4:1 Unbalance to Unbalance Coaxial Transformer

This is my first time to implement the 4:1 unbalance to unbalance coaxial transformer to match the input and output impedance of an rf power transistor. My prototype rf amplifier originally had a lumped LC network in both input and output sections and only have +/- 3MHz of bandwidth but after the addition of 4:1 coaxial transformer, the bandwidth has now increased to +/- 6MHz from the center tuning frequency of the amplifier.
The coaxial transformer is about 1/16th wavelength long from the highest frequency in which the amplifier will work and should have 25 ohm of characteristic impedance. In my prototype amp it uses a pair of RG-178 cable paralleled to arrive at the 25 ohm requirement.

This is the simplified diagram when looking at the transformer however, the characteristic impedance of the coaxial cable also plays an important role in the impedance ratio. For a 50 ohm to 12.5 ohm transformation, it calls for a 25 ohm cable but since I don't have this kind, I simply paralleled two RG178, a 50 ohm cable.

The schematic diagram of my prototype amp tells that aside from the 4:1 coaxial transformer, LC network was still employed to further matched the impedance of the base and collector of the power rf transistor. The only limiting factor that prevents the amplifier for a broadband operation is the LC network!

My prototype amp after the inclusion of the 4:1 coaxial transformer. I might replace the old PCB to facilitate the addition of the two coaxial transformer. ---73 de du1vss

Experiment With Different Wind Turbines

Two months ago I was tasked to build a wind generator that can be use to charge batteries and power LED lights here in our main office. The first thing that I was thinking is the vertical axis wind turbine (VAWT) suitable enough for our other projects that will provide back-up power in some of our remote  installation of met sensors. After a month long planning and fabrication, I came up with a Savonius turbine that drives our 350W motor hub from an electric bicycle. During the testing, I was a bit disappointed with the result, the wind generator starts to spin at 2.5m/s wind speed and only reaches at 15 to 25 rpm from a wind velocity of 5m/s, not enough speed to generate few volts of electricity!

This prompted me to think with other turbine designs that are more efficient and have less wind drag as compared to my existing Savonius turbine. Anyway, earlier this morning I was busy setting my small test fixture that will test 2 other types of vertical axis wind turbine. We have an old anemometer that we no longer use and found that I can use the optical sensor to measure the relative wind speed of the turbine under test.

The first is the small scale model of my Savonius turbine, I'm aware that this is the least efficient from the rest of turbines but i would like to see how well it does when seated on the test fixture.

Not bad at all, the Savonius turbine able to spin at 0.2m/s., rotation sometimes varies over time but is still provides me a consistent slow rotational speed.

The next is a Darrieus turbine installed in the fixture. It has a hard time getting to start initially even if the wind is moving at the blades at substantial speeds. It sways back and forth until it catches the right angle and start to spin rapidly. I was able to get an average of 0.5m/s wind speed as displayed in the anemometer unit but it is not self-starting, not a good candidate for my project.

The last model of turbine that I'm going to test is the Lenz turbine. From the appearance of the model, it looks like a combination of Savonius and Darrieus turbine.

The performance is very promising, it automatically start to spin rapidly and I was able to obtain an average of 0.8m/s of wind speed reading from the anemometer unit.

 Comparing all the three models that were evaluated, the Lenz turbine is very well suited in my project and just a matter of time, an actual turbine will be made and soon running in my wind generator project. 73 de du1vss

Callsign Plate

Beautiful callsign plate which I got from Philippine Amateur Radio Association (PARA) last year. The price is much cheaper than the one from the NTC (National Communications Commission) and that's why I bought two of these. One of the plate was hung at DU1VSS station and the other one is kept for future use. 73 de du1vss.

100W Dummy Load

A 100W quick and easy rf dummy load using a 100W 50ohm Caddock thick film power resistor. The large aluminum heatsink was bought at Rox Electronics. The dummy load was tested at 144MHz exhibiting an swr of 1:1.0 at 150W of rf power. ---73 de du1vss

LM35 Thermometer Project

LM35 is a precision IC temperature sensor with an output voltage linearly proportional to the Centigrade temperature. Thus for a 0 to 100 Degrees C, the sensor output will be 0 to 1V accordingly. The sensor works over a wide operating voltage from 4 to 30V with only less than 60 microAmperes current drain which makes it favorable on a battery operation.

My prototype is connected to a 9V battery and gives me a reasonable accuracy when compared to our calibrated ROTRONIC Thermometer using a type K thermocouple. A digital voltmeter is connected to the sensor to display the voltage output of the sensor. I also tried connecting the sensor into our Envitech Ultimate logger to record and monitor the room temperature in one of our AQMS  station.

Another possibility is to use the sensor to trigger some devices such as relays, blowers, fans, alarm, etc. I Had a great time building several automatic blower in some of my power supplies using this LM35 sensor, have fun with it.  ---73 de du1vss

Stripline SWR Meter

Another weekend project that I made is this SWR (Standing Wave Ratio) meter which rf pick-up is a directional coupler  fabricated from microstrip. FR4 double sided pcb was chosen  with thickness of about 1/16 of an inch on a 2mil copper thickness. Processing the microstrip starts from applying a masking tape on both sides of the PCB and manually drawn the three main important traces on the top side of the PCB. The masking tape on the bottom side was retained so that at the end after etching, it will serve as a ground plane of the microstrip. After etching the board, please verify that the dimension of L1, L2 and L3 are  closely followed so that our strip line impedance will be close at 50 ohm, the thickness, width and the material of the PCB has also a direct influence on  the impedance.

On my prototype I intentionally separate the directional coupler from the meter unit so that when time comes, I can easily swap between other types of directional coupler design. During the initial test, I found that the prototype respond well to VHF while it requires greater amount of rf energy to produce the same full scale deflection of the 100uA meter at HF. Germanium diode used in this project is not linear specially at different power levels therefore this project is best suited for QRP application only.

The face of the meter can be modified so that a calibrated lines can be drawn using a reference dummy load each with different impedance values.  73 de du1vss.

Two Stack Dipole Using Power Divider

My friend ask me If I can make a two stack dipole for his 92.9MHz FM station and I decided to construct this project using a home made coaxial power divider using a two piece quarter-wave long 75 ohm coaxial cable. The two 75 ohm coax was connected in parallel to obtain 37.5 ohm which is the required impedance for our power divider. I also use a 1 wavelength long 50 ohm coax feeding the two dipoles and along with it, a short piece of wire that act as a pawsey stub that makes the dipole feed point balance. The 1 wavelength long 50 ohm coax was also chosen so that each dipole can be vertically spaced at 1 wavelength for gain optimization.

A plastic element insulator hold the dipole segments on a 1"x1" square tube boom. The feed point connection was carefully protected with a layer of rubber tape and on top of it is an electrical tape for UV and water protection.

A  fabricated  bracket  will hold the antenna boom and this was made of a thick plastic material, the one used here was a chopping board 5mm in thick.

A close up view of the coaxial power divider during initial tuning and testing.

A quick frequency sweep was made to evaluate the antenna SWR after the trimming of dipole segments. The antenna bandwidth was quite narrow for this design and I was able to obtain an SWR value of 1:1.2 across 92MHz to 93MHz only.

During testing, the two dipoles were spaced 1 wavelength horizontally and was 1 meter above the ground  just only to evaluate the SWR and tune the antenna while at the ground. When installed permanently, this antenna is expected to yield  3dBd of gain theoretically.    73 de du1vss