, and the filter attenuation estimated based on S-meter measurements.

    As I meandered through Part 5 of this group of posts, I needed to find a way to calibrate the S-meter scale on the FT-920 to a 6DB reference. Lacking any real test equipment, this will allow me to do relative tests on the band pass filters to measure the filter attenuation on the harmonics and sub harmonic. So I used the attenuator pad(6,12, and 18db) to measure the delta between each S-unit from S-0 to S-9, “10 over S-9” and “20 over S-9”.

    Big shock(NOT!): The S-meter on the FT-920 is definitely not linear.

    Actual Big shock: The S-meter actually IS linear over part of its scale. I was a bit surprised by that.

    The S-meter on the FT-920 was “measured” by using the attenuator pad, inserting attenuation and noting the S-meter drop. It came out to something like this:

    • S0-S4 – 6db
    • S4-S5 -6db
    • S5-S6 – 6db
    • S6-S7 – 6db
    • S7-S8 – 6db
    • S8-S9 – ~9db
    • S9-10 over 9 – ~12db
    • 10 over to 20 over – **Not measured**.

    It was hard to decide if the drop from 20 over 9 to 10 over 9 was 12db or more, so I didn’t do any testing with any signals that strong. For the sake of an example, when the original signal was reading S4, adding 6db attenuation dropped the reading to S0.

    It seems noteworthy that the spread from S0 to S4 is only 6db. I can often work stations that are down around S0, and almost always work anything higher in decent conditions. I guess to me it emphasized how important just a few db difference might be to making a contact. Maybe a 1db loss throught the filters is more siginficant than it appears. To paraphrase the OM’s, “every db counts…”.

    Armed with that calibration scale, there now is a way to make an educated guess at the amount of attenuation a filter is providing on its harmonics and sub harmonics. By injecting a signal on the harmonic band, comparing the S-meter readings with and without the filter gives an easy way to approximate the filter’s attenuation on that band. It won’t be surprising to find that the accuracy of the measurements will be poor when compared to lab measurements, but it gives a rule of thumb guideline. Better than nothing.

    Amendment, 2009.06.21– Somewhere along the way I misplaced my notes with the measurements made during mid-May. It looks like I won’t have time to re-test the filters for a couple of months, so here are a few notes from an e-mail to N4YDU. The executive summary….

    K4VX filters – worst case is about 30 to 35 db of attenuation, through the 20m filter. The best cases are probably 35 to 40 db attenuation on the second harmonics.

    NVARC Ugly filters – Woooweee! These puppies may have a bit of loss, but they sure do a great job on the harmonics. All bands showed 6 to 12 db better attenuation on the second harmonics than their K4VX counterparts. An S9 signal is dropped to S0, still readable. An S7 signal becomes barely audible at the noise floor of the F-920. The guestimate here is better than 40db attenuation on the second harmonics. Higher order harmonics were disappeared, so no ideas on the attenuation there, except that is “Enough!”.

    Previous in series: Part 5 Guess-timating the filter efficacy

    Start from the beginning at the W4KAZ Band Pass Filter series.