Luigi Findanno Makes

I’ll not reinvent the wheel, I found this resouce that explain how a L network works as impedance matching, very clear explained: https://en.baltic-labs.com/2013/04/l-network-impedance-matching/

Designing RF Transformer can be scary but I found a simple and fast method to do the job done. At the and of this article you shoud be able to design a real RF transformer for impedance matching application. The know data are: input impedance, output impedance and frequency.

First of all we need a ferrite core. I had no idea how to find the right one, are they categorized? Which characteristic are important? I didn’t know. My priority is ever it must be easy to find. So I found a ferrite that works on my frequency range an I just bought a BN61-302 on ebay, it’s a little binocular ferrite. Now we need the AL value of the ferrite core, this is the inductance factor. On toroids.info I found the AL factor of the BN61-302 is 277 nH/turn^2. It looks complicated but is only the inductance per turn of the selected ferrite core. On toroids.info you can find the characteristics of many easy to find ferrite cores, so it’s a very important resource.

Now we have all the numbers we need:

• Input impedance Zi
• Output impedance Zo
• Frequency f
• AL of ferrite core Al

and we need to calculate the number of primary N1 and secondary N2 turn, so here the magic:

These formulas are an elaboration of other formulas. After calculating and roundind up the N1 value, we can calculate the secondary winding turn N2. If the value of N2 is too unreal like 0,3 you need to increase by 1 the primary winding turn N1 and than recalculate the secondary N2. It will increase the primary inductance, but the impedance ratio will not change.

If you asking yourself where these formulas came from, here there is the explaination. The rule of thumb for wideband transformer design is to keep the inductive reactance about 4 times larger than input impedance Zi. The formulas comes from the following equations.

• L = AL ⋅ N^2
• L = Z / 2pi ⋅ f

Replacing L we get Z / 2pi ⋅ f = AL ⋅ N^2 developing the equation and imposing Z = 4 ⋅ Zi we get the first formula.

From the formula Z1 / Z2 = (N1 / N2)^2 we get the second one.

Reference:
http://www.encyclopedia-magnetica.com/doku.php/al_value
http://www.foxcomputer.se/RF%20Transformer%20Design.pdf
http://toroids.info/

I was looking for a metallic protractor to mount on my DIY antenna rotator and I was sure chinese suppliers was the answer. As I found some interesting objects, thay had some limitations: too big or too expensive. So I decided to design my own protractor and engrave it an aluminum sheet using only free software and my little CNC.
First software: LibreCAD. It’s an open source general purpose 2D CAD and I’m very satisfied, there is not so much to say, it works well and it’s free.
But I was too lazy to design a protractor so I asked to myself “maybe someone did the work for me?” Yes someone already did it, dxf ready.

Next step is generate the gcode file. I spend a lot of time to find the right software. I tried makercam.com but it generate a gcode that not don’t work really good, almost with my CNC machine. The solution is dxf2gcode.

The parameters on the lower left are very important. Here the explaination.

• Z Retraction area: How height goes the tool at start
• Z Safety margin: How height goes the tool when it move to another engraving area.
• Z Workpiece top: Leave 0 (zero)
• Z Infeed depth: How depth goes the tool in the material.
• Z Final mill depth: Desired engraving depth.
• Feed rate XY: XY speed
• Feed rate Z: Z speed

Normally I set the same value to Z Retraction area and Z Safety margin, 2mm is enought. I set the same value to Z Infeed depth and Z Final mill depth to -0.1 mm, it means that the job will needs only one round. If we set Z Infeed depth to -0.05 and Z Final mill depth to -0.1, the job needs two rounds to be done. I set the 2 feed rate to 100 mm/min that is pretty slow but it’s right to get a good job on aluminum. After that, click on the Export menu, Optimize path than Export menu, Export shapes and the gcode file will be generated.

Another interesting software I found is CAMotics, an open souce CNC machine simulator. It reads the gcode file and show the tool path and the motions so we can test the gcode file and see if it’s really ready for the CNC job.

The last free software I use is Candle to control my little CNC. It’s a grbl controller software that sends GCode to the CNC machine.

Engraving is very usefull, the range of applications is really wide, front panels, plates, artistic stuff, etc. The software we need is totally free, my CNC is a 199 euro kit, I bought it essentially to make PCB but it shows good flexibility and I’m very happy with it.

The dish frame parts are got from a 2mm thickness aluminum sheet. The octagon has 62mm side, the long arms are 520mm x 14mm, the short arms 212 x 14 mm, M3 screws are used to fix. The feed’s arms are made from 6 mm diameter aluminum tube, 410mm lenght. One side of the tube is M5 threaded the other is ortogonally perfored with 3mm diameter.

Please read the the images descriptions for other dimensions and details. I think the most important informations are setted out, if not just use the logic and the fantasy.

• 

• 

• 

• 

• 

• 

• 

• 

• 

• 

• 

• 

• 

• 

• 

• 

• 

• 

• 

• 

• 

• 

• 

• 

• 

• 

• 

• 

• 

• 

• 

• 

• 

• 

• 

• 

• 

• 

• 

• 

• 

• 

• 

The simplest way to improve the RF receiving quality is to add a passive filter between the antenna and the receiver. The FM 88 – 108 MHz commercial radios broadcast strong signals and even if we are receiving on others frequency, they contribute to get worse sensitivity to the receiver. A stop band filter can be a good solution to improve the sensitivity but not in all situations and the reason are the condensators. The condensator are the critical passive component when we talking about RF. In the real world there is a point where the inductive components of a capacitor become so important that the capacitor is no more a capacitor. This situation occours after a certain frequency. If we want to receive after this frequency is probably better to not use the filter at all. The following image show how the impedence of a capacitor become inductive after a certain frequency.

Bigger capacitive values suffer more than little. After some investigations I found out if we want approach the GHz range, even the SMD 0603 capacitors can be problematic and the choice of the capacitors must be scrupulous.

The Bias Tee is a simple device that inject DC voltage in the antenna line to power on others devices mounted near the antenna. As the schematic is pretty simple, the choice of the capacitor type should be accurate.

When I designed the downconverter Convertino, I wanted to test it with a concrete application. The ADS-B airplan trasponders send signal at 1090 MHz so this frequency perfectly fit to test this kind of downconverter. After I set the local oscillator of Convertino to 1175 MHz, I could see the ADS-B signal downconverted to 85 MHz (1175-1090=85). The only thing I needed was an ADS-B software decoder with settable carrier frequency. ADSB# is a well knew ADS-B decoder but it doesn’t allow to set the carrier frequency because it is supposed to be always 1090 MHz. I was lucky to find the source code of an old version of ADSB# so using Visual Studio, I modified the program adding a textbox to set the carrier frequency as needed. The source code and the exe file of the modified ADSB# is available here, the exe file is in the ADSBSharp\bin\Debug folder. Using adsbSCOPE with the modded version of ADSB# tuned on 85 MHz, I could see some airplane positions. With this experiment I took the concrete demostration Convertino really work and I got the first data about its RF input sesitivity.

Since spring 2019 I began to play with RF. I had a cheap USB rtl-sdr stick and I would do something intersting. So I found out this page https://www.rtl-sdr.com/about-rtl-sdr/ a list of interesting things a rtl-sdr stick can do. Receiving NOAA weather satellite was very impressive to me so I immediatelly made a 137 MHz V antenna. After I received the first images I found other people on Facebook and Twitter they receive very impressive images from weather satellite, this stimulated me to improve my equipment. So after learning more about RF I succesfully made: a 88-108 bandstop filter, a 137 MHz band pass filter, a QFH antenna using an online calculator, a 137 MHz antenna preamplifier I designed. With my new tools I received the better images ever. Than I found out some weather satellite send better quality images using other frequency and protocols. I’m talking about 1700 MHz and I never played with microwave. So I focused myself only on teory and learning what other people do. Well, to receive L band signal I need at least: a dish with a tuned LNB and a downconverter. So I designed an L band downconverter. It was not easy I had to study a lot of stuff. Before the components choice I learned how to design an RF PCB. The component choice was challenging. My first approach was an IC for every function: PLL synthetizer, mixer, band pass filters. But my experience with microwave was to zero and a lot of component means a lot of uncertainty. Suddently I found out an IC the MAX2121 a complete Direct-Conversion L-Band tuner. This IC is a complete IQ downconverter, more that I need but it doesn’t need others RF parts. Basically using only one of the outputs (I or Q) my purpose is solved. So I designed my first 4 layer PCB, ordered all the components, assembled the circuit and it works! As demo application I succesfully decoded the 1090 MHz ADS-B signals downconverted to 85 MHz. I will update this blog with all my projects, stay tuned.