Measured Success

After installing the Homebrew PCB with the 10MHz TCVCXO Frequency Standard Clock on the Propeller, I finally got it connected to an Antenna as a WSPR, QRSS and OPERA 200mW Beacon.

The results are amazing, the Frequency is measure as very Stable and the Frequency is measured to be as set.  The reporting site, and measurement that I trust the most is W1BW (Operated from ARRL HQ);  they report Zero Hertz Drift and at the desired frequency of 10.140210MHz - exactly as I have programmed the Beacon.

Note: Not only does the TCVCXO provide for a stable and accurate RF frequency, but it also provide accurate scheduling for the Beacons. WSPR requires scheduled transmittion on two minute boundaries, starting the actual beacon two seconds (+/-1) past the minute. This Propeller Beacon is on-time, every time :-).

Also, the three beacons run in quick succession, that is: WSPR starts on a two minute boundry (runs for 2 min), followed by QRSS on the top of the next minute (runs for about 3 min), followed by OPERA on the next minute (runs for about 2.2 min). Then all repeats 10 minutes later. The 10 minute schedule allows the QRSS beacon to be displayed at the same place within the 10 minute remote Grabber captures.

I am receiving many reports (spots) from around the US, no DX yet, but the night is young.

My WSPR on the Map

Note: the W1BW report (at 3939 km) in the following list (click to view).

My WSPR Spots - as Reported by WSPRNET.org

Also, the following was reported by the KK7CC Grabber, note the frequency scale, which is out side of the normal QRSS Subband, I had to pick a frequency to allow my signal to show up on his Grabber. I picked 10.139930MHz, KK7CC reports the same !

My signal is the lower trace with Propeller graphic and my Call in slash code.

My QRSS on KK7CC Grabber - Las Vegas

I am a Very Happy Camper !!   :-)

---

UPDATE
New DX report at 7049km with only 200mW:

Timestamp	Call	MHz	SNR	Drift	Grid	Pwr	Reporter	RGrid	km	az
2012-03-17 03:54	WA0UWH	10.140204	-24	-1	CN88xc	0.2	WA2YUN	RK39hh	7049	270

And, Two more:

Timestamp	Call	MHz	SNR	Drift	Grid	Pwr	Reporter	RGrid	km	az
2012-03-17 10:04	WA0UWH	10.140208	-14	0	CN88xc	0.2	VKFRG	QF44np	12766	244
2012-03-17 10:04	WA0UWH	10.140199	-26	0	CN88xc	0.2	VK4TMH	QG56fj	11758	252

--

Stable Frequency for the Prop

More testing is needed, but it looks like the use of a 10MHz TCXO is going to work very well with the Propeller Software and Beacons. See previous posts, where the plans were made.

When I first installed the TXCO I could see that it was going work, as it did not produce the Hash (or fuzz) on the transmitted signal that was shown on previous blogs.

From turn-on the Frequency was very stable, after a few minutes no change could be detected. The output was NOT exactly on frequency, as I had not bothered to hook up the "frequency voltage control" line, I wanted to see the behavior without, and the single page manual does not provide much details of it's use.

The Frequency was low by about 1.4ppm as indicated on a previously calibrated Spectran. Within the Spin program I was able to correct the frequency and ran several more tests. All tests worked as expected (or hopped).

To complete the installation I hooked up a simple resister divider to drive the Frequency Control line; a 1K ohm fixed up to 3.3v, 2K ohm variable to Freq Control, and a 1K ohm fixed down to ground. Then I removed the previous Spin Freq Corrections and re-compiled the program.

The RF output frequency was within a few Hz of spot on, a quick twist of the variable Freq Control Pot put it on frequency. The pot is a little touchy for small adjustments; I will replace the two 1K ohm fixed resistors for maybe 2K ohms each.

Beacon Stability with TCXO
WSPR Above, QRSS Below 
as Seen by my Local (overloaded) Receiver

If you look very-very close, you my detect about a 1Hz drift between the two WSPR signals, but I am not sure that I can trust my receiver, or any part of my test configuration for that fine of resolution. Normally I check with http://wsprnet.org and other external sites for conformation of details, but wsprnet was down when I checked.

More tests are needed, but I think this is a excellent fix for the RF Frequency Stability Problem for the Propeller Beacons.

I will add an installation photo here, later.

--

Display ON

I have been running my 30m Propeller QRSS, WSPR and OPERA Beacon the last few days, and have received many local and DX spots as reported on the web.

A periodic restart (reboot) or re-program of the beacon is needed to keep the beacon on frequency, to change the mode, and set the right time sequence - which is a bit of a pain. What is needed is a GUI with display and controls to make adjustments as necessary while the beacon is running. To that end, I have been attempting to build a standard GUI for the Propeller Microprocessor - sounds simple doesn't it. Well it maybe simple after I get over some initial hurtles.

The first GUI hurtle was to get a LCD Display to work correctly, there are many contributed LCD driver in the Propeller Object Exchange Library, most use conflicting pin connection or they do not provide the 4bit data path that I desire for my display.

The GUI

The real problem here is there are Too Many LCD contributed drivers from which to choose. all with conflicting pin assignments and programming interface. It is difficult to understand the difference between them without actually downloading and trying each - which is very time consuming.

After much searching I finally found "jm_lcd4_ez" REV 1.3 (see previous post), which is a LCD driver that meet my requirements. In the Object code the author provided additional entry point for alternate hardware configuration. Many of the available LCD drivers require adjacent data and control line connections to the micro as the data and control are transferred via one instruction as a full byte. The Propeller is more flexible than that, and there is no need to be that restrictive. The selected author recently added the required flexibility within the top level Initialization routines, I tried his driver and it worked most of the time, but it had problem at power up.

I almost gave up on this driver and starting looking for yet another, even though I liked the documented flexibility and implementation. The author forgot to do the same flexibility modification at the lower "lcdinit" level. With a little work and experimenting I was able to make the correction to the low level routine. And now,  all is well.

If there is interest, I will publish the corrections here, else I will try to chat with the author.

The next step for GUI development is to program the two Rotary Encoders.

UPDATE
I received an email reply from Jon, the author of  the "jm_lcd4_ez" LCD driver object, he has now confirmed the bug, and is going to fix and update his Object (to REV 1.4) on the Propeller Object Exchange. - Thanks Jon.

--

Propeller on 30m QRSS

As I expected, the Propeller PLL Synthesizer Phase Jitter Shows up on QRSS (see previous post). The received signal appears "fuzzy".

In the next few weeks I will continue the experiments to see if there are modes or functions that will help reduce the annoyance to an acceptable level or irrelevancy.

30m QRSS

The top signal is my QRSS "WA"  (for Washington State) and an unseen CW ID. The lower received signal is from "NM7J" (200mw) in Las Vegas, at 905.1 miles.

Note: my "A" is not a true "A", it is an open "A" without a cross bar (it was quick and easier to program).

To receive the above, I am using; a Homebrew 30m Vertical at 0 AGL, Ensemble RX, WinRAD, and Spectran, connected via Virtual Audio Cable (VAC).

I am still using only short wires for the antenna, as I have not built the necessary low pass filter for connection to a real antenna.

--

30m Beacon In The Window

Here is my 30m Propeller Beacon running in (near) the window (see previous post).

Beacon in the Window

Now it is time to do some electronic design and build a output filter so that it can be connected to a real antenna.

---

UPDATE
I thought by this time I would have some examples of received signals from my Beacon, but my HF rig, audio link, sound card and software is not working. I need to spend more time with the problem.

I will be very interested in how much Propeller Processor PLL phase noise will be present. Will it be objectionable on QRSS?

For now, the received signals would have to be received by my own receiver, as the power level is too low and the Beacon is without an outside antenna.

--

A Propeller for Christmas

I got a VERY unexpected surprise for Christmas, . . . even though I ordered it for myself.

I have been playing with micro processors for the last few months, with the idea that at some point they will control my beacons or other Amateur Radio Projects and Devices. So far, I have been using the Arduino and Teensy micros for most of my exploration and learning.

One of the Blogs that I follow is Wardy's Projects, he posted about a new device for him. It is the Propeller from Parallax. My initial reaction was; yeah, right, . . another "me-too micro". And, it was made by same people that made the Stamp? The Stamp is one micro that I just never got around to exploring. But, as per Wardy, the Propeller promises to have some interesting capabilities that I would find useful.

Note: while writing the post, the Parallax site appears to be down, what's with that!

After doing a little more reading/research, I decided to order a Propeller "Gadget Gangster" it was about $50.00 and it appeared similar to the Arduino Shield that I had been using. I picked up the order at the post box on Christmas eve.

Some of the attributed of the Propeller Micro Processor that I though would be interesting, are:

• The processor has 8 cores, known as COG's, each is an independent processor which can run it's own code.
• The master clock can run up to 80mHz (or much slower, to save power), and some people over clock it even faster.
• Each COG can control two Frequency Synthesizers, from DC, to an amazing 120mHz.

Some of my concerns were; it does not use "C", it has it own languages called "Spin" and "pasm", although Spin looks a little like "C", . . . it's just another new language to learn.

The Propeller Processor looks like it maybe an ideal Amateur Radio RF Frequency Synthesizer, just ideal for my planned Transmitters, Receivers and Beacons. The raw Propeller Chip is available in DIP and SMT format, for about $8.00.

It took a bit to get the Propeller running (i.e., I am familiar with the Arduino, which obtains its power from the USB cable). By adding a 9Volt Battery and Clip, the hardware installation was complete. Installation of the Propeller IDE software was easy, a download and a few mouse clicks. It did not run, under WINE on my Ubuntu system, it ERRORd with something about the USB could not find the ports. But, there is a download for Linux and it installed quickly. The example and library directories are easily found in the MS directory structure of the first install that I did under WINE.

I clicked on the Frequency Synthesizer example, and its required library, which were loaded into two source TABs on the IDE GUI.

The source files were short and simple, I could see in the source file where "they" set the desired frequency, so therefore, with a quick edit, I picked my favorite frequency: 10.14mHz (typed as "10_140_000", the compiler ignores the under-bars).

I pressed F11, an ERROR! The compiler could not find the USB, pressing F7 fixed the problem. Pressing F11 once again --- out of the speaker across the room came a TONE!!! I had my HF receiver set for upper sideband on 10.1387mHz (the normal QRSS and WISPR receive frequency). - IT WORKED !!

My very unexpected Christmas surprise? - It was so simple.

After modifying the programs a little, I now have a Beacon sending my Call sign with CW, and sending multi-level FSK QRSS signals. Before going live with an antenna, I will need to build a low pass (or band pass) filter for the output. Currently it is just; a Battery, the Propeller, and two short 3 inch wires for an antenna.

And just think, I am only using one of 8 COGs and one of the 16 frequency synthesizers on the processor! I can do a lot more with this thing. Spin is going to take some time to learn, but I think I can like it.

The only downside that I currently see is the Propeller is not 5v TTL, it is a 3.3v CMOS processor. If real TTL levels are necessary; an input 5v TTL can be connected via a 4.7K resistor, for output a TTL driver circuit maybe necessary, and yet some TTL device will work with the CMOS 3.3v output.

I am just getting started with this thing. I am going to have fun with my Propeller in the coming mouths, . . .

Thanks to Wardy, and Santa!

--

Boards Received from China

I am an avid Homebrew PCB builder using the Toner Transfer Method, I can build a single sided board (similar as shown below) in about 20 minutes, and a double sided board in about one hour. I can also apply a green UV Curable Epoxy Solder Mask on either or both sides. A solder mask takes about 2 hours to apply.

As an experiment, I wanted to try the services of a commercial PCB production shop, the following is the initial results.

My 30m Beacon circuit boards was sent out for processing, it just arrived back from the PCB manufacture in China.

发件人： IteadStudio

发送时间： 2011-04-15 23:44:47

收件人： order

抄送：

主题： Fw: Re: Order Update #1550

After receipt and close inspection, I am very pleased with the results.

The schematic and board was designed with DipTrace Software and then six Exported Gerber Files and one Drill file were sent via e-mail to IteadStudio.com in China for manufacture.

To access the PCB service at IteadStudio.com; click on the tab on the site's top; "iStore -> OpenPCB -> PCB Prototyping Service".

The Original Layout as Designed in DipTrace PCB Software

As shown via the DipTrace 3D Preview Software

This is a photo of the resulting PCB as Received from Iteadstudio.com

The results are amazing. This is a; 1x2 inch board, with 8 and 6 mil Traces, 10 mil Clearance, the Ground Grid is 8 mils on 24 mil centers, and the Plated-through VIA's are 40 mils with 15 mil holes.

As shown in the magnified view below, there are multiple ground grid traces between the 1/10 inch on center Header holes. The square SMD pads are for 0603 LEDs and their 0603 limit resistors.

       

As Designed                                                    As Shipped

I received 10 boards for $12.00 plus $4.00 Shipping, that is $1.60 per board - a very good value. To obtain this low price, boards must be equal or less than 5x5 cm. Larger boards are available, but at higher price.

The only down side is the shipping time from China, it was 15 days after manufacture, 22 days total. But the results for this hobbyist was well worth it!

 Five of the 10 Boards are Marked as  being Tested

Chris and Dave of the TheAmpHour.com put me onto this great available service - Thanks Guys.

Note: I did not experience the Ground Plane Dropout as Dave Blogged, but then maybe I did not use similar geometry.

I am very pleased with the quality and results, and plan to use the service again, where multiple copies are needed and turn-around delay is acceptable. But, for my "one off" projects, I will continue to use the Homebrew Toner Transfer Method.

OK, one nagging question, what am I going to do with ten of these circuits, when I really only needed one or two? Well, this was just a test or exercise of the process, for later and more sophisticated circuits, and where copies will be sold.

Soon, I will post the results after loading a board.

---

UPDATE

I found a similar priced and faster service, based in Portland, OR. See: my follow-on post.

-- Home Page: https://WA0UWH.blogspot.com