9/9/14

TenTec Rebel & the QRP Skunkwerx Universal Shield

Huntsville Alabama has a great hamfest every August.  In addition to a nice air conditioned indoor boneyard and loads of commercial vendors, Craig, NM4T, puts together a great slate of QRP-related forums.  Not only that, but he also hosts a gathering of QRPers on Saturday night at nearby Monte Sano State Park.  The festivities include an FB BBQ, along with an event or two.  This year, one of the events was the "Huntsville Rebelthon".

Craig, along with Glen, KW5GP, and some others, have formed the QRP SkunkWerx - a design group that, so far as I know, has concentrated primarily on expanding the TenTec Rebel Open Source QRP Transceiver.  The SkunkWerx have come up with a "Universal Shield" for the Rebel, and have pushed the little rig beyond its CW limits by coaxing the rig to work on JT65.

The Rebelthon was limited to a handful of Rebel-owning participants.  Glen provided the Universal Shields, some miscellaneous parts to slightly modify the Rebel, a Nokia 5110 display, the wiring needed to hook it all up, the revised sketch, as well as assistance and general moral support for several of us to start down the road of personalizing our Rebels.

SkunkWerx Universal Shield and Bandswitch Mod
The SkunkWerx mod involved replacing the SPDT on-off switch with a DPDT-center-off switch to enable hardware bandswitching.  The switch sets the band jumper open or short, which tells the chipKIT UNO32 which band the rig is on.  In my case, the bandswitch is wired so that up is 40 meters and down is 20 meters - center is off.

The band changing requires that the DDS tune the appropriate frequency to mix with the IF so the rig will tune on the chosen band.  Also, the appropriate bandpass filter must be selected.  As the rig comes, the bandpass filters are selected with four jumpers.  The SkunkWerx Universal Shield uses two latching relays to select the proper filter.  The RebelAllianceMod does much the same, but with a couple differences.  First, they do not use latching relays - when powered, their relays will constantly consume power - not a lot, but for a QRP rig that often runs on batteries, every little bit helps.  The other difference is that they use a long press of the FUNCTION button on the Rebel to change bands. This has the advantage of not having to re-start the rig when changing bands.  Perhaps someone will incorporate that switching method into the SkunkWerx sketch.

That's one of the beauties of Open Source - many different people contributing different solutions to the same problem - and having these solutions available, others can combine them in ways that appeal to them.

SkunkWerx Nokia 5110 Display
Back to the Universal Shield, the SkunkWerx have included a connector for a display, and there are options for a variety of display types.  Their sketch includes code for the Nokia 5110, as well as a "standard" LCD...which is included in the original TenTec sketch.  Glen provided a Nokia 5110 display, and the information that their sketch provides is shown to the right.  Namely, TX frequency, RX frequency, RX Bandwidth, DDS Step Size, and frequency Band.

At the Rebelthon, Glen made sure everyone had success in modifying their on/off switch into a bandswitch, and also had all the jumpers wired in correctly.  He then uploaded the SkunkWerx sketch into each rig, and tested to make sure all was well.  In my case, the bandswitch was a little flaky, so Glen pulled another pre-wired switch out of his stash and we had my rig flying right in just a couple minutes.

So, with a big push from Glen, it was time to get down to playing more with the Rebel.  As the Rebel comes, it's an OK, but not great, QRP rig.  Certainly adding a display adds a lot to the enjoyment of using the rig.

One of the first things I noticed when I got my Rebel was how bad the RX sounded.  Listening on earbuds was extremely uncomfortable.  There is also a fair bit of noise just from turning the main tuning encoder - or pressing the switches or turning the RIT knob for that matter.  I know that Michael, KD4SGN, had some similar experiences with his Rebel and has made some modifications that have helped.  I need to get with him to learn about what he's done, but in the interim, I decided to connect a speaker to the rig, and the radio is now usable - not the equal of the Drake R-4, but at least now I can stand to listen to it.  In fact, I'm listening to a bit of DX on 40m as I type this.

With the sound of the RX tamed down a bit, I felt better about messing with the sketch.  For starters, I find the Rebel's Wide and Medium bandwidths pretty useless, so I changed the default bandwidth from Wide to Narrow...pretty easy!  Another thing I didn't like about the Rebel is that the finest DDS step is 100 Hz.  I knew 10 Hz steps would make for a more analog-VFO-like experience, and allow me to tune the pitch in just so.  Well, it is easy enough to change the tuning steps, but the Rebel's encoder is so coarse, that the tuning rate is much too slow when using 10 Hz steps.  So, back to 100 Hz for now.

Now, on to the SkunkWerx sketch.  The first thing I didn't really care for with Glen's display was the lack of decimal points and thousands separators on the frequency.  A frequency readout of 14021035 was just too much for me...I could much better deal with 14.021.035 and would prefer 14.021.0 or .04.  Well, one thing I learned was that putting those decimals in isn't the easiest thing to do!  I tried several different methods, but never came up with anything satisfactory.  So, I changed the display to read in KHz, with no thousands separators and no fractional KHz.  Again, not ideal, but something I can re-visit in the future.

Next, the Nokia display is pretty small, and the library Glen used results in six lines, so the type is small as well.  Part of this is because of all the information displayed - Glen uses five of the six available lines. Looking at all the info displayed, really a lot of it is redundant.  I didn't need the Band - that info is available from the frequency readout.  Likewise, the Step Size and Bandwidth are both indicated on the radio's front panel via LEDs, so that could go.  And the main advantage to me for showing both TX and RX frequency is to see how far from zero the RIT is tuned - the RIT doesn't have an "off" position, so it's easy to bump the knob and be a few Hz...or few hundred Hz off frequency.  So, some kind of RIT display would be nice.

AA4GA 5110 Display - as of early September 2014
For my changes, I initially removed everything but the TX frequency from the display.  I then changed to the largest numeric font in the 5110 library, which is effectively three rows high....resulting in a nice, easily visible frequency readout.  Below that, I added a numeric indication of RIT tuning +/- in Hz.  To make it more readable, I skipped a line below the frequency.  That still left a line free at the bottom of the display, and I decided that would be a great spot for an S-meter bar graph.  So, I added that using asterisks for the graphing elements.  I would prefer something more "solid", but that can wait until another day.  I am also displaying the raw signal strength provided by the Rebel so I can get a feel for how high the numbers actually go.  Someday I may use this for "calibrating" the S-meter, but for now, it's nice to just have a relative display of received signal strength.

I do want to mention that my fault-finding of Glen's display isn't a negative reflection on Glen - he actually designed this as a springboard for others to mold into their vision of what the rig should be.  In fact, on Glen's personal Rebel, he's using an OLED display...an option worth considering!  In fact, Glen is not only a capable programmer, he has recently authored a book on using the Arduino for Ham Radio that the ARRL has just published.  I bought the book and have ordered sensors for some of the included projects.  Glen's motivation is definitely educating people on using the Arduino (and by association, the Rebel).

So, if you've got a Rebel, consider using the SkunkWerx Universal Shield and sketch as a start for your own Rebel excursions.  The complete SkunkWerx sketch and documentation of how to build the Universal Shield are available at the Rebel Yahoo Group's file section.

My main purpose in getting the Rebel was to use it for learning about rig control (primarily VFO and display) for use in some future home-brew projects.  I think the rig will come in handy for that and will provide me with a playground to test different ideas - I've already got several more that I want to try when I get around to it.

Thank you to Glen, Craig, and all the SkunkWerx....and to TenTec for the fun toy!

11/24/13

MacLoggerDX

With the acquisition of a Mac Mini to replace a very tired Windows XP laptop, came the need to convert my daily logging system from Logger32 to an OSX solution.

I have used Logger32 for a long time, it has a decent feature set, and is highly configurable.  Relatively complicated to set up, but once that is done, it works pretty flawlessly.  There is good documentation by a dedicated team of folks, as well as an active Yahoo group that includes regular participation by the program's originator Bob Furser, K4CY.  As a DXer, it does everything I need a logging program to do...not as fancy as some of the programs out there, but the features it has are solid and useful.  All business, no fluff.

We converted to Macs at work about four years ago, so I'm fairly familiar with them, and have had my own Mac in my sights for a long time, so I'd done some research regarding logging software already, and knew there weren't nearly as many players as in the Windows world.  I also knew based on personal experience and anecdotal reports that I could probably keep running Logger32 in a virtual machine on the Mac, but did not want to do it that way if I could help it.  Keeping that as a fall-back solution, I wanted to run a native OSX application, so I began seriously looking for a logging program.

The two that seemed most interesting were RUMLog by Tom Lindner, DL2RUM and MacLoggerDX (MLDX) by Don Agro,  VE3VRW.   Both applications have user forums online in which the developers participate, but the MLDX forum seems to have the larger and more active user base of the two.  There are things about RUMLog that remind me of Logger32, and as freeware, the price is right, so I decided to give it a try.  It's not a bad program at all, but I had had many recommendations to try MLDX, so I decided I would.

MLDX is $95, but has a trial scheme in that you can use the non-crippled program as much as you'd like prior to buying, with the only limitation being that the program times out after 15 minutes.  Fair enough, I can give it a try before buying.  The first thing I noticed is that 15 minutes is a very short time to spend with a complicated program when you don't know anything about how it works!  Most software packages I've run across will do something similar, but give you a two-week to 30-day trial period.  I can see advantages to both schemes I suppose, but when actively evaluating, the 15 minute sessions are a PITA, but I sat down with the program, manual, and online forum to give it my best shot.

After importing my 5000+ QSO QRP-only log from the past three years, the first thing I noticed is that the QSO entry area is very busy on MLDX.  There are fields that, IMO, just aren't necessary to display for average QSOs.  For example, in addition to "normal", expected fields such as First Name, State, Frequency, Mode, Reports and Notes,  there are questionable fields such as Last Name, Street, City, State, Email, IOTA, SOTA, 10/10, etc.  MLDX pretty much assumes the user will be using an online callsign lookup service such as QRZ.com.

Also, the formatting of some items is fairly questionable as well.  Particularly, the Time fields are a full timestamp displayed as YYYY-MM-DD HH:MM:SS.  That is very busy to look at and needlessly detailed for most purposes.  But I figured I could live with it - IMO though, it just detracts from the appearance of the log - a log whose main advantage is it's appearance.

Another problematic item is the Notes or Comments field.  It is automatically populated with a bunch of needless "stuff" when doing a lookup on a station.  For instance, the following was entered with no prompting from me in the last QSO I entered into MLDX:

Class: E, efdate: 2005-02-01, expdate: 2015-03-17, GMT:-5, qrzloc:user, Uploaded to ClubLog 2013-11-22 02:16

Whew!  Now, there may be a flag in the preferences that allows me to turn this off, but I never did see it.  But really, why would I want to put in my log when the other station's license is going to expire?  Crazy.  Just because some information is available doesn't mean you need to log it!

But I kept looking at the program, and while it did do some fairly stupid things, it did some other things very nicely indeed.  There is a Stacking register that allows you to keep up with calls/frequencies and easily tune back to them which is nice.  There are a lot of map options that are fun.  Like most hams I think, I like maps.  While the Grayline map was OK (although it should be a vector image like most loggers use, rather than a bitmapped image), especially with the ability to plot spotted DX stations, the map I was most enamored with was the Google Map - the station worked is plotted on Google Maps, which can be zoomed in or out to get an idea of the other station's locale.  When I was a new ham over 35 years ago, I remember looking up all my QSOs on a map just to see where French Lick, Indiana was or whatever - great fun, and this automated it.  Nice.  In fact, this mapping feature is probably the single feature that made me decide to send Don Agro my $95.  I figured all the other weaknesses I'd encountered to date were tolerable, and/or were simply a matter of me not setting the preferences correctly yet.

So, I paid my 95 bucks and started using the program.  Don Agro was very quick to issue me a registration number and I was off and running.  The first thing I noticed after the program had been running for an hour or so was that my display of DX stations in the Spots window was not showing much in the way of useful spots.  Almost all the spots were countries and band/countries I'd worked before.  So, I started digging into the preferences, the manual, and researching old posts on the Yahoo group, but I could not figure out a way for the Spots window to display only new countries or new band-countries.  Instead, it was showing every DL, G, I, etc. that had been spotted, which cluttered up the Spots window tremendously.  While the spots were color coded, each of several columns had to be reviewed for each spot to determine whether or not it was a spot of interest.  So, I posted a question about this to the Yahoo group and was quickly answered by Don Agro, who basically said that the filtering I was looking for was unavailable.  I then asked that a feature request be considered to allow such filtering, as without that, the Spots window is very difficult to use.  At that point, I was quickly admonished by Don Agro that it was an inappropriate use of the Yahoo group to post feature requests, and that any posted in such a manner would be ignored.  And, fair enough, that information is included on the group's "About" page and I'd missed it.  So, I apologized and sent a feature request directly to Dogpark Software and Don Agro.

And that is when the shit hit the fan.

Via a rapidly-deteriorating exchange of email between me and Don Agro, I was told that I was the only one who had ever made such a request, that my request was "nonsense", and at one point "Feature bloat is the single worst thing a Mac application can be accused of".  Which is really pretty funny if you look at the program on a whole and see all the useless things that it does do in an attempt to be pretty vs. functional.  At one point in the exchange Don Agro, to his credit, offered to refund my money, which I thought was more than fair, and a bit surprising actually.  This was after he had said he wouldn't add the filtering capability that seemed so basic to me.  So, I decided to see if I could successfully install a virtual machine and get Logger32 running acceptably on my Mac, and if I could, I would probably take him up on his offer.

A few hours later, Don Agro sent me an email stating "Version 5.50 Beta 64 has worked and confirmed filters added but I would still be more than happy to issue you a refund. I don’t need your $95.00".  Wow, what a class act!  That pretty much made my decision for me - I would put up with the limitations in RUMLog (which really is a good program and not all that limited) before using MLDX!  Hell, I would resort to a pen and ink log before using MLDX!  (OK, that's not much of a stretch - except for contests, I still keep a paper log - I'm a fountain pen user/collector!)

I had been warned that Don Agro was a bit of a curmudgeon.  Bob Furser is too, and so am I.  But I think it's more than curmudgeonliness in Don Agro's case - I think he may have some issues that could use professional attention.  He certainly does not know how to address customers.  To his credit, Don Agro did rapidly refund my purchase price when I accepted that offer.

So, I downloaded Parallels and installed Logger32 on the Mac, and am working along very nicely indeed, with all communications between the radio, keyer, and software working exactly as they should.  Parallels works well with OSX's new full-screen mode, and I've got my smaller monitor off to the side with full-screen Logger32 and the big monitor used for Mac stuff and everything is working great!  No, it's not as pretty as MLDX, and doesn't have quite the same feature set, but it does what it should do, and does it well.  And the developer isn't an asshole to me to boot!  Life is good.

MLDX may be a good choice for you - I will warn you to test it fully before buying.  Oh, and one more thing:  I wanted to leave my install of MLDX active on my machine until I finished this review, before deleting it forever.  So, when I started writing the review, I cranked up MLDX to review how it did some things, and after 10 minutes, I got the familiar "this trial will stop after 15 minutes" pop-up.  Wow, Don Agro has the ability to rescind MLDX registration even after you've purchased the program.  So, if you really piss him off, be warned, he could disable your registration at any time.


A little addendum:

OK, so I saw nothing in the manual or discussed in the Yahoo group about uninstalling MLDX, so I decided to post a final message to the Yahoo Group to make sure I was getting everything removed properly.  Well, I already realized this was a private group and thought all posts were moderated by Don Agro himself, but figured I'd ask anyway.  Here is the automated response I got:

We are unable to deliver the message from <ME>
to <macloggerdx@yahoogroups.com>.

You are not allowed to send email to this group. There are two possible
reasons for this:

1. This group may only accept postings from moderators.
2. The moderator of this group may have removed your ability to post to
   this group.

To contact the group moderator, send mail to macloggerdx-owner@yahoogroups.com

For further assistance, please visit http://help.yahoo.com/l/us/yahoo/groups/original/members/forms/general.html


---------- Forwarded message ----------
From: Lee Hiers <ME>
To: macloggerdx@yahoogroups.com
Cc: 
Date: Sun, 24 Nov 2013 12:50:30 -0500
Subject: UnInstall
I did not see any instructions in the manual referencing removal of MLDX, nor in any of the messages here.

Other than deleting the application file, I see there are four dogparksoftware entries in the Library/Preferences to remove, as well as the log and some associated files in a directory in my Documents directory.

Is there anything else to remove/do to delete MLDX from my Mac?


Having previously posted a few messages successfully to the group when evaluating and installing the program, I can only surmise that number two is the reason my message was not accepted.  Nice.

8/15/13

CW Open 2013 QRP Team

This is a quickie about the CW Ops CW Open contest that is coming up on 8/31.  Actually, it is three contests in one - there are three four-hour time slots in a single 24-hour day - each of the three periods is scored individually, as well as an aggregated score.  The CW Open has team contesting, and last year we put together a team of QRP-only operators.  It was so much fun (and we didn't come in last place!) that we're doing it again.

Szabi, VA2FDT, was also working on getting together a QRP team, so we've joined up our efforts.

Here are the two teams we've put together:

QRP RockcrushersSession 1Session 2Session 3Total
VA2FDT2,3322,8083,2338,373
WB3AAL0000
WB5BKL3,2242,3852,5208,129
N6GA5,7962,9765,30414,076
W7ZRC23,56009,43432,994
N8XX5,1203,8284,24813,196
N0EF6,6961,1225768,394



QRP BulldozersSession 1Session 2Session 3Total
AA4GA1,024001,024
AF9W1,56009282,488
KU7Y1,0563425,9647,362
N7RR????
WA8HSB05,07005,070
K2UFT10,4765,57690016,952
WC7S8,3644,3503,54016,254
Please make an effort to give us all a few QSOs in the CW Open this weekend!


3/14/13

AA4GA - March 2013


The local radio club (Athens Radio Club) is having a "show off your shack" show and tell at the next meeting and asked everyone to send in pics of their shack...so, this is mine as of the middle of March 2013:


2/19/13

ARRL International DX Contest 2013

Like last year, the condx seemed to go down a bit on Sunday compared to Saturday. This year seemed a little worse than last - hard to say though. One thing that was down this year was BIC time - only about 18.5 hours compared to 24 hours last year. I did pick up a couple extra mults compared to last year, but QSOs and points were down slightly. But this was with 25% less operating time, so either condx were actually better this year, or the station was better. I'm voting for the second. The weekend before the contest I put up a full-sized 40m ground plane, with the feedpoint up about 12' or so. This antenna consistently beat the 80m doublet into Europe on 40m, so it got a good bit of use. Also, on Saturday, during the contest, I put up a trapped EFHW for 10/15/30.

This antenna has generated a number of questions, so in case anyone else is curious, here's a little more info. The Cushcraft R-series of electrical half-wave antennas has always gotten mixed reviews, but I think overall they are a good idea. But I didn't want to shell out the money to try one in case they really were dummy loads. So, I figured why not build what I wanted? Since the R-series are effectively half-wave antennas, they have a high input impedance - this is corrected for with the large base-mounted matchbox that comes with the antenna. A similar arrangement exists for the Par End-Fedz 10/20/40. It is a half-wave on 40 that has a broad-band matching transformer used at the base; in the Par literature, they even mention that this matcher may be used for single-band antennas by simply replacing the wire with one appropriately sized for the band of interest. The 10/20/40 resonances all stem from the fact that these bands are harmonically related. My idea was to build an R-3(-ish) antenna with a trapped half-wave element to allow operation on 10/15/30m. 10 and 15 because those are the bands I need to pay attention to for my 5BDXCC, and because my antennas from 20m on down seem to work OK as is...and, I was anxious to get back to the contest and didn't want to bother with tuning the new antenna's low band up to 20m.

In March 2012, QST published an article by W6NBC titled "Better Coax Traps", wherein the author presented an alternative method of coax trap construction that was free-form, rather than using PVC or some other form on which to wind the traps. This looked like an easy method, so that's what I did, using the VE6YP trap calculator. I used RG-174 to build coax traps, and was aiming for a frequency just below the targeted bands. Not having a way to measure the trap resonance, I just put an antenna together and checked SWR - generally the SWR was OK after trimming the 10m section to lowest SWR, but could be better. I will likely trim the antenna closer to resonance on all bands, and also am planning on trying some different traps. Before doing too much with the traps though, I'd like to buy or build a dipmeter so I'll have a reasonable way to check the traps' resonant frequency. Oh, I also fed the antenna through a Par End Fedz 10/20/40 matchbox, as I had one laying around, not in use. There are a number of plans for a similar matchbox on the internet...Google is your friend!

I hung the antenna up as a vertical, with the feedpoint up about 10 feet or so. How did it work? At first, on 10 meters, when the band was just opening, the 80m doublet sounded better than the wire vertical. But, within a half hour or so, the band settled down, and the the vertical was actually better to Europe about 95% of the time, and the advantage was usually at least two s-units - at times, the advantage was 4 or 5 s-units. Performance like that makes the antenna a keeper - I do hope to do a little more tweaking on the antenna before next fall.

For the contest itself, I operated about six hours less, and got a couple more mults, and otherwise, my score was down only slightly compared to last year. The bands seemed a little worse, but with the new antennas, it was pretty hard to tell.

Here's the band breakdown:

Band QSOs Mults
80:   30   26
40:   85   47
20:   91   51
15:   87   44
10:   58   35

Tot: 351  203 Total Score = 213,150

Down a little less than 5%, with only 75% of the operating time - I'll take it!

KX3 at 5 watts, 80m doublet at 45', 40m Ground plane fed at 12', and 10/15/30m trapped EFHW, hung vertically.

2/7/13

MagicBox

The Four State QRP Group (representing MO, AR, NE, and OK) sells a number of neat QRP kits including the K8IQY MagicBox, a T/R controller that lets you connect a QRP transmitter, receiver, antenna, and key and gives back full- or semi-QSK, RX muting, and sidetone. It provides an easy way to interface all sorts of little transmitter and receiver projects so that they can be easily used on the air. The Four State website goes into greater detail.

I got my kit from someone on QRP-L who decided not to build it, and found that the kit was well documented and went together easily and quickly. I tested it and made a QSO using a Tuna Tin 2 I'd built up Manhattan style. The board is designed such that no enclosure is needed and uses PCB-mount jacks for all connections, so it was easy to try without an enclosure. Which was great, because for a lot of us, finishing the project by mounting in an enclosure is the hardest part. It certainly is for me, and I had a bare MagicBox for months.

 I left it at work, thinking I'd be able to get one of our shop guys to help me with it, but that never happened. I finally conned NA4SO into helping me drill the holes, but we did run into a problem with the drilling template. It was a little off, although I thought I'd confirmed it was OK (measure twice, cut once!). Not too bad though, and the easiest solution was to pull the BNC jacks off the board and use panel-mount jacks.

 I set mine up with a power LED, a semi/full QSK switch, spot switch, and tune pushbutton. The only problem with any of this was with the spot switch - the early kits were supplied with a "bleed-through" resistor that was too large a value - I paralleled a 100 Ohm trimpot with this resistor and now the spot function works as it should...with the added benefit of being adjustable...this could even be routed to a pot for front-panel control if desired.

Right now, I'm using my old Drake R4 for an RX, but eventually hope to build something up from scratch.  All-in-all, I think it's a great kit and well worth looking at if you need to interface QRP "seperates".

2/2/13

A Case for the KX3

Not an argument extolling the virtues of the radio - it does that pretty well by itself.  Instead, a physical case for carrying the KX3.

I still haven't taken my KX3 out in the field to speak of...although I did carry it to NA4SO's shack last week, and apparently while in my pack, something banged into the RIT knob and cracked it.  I only had the rig in a thin neoprene shell, and there were a couple of heavy crimpers in the pack, one of which I'm guessing banged into the KX3 - glad it didn't hit the clear bezel!

So, I'm trying to get back to deciding what case would be best.  I'm looking for something that protects well, but I don't need - or want - a bunch of extra compartments for carrying "stuff".  I just want good protection for the radio itself, and for that package to be as small as possible for SOTA hikes.

Here are some possibles:

 This case is being used by W7GJ.  I think it would be great for carrying the KX3 along with more stuff...but that's not what I'm really looking for.  This bag doesn't look ideal for using in conjunction with a backpack, which is what I need.  Although, I've had a number of LowePro bags over the years and always like them, so I'd say this is definitely worth considering if you have the need for this type of case.








Here is a better candidate for my needs, and one I'm giving serious consideration.  It has a better form factor, and while having some extra compartments, the bag isn't grossly over-sized.  I learned about this one from G4ILO, who apparently learned about it from N6KR himself, so it's got to be a good candidate.









Obviously something like this Pelikan would offer great protection, but it's far from petite, and the case alone weighs in at 2.5 pounds, so it's not exactly backpack friendly.  But I'm still looking at the weight vs. protection tradeoff and may go for it.








If anyone has any good ideas, let me know via email at the address listed on qrz.com