Unode Biased Push-Pull FET

This is my new Homebrew "Vacuum Unode Biased Push-Pull FET Power Amp" of my own design.

Unode? - with the vacuum tube containing only one element, look closely, you will see, I did not know what else to call them :-)  And, I like the name and the glow.

I have experimented with this type of bias for FET Power Amplifiers before. Building some Amps with varying degree of success. The Unodes (lamps) are used as a Positive Temperature Coefficient (PTC) element providing a nice stable bias environment for power FET's.

Unode Biased FET Power Amp

This is the prototype, with Homebrew PCB being planned (and soon created). Performance measurements, Low Pass Filter, and other part modifications will most likely follow. For now the Amp is doing a smoke test. Power and Spectrum measurements will soon follow.

A real Toner Transfer PCB will eliminate the Grounding Solder Globs and make the amp much more presentable. Soldering over the edge of a Manhattan pad is always difficult. Next time I build this style of prototype board, I will use Solder Wick for the over-the-edge ground connections.

This power amp (in it's final form) will be used with my Propeller WSPR, QRSS, and OPERA Beacon.

--

QRSS Report

A new QRSS report, this time from Vince - N9VN near Chicago at 1676 miles, Thanks Vince.

QRSS Report
1676 Miles with 62mW's

That is my Propeller Graphic, SlashCode Call, and followed by "WA", as seen above the Flying W Graphic from W1BW.

--

WSPR, QRSS and QPERA Beacon

I am have way too much fun with my multi-mode Propeller WSPR, QRSS and OPERA Beacon.

WOW, I Just got a Spot from Antarctica at 15,825km with my little 62mW Beacon !!

For specmanship, that is 157,328 miles per watt.

WSPR Spot with 62mW's
see: http://wsprnet.org/drupal/wsprnet/map

Timestamp	Call	MHz	SNR	Drift	Grid	Pwr	Reporter	RGrid	km	az
2012-01-30 03:02	W5CGC	10.140234	-21	0	EM13ka	5	DP1POL	IB59uh	13407	157
2012-01-30 02:56	VE6PDQ/1	10.140161	-16	0	DO33fl	5	DP1POL	IB59uh	16025	147
2012-01-30 02:36	DJ6OM	10.140164	-21	0	JN58se	1	DP1POL	IB59uh	13314	187
2012-01-30 02:30	DL3KAV	10.140281	-26	0	JO30ms	0.5	DP1POL	IB59uh	13561	186
2012-01-30 02:30	K5XL	10.140252	-19	0	EM12kp	1	DP1POL	IB59uh	13369	158
2012-01-30 02:30	ON7BBR	10.140225	-23	0	JO21fh	1	DP1POL	IB59uh	13604	185
2012-01-30 02:30	C91PM	10.140244	-15	0	KG64ha	5	DP1POL	IB59uh	5599	196
2012-01-30 02:28	WA0UWH	10.140176	-21	-3	CN88xc	0.05	DP1POL	IB59uh	15825	150
2012-01-30 02:28	G0MJI	10.140185	-13	0	IO83ni	5	DP1POL	IB59uh	13800	182
2012-01-30 02:22	VA7SK	10.140205	-20	0	CN89lf	5	DP1POL	IB59uh	15970	150
2012-01-30 02:20	KF7O	10.140155	-25	-1	DN06tb	1	DP1POL	IB59uh	15492	152

Each mode (WSPR, QRSS and OPERA) is transmitted once each 12 minutes.

My Propeller 30m WSPR, QRSS and QPERA Beacon Station

--

OPERA Running on Ubuntu, It "sort of" Works

I have OPERA 1.1.6 . . .  "sort of" working on Ubuntu via Wine :-)

NOTE: The following has "Hung" the Ubuntu Gnome Display System, I had to ESC to command line to recover (ctrl-alt-F1), kill the OPERA process, and then return to the Desktop via F7.

CAUTION: Use at Your Own Risk

Two DLL's were needed; MSVBVM60.DLL and pdh.dll. I found them online, but was advised not to download from un-trusted sites.  I have a local XP system from which the files were found, and I copied them into the Ubuntu OPERA directory,

To make things easier, I change the default permission of the executable to "mode 755", with:

$ chmod 755 Oper*.exe

Then I ran the Install as:

$ wine ./Opera*Install*.exe

To run OPERA, execute:

$ wine ./Opera\ v1.1.6.exe

After adding my Operator Information and selected the sound card for input, the waterfall came alive with signals. The sound card was over driving the Decoder, but a quick adjustment to the Ubuntu Volume control fixed  that.

I was able to decode several signals on 10135mHz.

Sill known problems that I have not solved:

1. The CPU Usage meter does not work.
2. The CAT Rig control does not work for my FT-817, but Tones are created on the output as can be heard with a temporary speaker connected.
3. Attempting to Set Configuration sometimes Hangs the Display System

Like most MS programs running under WINE, there will be a lot of WINE "fixme:" errors logged to the command line window form which OPERA was started, most can be ignored.

Maybe these problems will be solved by someone knowing much more than I.

--

Mackinac

Tess and I just survived 4.5 days without Power, Heat, and the Internet.

Without the Internet, I only had my paper "Sprat Magazines", "Radio-Frequency Electronics" book and few catalogs to read for entertainment. I will stock more printed material for next time.

Driveway
(the way out)

We are still snowed in (10 inches in a 300' driveway), but the power is back on, and the Hanger/Shop is warming up.

Life may soon return to normal.

---

Update: My friend Jeff - KO7M tells me I misspelled "Machinac" - OK, I am not a local :-)

--

Propeller Beacon Feld-Hell Font

The Propeller QRSS Beacon gets a Feld-Hell Font, the mode and font is known as Hellschreiber.

Check it out, on Jeff's Blog:

http://ko7m.blogspot.com/2012/01/more-propeller-madness.html

Soon, Hellschreiber mode will be added to my Propeller QRSS Beacon.

Note: Jeff created the font from scratch.

--

Propeller Tests on WSPR

My friend, Jeff - KO7M, has been experimenting with WSPR on the Propeller. Check out his blog at:

http://ko7m.blogspot.com/2012/01/wspr-on-propeller.html

We have been trading Propeller code segments for experimental Beacons. Due to his lack of a Low Pass filter, I do not think he has actually put it on an antenna yet, but he is working on a HB design. Soon I am sure on-air experiments will follow.

Jeff and I talked about how we could provide non integer frequency shifts as needed by WSPR, I suggested that maybe we could just use "1Hz" shift and let the proclaimed PLL Jitter do the rest. It looks like it has worked - now that's a hack!

Thanks Jeff, I will have to include WSPR into my current Propeller 30m Beacon, maybe after you completed your tests.

--

Beacon Stacking via 10 Minute Sync

I received the following via Google+ and emails, it is archived here for myself and my blog readers.

---------------
From Peter Mulhare

Eldon

Just got in from Dinner........here is your first Stack pretty amazing, absolutely perfect copy from a stack of 6 images - 1 hours worth. I think you will be suitably impressed. Note that ZL1EE, W4HBK have also shown up stacked, also in there are G0MBA (@ '090), KC7VHS (@ '070), G0PKT (@ '980), and another station @ '925. These stations are not stacked because they are not running 10 minute frames.

Excellent result Eldon!

Pete
iK

Six Stacked Frames

Single Frame

--------------
From Eldon Brown

Peter,

Would you mind if I share your message and Initial Stacked Results on my blog? I have many readers that do not have Google+ and I want to Archive the results within the blog.

When we establish all is working as expected, I want to disconnect the Power Amp (62mW) and run directly from the Propeller with only 3mW. Your stacked images maybe be the only way to see the results.

Thanks Eldon - WA0UWH

-----------------
From Peter Mulhare

Hi Eldon

I don’t mind if you share the results, that’s fine by me. The best I’ve done at those kind of powers was to Western Australia, 5500 km to VK6JY, running 7mW from the output of my FEI 5680A Rubidium Frequency Standard, keyed FSK with an ATmega16 keyer I designed. I think we would need pretty good conditions to do it, but let’s give a shot!

The windows of opportunity at the moment are pretty short, limited to grey line propagation around one or other of our sunrises or sunsets, with the other one in darkness. I will have a play with DXAtlas, and see if I can come up with some dates to try.

Note that Bill W4HBK was in nearly all night on the Grabber, now that’s due to the way the grey line effects work between me and Florida. The problem with Washington is that you’re so far North and much closer to my Longitude, making us a north/south path, where to Florida the path is so much more east/west.

I’m impressed with the accuracy of your timing! the stacked signal has no thickening due to timing jitter, which is excellent, the computer clock here, is sync’d to the UTC second with several NZ University Caesium Beam controlled NTP servers, over the net. Software is Meinberg V4 Timeserver.

Pete
iK

-----------------

The First Snow - 2012

This is the First Snow of the year has just started, it is a good time to be inside with Ham Radio, and having fun coding my Propeller Beacon (see previous posts).

The First Snow - 2012

The photo was taken from my Loft (shack) window, with the 450 ohm Ladder Line in the foreground. The Ladder Line feeds a Off Center Fed (OCF) Dipole 90 feet above ground. So far the 30m 62.5mW Beacon and Antenna have received good reports from many remote locations. The beacon is transmitting both QRSS and OPERA on a 10 minute sync cycle.

A nice map of Generic OPERA reports are available at http://pskreporter.info/pskmap.html

--

Beacons in Sync

Propeller "Multi-COG" Programming is much simpler than I had imagined.

To date, I have used time delays to to pace my QRSS and OPERA Beacons, both running on same Propeller Micro Processor (using only one COG). The timing was typically just a 6 minute delay after the last transmitted data.

Peter and some other Grabber owners can Stack grabs to provide even more sensitive reports. A similar technique is used by Astronomers to photograph very faint stars. The typical QRSS Grabber is programmed to capture a frame each 10 minutes. To play the game, QRSS Beacons need to sync to a 10 minute interval, it does not matter where in time, but repeat in sync with itself.

So, this requirement provided me an opportunity to learn more Propeller Spin Programming. I tried several approaches to provide a Real Time Clock or Sync Timer Function, but nothing I programmed was working very well.  With more reading, learning, and experimenting, I decided to try to build a "Epoch" counter in a second COG.

It was so easy I could not believe it. To start a new COG, the following line is all that is needed within the first (default) COG:

cognew(procedureName, @someStackSpace)

The "cognew" statement specifies  the new COG's default starting procedure, in the same way as the default GOG (0) always starts with the first listed procedure in the Spin file.

The new COG has access to all of the routines, variable, and constants listed in Spin file. In my case, I need a "delay(1000)" routine to update the Epoch counter within the new COG.

PUB UpdateEpoch                 ' Runs In its own COG
    repeat
      delay(1000)
      Epoch++

Here is the First COG Code for "Main" that starts and uses the Epoch for Sync:

VAR
    Long RFFreq

    Long Epoch, EpochStack[16]

PUB Main | Sync

    doInitialize

    cognew(UpdateEpoch, @EpochStack)

    Sync := 60 * 10             '10 minutes
    Epoch := 0

    repeat

      RFFreq := Qrss_Tx_30m
      doQrssBeacon
      noTone
      delay(10000)

      RFFreq := Opera_Tx_30m
      doOperaBeacon
      noTone
      delay(10000)


      ' Delay for Sync
      repeat while Epoch // Sync
        delay(100)

Note: The "//" operator is modulus function.

The new and much improved (published) delay routine is:

PUB delay(ms)  | t

    t := cnt - 1088             ' 1088 is Published Time for Overhead at 80mHz
    repeat ms
       waitcnt(t += MS_001)

Now, my Beacons should now show up on the Grabbers in 10 minute Sync.

I am enjoying Programming the Propeller !


--