Wednesday, February 10, 2021

The Bullseye LNB for QO-100

It finally arrived in the Netherlands, the most expensive LNB I ever bought. For a whopping  €25 and many weeks of waiting, the Bullseye LNB was mine. Well, nothing to complain about really, I knew how long shipping will take, and, seen reviews on YouTube, the LNB seemed worth the money and the wait, in particular when compared to competing products.

So, the shipment came in, finally. Of course, I waited until after midnight before I finally got the LNB installed in the focus of my downlink dish. You can't do those things in daylight... it just would not be cool... literally, with have a winter wave with temperatures about -9C. 
Anyway, the LNB is installed and was surprisingly easy to find in the spectrum. Well done! Also, for the time being, the LNB appears to be very stable. Again, well done!

If the positive experience holds true, I will be hooking up my FT-736 again. Something I really looked forward to since I first learned about QO-100.

Sunday, January 17, 2021

QO-100 Up-Link -- So Many Things Wrong

There are so many things wrong with my up-link feed:

  1. RG-6 75 Ohms feed line from a 50 Ohms PA
  2. the base of the helix is embedded in hot-glue
  3. the material is Zn-plated Cu-wire (Zn corroding)
  4. the spacing of the helix is off and narrows towards the end

2.4 GHz up-link helix feed

You are seeing the result of ignorance and tweaking, which resulted in the feature of point 4).

I hooked a nanoVNA2 up to the PA side of the RG-6, then pushed and pulled on various parts of the helix to minimize the SWR. This procedure included the 75 Ohms transmission line of course. Thereby, I improving the SWR from 4.6:1 to 2.4:1.

Next step you probably be to play with a wire-cutter again. As you can see, the helix has 3+3/4 turns, so, there might be room for improvement still.

For the time being, I am pretty happy with my up-link signal anyway.

Sunday, November 29, 2020

Ender 3 and the Black-and-White Dog

In the past, I was proud enough to fabricate every little piece of hardware from either plastic, metal of wood, with traditional tools. It's 2020 now, so the decision was made to modernize my tinkering and add 3D-printing to the arsenal.

My Black Friday choice of purchase was a Creality Ender 3. Mainly for the reason hat it is so widely spread. 

The printer came semi-assembled and was easy to put together. There are some very good tutorials how to do that on YouTube. I followed one by "3D Printing Canada", which explains the alignment and bed-leveling very well.
For all you experts out there, the model I got is version with a Melzi board.

before assembly

The printer came with a small roll of white filament. With the printer all being setup and aligned, I dcided to print the dog for which a gcode file is supplied with the printer. The dog took 3h 01m 46s to print. In the process, the printer ran out of the white filament. In principle a very good thing, so I could practice changing filament in mid-print on a object of low importance. That of course created the black and white dog, as referenced in the title of this post.

the black and white dog
Note the little white lock which came from the remainder of the white filament in the extruder chamber.
After having changed the filament, still 2 or so layers of white came out, perfectly aligned with the first part of the print. However, the first black layer is slightly shifted in the Y-axis. Could that be due to the different filament? It is still PLA... 

By now I printed several essential improvements found on thingiverse. The STL-files of those parts were slized by Sliz3r on a Raspberry Pi 400. 
The first one came out a bit rough, but usable. For nozzle temperature for this print was set to 185C. The dog was printed at 200C and came out fine. So, for the next few parts, I created gcode files for a nozzle temperature of 200C and a bed temperature of of 60C.

As a design software for my own projects, I found that solvespace would serve my needs perfectly. The program is available from the Raspberry Pi OS repositories and work fine OOB. Solvespace is similar to many other CAD programs.

This is my first attempt of becoming less of a dinosaur, however, I am not sure about this yet. With a decently equipped workshop, with materials at hand, making a part from a block of metal or plastic appears to be much less time consuming. Not only do I not have to use a computer to design the piece, I also do not have to wait until the printer is done. Even worse, one has to babysit the printer, just in case something goes wrong. Further, if the printed part is not perfect, there is very little one can do to improve the part. When working with a lump of metal, I tend to make the part just a tiny little bit bigger here and there and use a file, the one from the toolbox, not the one sitting in your computer, and adjust to the proper dimensions manually. 
Well, his is my first day playing with 3D-printing, however, I already feel the disadvantages.
To be honest, there is an advantage to CAD and 3D-printing, which is huge. Once a part is successfully made, it is easily shareable with others. With hand-made parts, this might be different, in particular when a single piece is made in some sort of workshop.
Concerning ham radio, one might think of driven element holders for Yagi-Uda-arrays or base plates for a variable capacitor, printed in ABS.

So, in conclusion, I look forward to becoming more efficient with 3D-printing over time. For now, there is a learning curve ahead, which might slow down my projects by some good amount of time.

Monday, November 16, 2020

QO-100 observations

Having a satellite in a geosynchronous orbit available is a great privilege for the amateur radio community. In my few months being active on the transponder of this precious satellite I observed the good, the bad and the ugly.

Writing about the ugly. I was listening to a conversation in Spanish. Si puedo comprara... At the end, the DX station was calling CQ. Well, someone else was faster than me. Fair enough, I waited. At their end, I called and was picked up by S0S in West Sahara. What a pleasant QSO, he even spoke some Dutch, respectful of my call sign. What a gentleman! Our QSO ended undisturbed.
I stayed on frequency, since I had some in shack business to conduct. Moments later, some from OE-land (aka Austria) SHOUTED his call (noted but not disclosed) at the transponder, over everybody else, of course triggering Leila. However, his signal was that strong that his call was audible the entire time. HOW RUDE! Of course he had the next QSO, so that he shut the f##k up. 
Is that how we want this transponder to be for the next 15 years?!

Honestly, the behavior of our Austrian friend reminded me of the reasons for which I decided to ignore the 80m-band 40 years ago. Unfriendly, unpleasant, un-HAM-ish behavior.

Dear friend from Austria, YOU SUCK!

Tuesday, November 3, 2020

QO-100 Uplink Setup Changes

My original setup used 10m RG6 between the PA and the feed, with a 100cm SMA patch cable between the converter and the PA.
When I reduced the length of the RG6, I also reduced the length of the SMA patch cable to 10cm. After this change, I had some QSOs, no complains. However, when I looked at my own spectrum, I was not so happy seeing splatter on the wrong side-band.
Probably, my converter, outputting 200mW, saturated the PA, which input is rated 100mW. Just as an experiment, I changed the patch back to 100cm. Having had some test transmissions, I can now state that my signal is clean again.
Sometimes, it only needs a little bit of coax in the right place to solve a problem.

Wednesday, October 21, 2020

QO-100 LNB Comparison

My downlink (DL) is realized by using an old 35cm off-set dish on my roof deck, something I mentioned before. On recommendation of an old friend, I bought a Venton EXL-S Rocket Single LNB. So far so good, this LNB served me on many contacts. Using the beacon feature of SDR-Console made it possible to have those contacts. However, during my tests using analogue equipment, drift from this particular LNB was noticeable.
I understood the Venton LNB is primarily useful with with POTY "dual band" feed, as it keeps the centre hole to a minimum. Fair enough. 

For my setup, I decided to use two separate dishes for UL and DL, hence, the size of the DL LNB is irrelevant to the UL feed geometry. As a matter of experimentation, I replaced the Venton LNB with an Octagon OSLO LNB in the 35cm DL dish setup. This LNB was purchased many years ago, when I was active on 3cm rain scatter and wanted to experiment a little bit with commodity gear. To my surprise, the Octagon LNB is a lot more stable, once at temperature. It also appears to be more sensitive than the Venton. Aperture is the name of the game here, I figure. Hmmm, maybe I should have used this LNB from the get-go and saved the few Euros by not buying the Venton LNB.

Still, there might be a use for the Venton LNB. As mentioned in my previous post, I am considering to use a single dish for DL and UL. With its smaller form-factor, the Venton Rocket might be the better choice when placed inside the UL feed helix in a single dish arrangement.

I will post my further experiment on this blog, so, stay tuned.

BTW, occasionally, you might be able to hear me on the transponder. Feel free to give my a call ;-)

Monday, October 19, 2020

QO-100 update

Not much to report, other than I shortened the RG-6 coax for the Up-Link. The length should be about 3.5m, however, I have not measured the length yet. In order to achieve the shorter transmission line, drilled a hole into a window frame relatively close to the UL dish. This allows the power amplifier to be indoors. 

By now, I made contacts in side-band from Brazil to Thailand. The occasional rag chew QSO confirmed that my signal is loud enough to not be difficult to listen to.

There is a plan of using the UL dish also for reception. The added gain in the DL might make it easier to operate with a full analogue setup. Although, the LNB being placed in the center of the feed-helix might compromise the UL signal. Something to experiment with...
Further down the line, I might acquire a larger off-set dish, e.g. 1m, mainly to add some gain to the UL.

Sunday, October 11, 2020

QO-100 thoughts

The humble listener/follower of my station or posts might have experienced a certain lack of enthusiasm about operating on QO-100/ Please do not interpret much into this. The satellite is great. Probably the greatest any Oscars had ever been by this date.
However, once the setup is done, the fine tuning is somewhat frustrating.

What to look for?

Ideally, in my world, one would be going fully analogue, i.e. using only pure receivers and transmitters.
Does that justify the effort by the operators of the satellite? I don't think so!
Hence, we also should look into modern techniques, such as SDR.

Well, this is exactly the spot I am struggling with. I got the equipment for both, analogue down link and digital down link. I operated with the satellite using both. Concerning the up link, I an analogue only. 

What I used so far (RX - TX):

  • RSP1a - FT817
  • FT290R2 - FT790R2
  • FT736R (RX+TX)
Any of the setups has advantages and disadvantages. 

For the time being, I am using the RSP1a - FT817 the most. However, I have a strong desire to move towards using the FT736R only. 

Tuesday, September 22, 2020

QO-100 Analogue Ground Station

Yes, it can be done. One can use a fully analogue ground station for QO-100 operations.

My first attempt included a Yaesu FT-290R2 for reception, along a Yeasu FT-817 for transmission on 70cm. Once a TX frequency has been established, following the drift of the LNB was an OK-ish task during QSOs. Changing frequencies was an unpleasant juggle of setting the receive frequency and trying to find the transmit frequency without annoying everybody else of the transponder. This ballet is performed with VFO controls of two different radios.

Second attempt: Yaesu FT-736R, the satellite machine from the past. It was doing great, in particular when changing frequencies, using the satellite mode. However, my cheap LNB kept drifting in sunlight, while the up-link converter stayed rock-solid. Hence, using the FT-736R was a constant switching between SAT-mode and RX-QRG. Once a frequency was established, TX-QRG needed to be adjusted. I am not sure how many actuations the little SAT-switch can take. What I am concerned, it took a lot of focusing in order to not adjust the VFO in the wrong setting, thereby messing up the RX/TX-offset.
With a stabilized LNB, this might be a different story.
I liked the possibility to use the VFO-knob to scan up and down over the band, knowing that, whenever I pushed the PTT, the frequency would be at least close.

The FT-736R is back in its box for now. In terms of having contacts, the SDR-option seems a lot more convenient, in particular when locking the reception to the satellite's beacon. Transmission is stable enough, in my experience, so nothing to be bothered about.
As an added bonus of the SDR solution of the analogue way, there is signal processing available when using a computer for reception. 

Further, I tried to use a Yaesu FT-790R2 as baseband up-link transmitter. It did work. However, there are two observations to share: 1) some weird spurious transmission were visible in A1A, 2) the audio in J3E was OK but not punchy at all.

Consequently, I will be using the Yaesu FT-817 as a baseband TX and the SDR receive setup, for the time being.

Saturday, August 29, 2020

QO-100 A1A up-link idea

This is something I had in mind for quite some time by now. What about an A1A only up-link for QO-100?

This is an old idea. At the time, this idea was to create a CW transmitter for 70cm based of a 2m signal. I have seen similar attempts to get something transmitting on 3cm. So, why not on 13cm?

Having done some "maths", I came up with a carrier of 96MHz, which is available as a regular crystal.

With a factor of 25, this will result in a 2.4GHz frequency.

Now, why is that interesting? The multiplication of 25 can be achieved with 5 x 5 .... dahhhh. Yes, sounds simple enough. Here is the interesting thing, it should be easy with a square-wave oscillator to generate 480MHz. At this frequency, it is pretty easy to build a strip-line filter or cavity resonator to clean the signal. Now, we just have to multiply the signal again. So, we might want to use a non-linear amplifier again, in order to create all the wonderful harmonics. Again, a 2.4GHz resonator should not be to hard to build.
And here is the idea. What if we used a rather high powered amplifier at 480MHz, which is easy to build, and a non-linear element to generate all those harmonics at decent power.

With a VTCXO of 96MHz, slightly pulled by a few kHz, one should be easily creating a signal in the CW band of QO-100, using the above mentioned method.