More Knob

Origins & Refinement of the Contest Knob - II

Stu continues with part 2 of his recap of the development of this fine contraption. Just for the record I am never satisfied but I am willing to be bounded :)

73

4K8M UN3F

I was doing a little correspondence when UN3F in Kazakhstan popped up on 30M. Recently I have increased my 30M operating. I am usually on 40, 80 and 160 and had only worked about 100 countries on 30M till about a month ago, so I have been trying to pick things up on 30 since not too much new in the way of spots comes across the cluster for the other bands. I could do a lot more on 160 but my antennas are still down on that band.

Just as I tuned him in he was working a WB8 and I was getting set to work him and he seemed to disappear, probably going QRT. He was past his sunrise so he was probably heading out to work. At the same time I saw 4K8M in Azerbaijan spotted so I loaded up the memories and had at it, clicking back and forth.

Unfortunately UF3F was truly gone but I had 4k8M in the log first call.


I have 3 antennas available for 30M and none of them are great. I have a 135ft flat top at about 50 ft fed with open wire to a Johnson KW matchbox. That antenna has as pattern that looks like this on 30M (north is strait up)


This antenna was not a very good match for UN3F but a better match for 4K8M

The second antenna is my 65ft 80M vertical. On 30M this antenna approximates a 5/8 wave vertical. This antenna has an extensive radial system and exhibits a SWR of about 2:1 on 30M. I better match that with an Ameritron tuner. This is the antenna I use the most on 30M


Finally I have a 45 ft vertical wire out of a tree that has a MFJ 929 auto tuner at the base of the tree. It has a beautiful pattern:


unfortunately the ground system on this antenna is abysmal since I mostly use it as a RX probe antenna for diversity, so I don't transmit on it much. UN3F is on fairly often so I'm sure I will have another shot at him before summer

73

Filters

My old buddy K3RR and I were corresponding and he mentioned the difference between his cascaded 250hz filters and my 12hz DSP filters. The theoretical difference is 10log 250/12 =13.2 dBm. I was sitting on 80M seeing if I could hear anything out of A52J:

(after all hope springs eternal) and since the band was so quiet I decided to measure the difference and make a little document.

Here is the radio with 12hz filters:


The RX2 receiver's vertical antenna has a little less output than the full size vertical on 80M connected to RX1. Best case today is in the -118dB range. For the 250hz case:


-107 dBm was the best I could do which is about 11 dBm difference. So the filter difference extends me another 11 dB into the noise floor. I decided to look at 500hz filters as well:


Predicted is 16.2 dBm and I got 14dBm difference, very consistent with the data point above. Some of the difference between predicted and measured could be due to AGC action since certainly by the time I get 16dB above the noise there is some AGC starting to flatten out the signal. I had "best" AGC-T (agc threshold, or the point at which the AGC just starts to reduce gain) set on the 12hz filters:


When I readjusted on the wider filters it didn't make any difference. The AGC-T adjustment is one of the best features of this radio. The AGC is designed to effectively turn down the sensitivity of the radio. If the AGC is not active the radio is most sensitive. By adjusting the AGC-T you can set the AGC to turn on at some point considerably higher than the band noise. This is a big deal when working DX since in most radios the AGC is already well engaged and turning down the sensitivity due to the band noise itself, so you are never really able to realize the full sensitivity of the radio or the full dynamic range. Since you don't need the radio to be more sensitive than band conditions allow by adjusting the AGC-T your chances of hearing someone right at the noise are improved.

Here is a shot at 500hz with the noise blankers turned off

-83dBm

and now at 500hz with the blankers turned on


Without the noise blankers, the AGC-T and the 12hz filters (and diversity) I basically would only be working G's, OK's and DL's on the bands.

Here is a shot of the radio on 10M with a resonant vertical attached

and here is a shot with no antenna connected to RX1

Still no Bhutan but VU4PB just showed up on 40 to tease me:

A nice European pile up and a K1 who is hearing him. It's the K1's gray line time so maybe tonight is the night:


73

Debut of the Knob

This is a project that is coming to fruition. It started as the brain child of Stu K6TU. I'm going to spend some time on the history and development of this project because I think it is a very interesting story. What it shows is how some people who don't know each other or know each other loosely, and who are scattered across the USA can come together around an idea and make it into a reality. Its a story also about a company Flex Radio who allowed for its software to be modified through the exposure of new CAT commands in PSDR so that this Knob could come to fruition. I asked Stu for some background and he published:

Origins & Refinement of the Contest Knob – I

I will follow with some more of Stu's history as it is published. What you see pictured above is the end product of this project. Once Stu K6Tu and Kevin K6TD had a design it was forwarded to Steve K5FR. Steve made up 5 prototypes. It was about this time Steve and I were hot and heavy into creating the SO2R interface using N1MM. I was totally unaware of the Knob development. A bit later I received a copy from Steve. Steve wrote the interface which now appears in DDUTIL and it allows the Knob to control PSDR

Stu is a master of PIC micro-controller code. A PIC chip (short for Peripheral Interface Controller) is a device which can be programmed to read inputs from the world outside itself, do some processing on the data it receives, and then output the results of its processing. Here is a short tutorial on how a PIC works

What was created was a USB based Knob controler for PSDR that could send up and down signals to DDUTIL as it was turned. The Knob also has a switch connected to it, like a big push button. In addition 3 more switches and some LEDs were added to the mix. The PIC scans its inputs and looks for state changes for example a knob twist. The knob when its twisted causes a series of signals to be sent to the PIC like the dits of a Keyer. If you twist a lot you generate a bunch of "dits" if you twist only a tiny bit only one "dit" gets sent. The PIC can also tell direction If you twist CW the function being controlled goes up and if you twist CCW it goes down. The most obvious thing to control is a VFO but the knob can also control things like audio gain and RIT and XIT. The button on the knob can control various functions for example if you push once you can toggle between VFO A and VFO B.

Stu put in a feature that allows for the Knob button to control up to three functions and and to be able to toggle between 2 banks of functions The original code Knob code only had three functions for the buttons, but at my request Stu recoded the firmare to include 9 total functions for the 3 buttons. Here is a shot of the DDUTIL interface


The 6 boxes grouped together are Knob click controls. Off and Sgl-clk are a toggle pair short click once and VFO A switches to VFO B. Short click again and you're back at VFO A. If you double click you will turn on and off variable tuning which is an acceleration feature of the knob. If you twist fast the software recognizes that and will move faster in the direction of twist. This way you can cover a lot of ground with a twist. If you then light on a station where you want fine control the software automatically decelerates and you get very fine control.

One of the commands we had to be able to access was the ability to tune the TX VFO when in diversity mode


Bob K5KDN at Flex created for us a means in PSDR to do that using CAT commands

If I LONG click the knob the B row of functions is activated


and I can once again do the various toggles I described on a second bank of commands. Here is a drop down of some of the commands available on the Knob button


All button presses are not created equal however and Dbl-Clk has some unique choices


Now comes the 3 buttons on the top of the box.


As you can see each of these also have the ability to control 3 different commands

and the command choices is pretty extensive


We included up to 8 Macro commands so the real functionality extends to the entire CAT set of PSDR which is close to 150 commands. In addition you can do complex commands in the macros like one button press for diversity on or off.

Here is a shot of my macro file for the Knob macros:


What the first 3 macros do is address memory 2, memory 3, and memory 4 in CWX

This turns my knob into a memory keyer for working DX!! The first three buttons are now all I need to break pileups and work the DX! In addition the knob does NOT need to be in windows focus to work its magic!

CWX is in need of some serious redesign, but Flex did some improvements in the TX/RX turnaround time and it now works in a fashion capable of breaking a pileup. I now have a memory keyer with basic functionality and push-buttons instead of mouse clicks! I HOPE someday they will tie the speed control to the front panel speed control so you don't have to adjust 2 boxes to change speeds as often happens when there is low band static and the DX is having a bit of a hard time decoding your signal. I would also like at some point the typical functions of a contest keyer.

Since the Knob is independent of the Windows focus, I can be in a logging program or a contest program and still control PSDR from the knob.

The last 2 macros allow me to change step size between 1hz and 100hz. I use 1hz to fine tune in a DX station, and then I switch to VFO TX and use 100hz to quickly cruise up and down the pileup, and I can do all of this without PSDR or DDUTIL being in focus

If I get confused as to what click switch does what, Steve created a little sub window as a guide and it is always visible. (I can not tell you how responsive K5FR was to the development of this control device. The guy is simply amazing)


It shows what VFO bank is active

and it also can memorize a custom label, so W9OY is displayed in the first button instead of Macro M30

Here is what it looks like on my desktop


This project was a blast to participate in. Four guys widely separated across the country from California to Texas to Florida, with widely disparate skills, came together and made it happen. Basically I didn't know Stu or Kevin at all. I knew Steve from other projects we had worked on. I was a late comer to the party give credit to the others primarily, but I did have considerable input in the Knobs final behavior. Thanks also to Flex for exposing the CAT commands we needed to make it happen. The development is ongoing and I will update as time goes on and more becomes available.

This is the reason I wanted take a little time and look at all the means there are to control the Flex Radio. To recap we have all the internal means to control the radio. Then there is CAT as demonstrated through Skimmer. Then there is the effectiveness of this external control on the overall behavior of a radio station By judicious design you can make the F5K and DDUTIL control central for a very complex radio station yet make the station behaveior effortless to use. Then there is further integration with other programs There are improvements and streamlining that improves productivity. Finally there is the Knob, which is a piece of hardware controlled by firmware and provides a windows focus free means to manipulate PSDR. All of this is tied together by K5FR's mighty DDUTIL.

Frank AB2KT, the developer of DttSP used to talk about using a midi surface to control the radio. That concept never really appealed to me because of the ergonomics of a midi surface and a keyboard and a paddle and a mouse and a MIC all spread out over the operating table, but this little knob is totally functional and takes up little room and it is also totally plastic (in terms of being programmable). Change the firmware or the software and voila' new things can happen. Maybe if we do two button pushes together we could get another 9 functions.... Naw Steve and Stu would kill me....

73

Andaman II

I emailed the link of yesterday's post to the man who collects the dough. Tonight Andaman was spotted on 160, 80 and 40 in the time frame consistent with the gray line propagation.


HURRAY!! I did not work them. I barely heard them, and of course I was greeted with a passel of US and especially EU stations calling the VU4 ON HIS FREQ, even though he was working stations up 1-2.


A few divined they worked him on his freq (like the WA5 seen on skimmer and a bunch of DL's OM's OK's etc) even though the real people working him were a khz higher. The 599's were popping up like crazy. Then after about a solid minute of UP UP UP UP by the cops and they decided to get off the DX and go UP but by then it was daylight on his end. I gotta give them credit for showing up!! I don't know if my rant had anything to do with it or if it was simply random, but it was good to finally have a shot.

It is so informative to have this technology. It completely takes the guess work and confusion out of analyzing a situation.

73

Andaman

So far this DX pedition has proven to be a real disappointment. They have been foraging in large pileups on 17 and 20 but have barely taken the time to go lower, and when they do the US is largely in daylight.

Here is the 40M "attempt" this morning


They showed up on the band at 7:59 EDT when about 1/3 of the US was already in daylight and here is the resultant "pile up". I heard them for a few minutes before they faded in this opening. If they had been here 30 minutes earlier (as in sunset is approaching ) I may have actually heard them well enough to work them.


There is a LP opening starting around 19:30 EDT and proceeding for about an hour and most of the time (as I have chronicled in the past here and here) they don't stay on long enough to even make it to the gray line. One would think that some planning would have gone into gray line operations but apparently not. Its actually pretty simple: When sunrise is approaching, while still in darkness, you switch a couple of the stations down to 160, 80, and 40. When sunset is approaching, while still in daylight, you do the same thing. It's not rocket science.

The cost of the DX pdeition is about $35,000. If you go to their website the first thing you are greeted with is a pretty much in your face DONATE page. Most DX peditions don't open with a donate page, but its fair enough. The guy collecting the money lives in Ann Arbor, MI. Interesting it's not some guy in Germany or Japan, or India for that matter. You would think 11 guys who want their exotic vacation paid for by what it would seem is mostly US dollars, would plan a little better to entice those who they expect to pay for that exotic vacation.

73

ST2AR hotter n a pistol

I was sitting here in the shack listening on YOUTUBE to Lee Ritenour and Pat Martino play a song Lee wrote for Les Paul RIP, and St2AR pops up on the cluster calling CQ.

I click the spot. Usually I would turn on diversity and set up the probe antenna and all that, but the guy is SO LOUD he's running as high as -75 dBm. I flip on the amp (instant on, no tune, solid state), the antenna is automatically switched from my 40M 1/2 wave vertical to my full size 1/4 wave on 80M and give him a call simplex. I'm the first one in line and in an instant he is in the log. Very good operator.


The guy is about 66hz higher than the spot and completely out of my pass band so I use the Knob to tune him in. I really like the addition of this accessory to my armamentarium.

In less than 15 minutes he was gone. He apparently had moved up to 40M. Ham radio blitzkrieg (German for lightning war) and a new one in the log on 80M.

I was reading a couple of old fuddy duddies on eham in the DX forum decrying all the new fangled gizmos bla bla bla old days bla bla bla ain't like it used to be bla bla bla. My first ham radio rig was a S-40B, a crystal controlled 6L6 and a knife switch to change between TX and RX. I had to turn down the RF gain switch the antenna and then turn on the transmitter before I could transmit, and then do the reverse to receive. In the old days I never would have heard this guy much less worked. 80M was about 1/2 inch long on the S-40B's dial.

Without the new days, I would have just finished listening to Lee and Pat play some bebop and headed off to bed, but then in the old days there wasn't any YOUTUBE. I got my license when I was 11 and I didn't have a clue who Lee and Pat were in those days, so I guess I would have just headed off to bed.

I like the new days much much better, and so does my logbook.

73

Russian DX

I was fooling around this morning and 15M was HOT! It's the Russian DX contest. I'm not much into contesting but I decided what the heck. I loaded up my 135ft open wire dipole as my primary antenna (RX1), and my 43ft vertical (RX2), loaded up N1MM and set up the log

logged into the cluster that I use with N1MM (W9AZ) and I was ready for some radio sport! I normally run 2 clusters K3NC and DX Central but when I have N1MM going I switch to W9AZ in Kankakee in both N1MM and Spot Collector Too many active telnet ports really bogs down the CPU so I have found using only one source for both programs keeps the extra CPU % down to a dull roar.



I mostly am interested in increasing my countries total. I certainly do not have the kind of antenna system or the skill set to effectively run a contest like this. I'm more interested in developing the systems involved in building effective contest stations (like the SO2R set up Steve K5FR and I created) than I am in actually running contests. I have spent way too many months up all night in my medical practice to stay up all night voluntarily.


This is the business end of things PSDR and N1MM. It's a point and shoot operation. Click on the band map and bingo you are on freq, the call sign is pre-entered into N1MM and if you can hear him you click the callsign and he is loaded for the contact. For example you can see my mouse pointer pointing at EM0K in the band map and I click and its preloaded to the spot above the station entry window in N1MM. I click that preload and the station is then loaded, make the contact enter the exchange and hit "log it" and I'm off to the next station. You can also see if you look at skimmer EM0K is indicated next to the green arrow. If I were to notice someone in skimmer I wanted to work just clicking his freq would bring him up in the contest system. If he was not logged in the N1MM band map I would just enter him and the exchange "log it" and move on.

If I copy a signal off the air I just enter it after I work it and hit log it, and it is remembered in case of dupes The other aspect is Spot Collector



If I notice something I want to work coming across this screen, I just click it. Since my data source for both N1MM and Spot Collector is the same (W9AZ), when I click spot collector what ever station I click is also in N1MM and all the automatic stuff happens, the station is preloaded etc. Spot Collector also loads the station in my normal day to day logging program DXKeeper so I can log new countries in that log book while I progress through the contest. If I am a little off freq I have a new toy that helps that


This is a contest Knob that is the brain child of Stu K6TU, has been in development by K6TU, K6TD, K5FR, and myself for a period of time. The Knob will be the subject of a further blog post when the time comes (in the fairly near future) It's MAJOR feature in it can control 16 functions of PSDR, without PSDR holding the windows focus, so if I am using a 12hz filter (which is what I do all the time) and the stations is 50hz or 100hz off the freq reported on the cluster I can simply grab the knob and tune up or down and get him on freq. I do not have to click to PSDR which makes N1MM disappear, tune him in and try to get N1MM back in focus. It makes contesting effortless. (Actually the way I have the screens set up it's not a major effort to change focus BUT I love the knob)

Well back to the contest before 15 fades into the noon time dust and all the Russians mosey on down (up?) to 20M and 40M

73

No VU4 but Singapore

VU4PB disappeared. S21YZ disappeared, BUT James 9V1YC decided to party.


His sunrise and my sunset perfect timing excellent signal

and in the log in less than 5 minutes

The bands are heating up.

73