Username

Password

Remember me
Password Reminder
No account yet? Create one
Main Menu
Accueil - Home
Articles
DX cluster
Photo Gallery
Forum
Liens - Links
Contactez-nous - Contact us


TX4T
We have 258 guests and 1 member online:
  • va2wdq

wrtc2018logo.png



rigexpert.jpg

Accueil - Home arrow Articles arrow Articles arrow RX ANT Keyboard 4 x 8 plus 8
RX ANT Keyboard 4 x 8 plus 8 Print E-mail
Contributed by VE2TZT   
As you can have seen I was not mush present on the CGQ site those last weeks. Following is a part of the reasons:

At last, during the transition between the two years, I have decided to allow some time to the learning and the setup of a complete electronic design and realisation process. It includes PCB design software, a modern personal PCB etching process and metal box making.

PROCESS

-For PCB design, I have chosen KIKAD because it is free and not too complex as for example ORCAD can be. With KIKAD, it is easy and fast to do add one’s own components for schematic as well as for routing. The rules for the size of vias and tracks are easy to modify.

It is the perfect tool for the needs of the average circuit designer that I am.

-For the etching process, I have chosen to try Pulsar tonner transfer system. At the very first beginning, I was somewhat disappointed by the result, I was not able to correctly transfer the tonner on the copper board. After 2 bad tries, I discovered that the quality of the tonner is very important and very variable from a brand or a printer make to another one. Another discovering was that the Pulsar Website is very optimistic on how fast you can manufacture a board they are claiming 7 minutes of total etching process from printing to etched board where I need around two hours... For example it is said on the site that we need only one pass through the laminator and a second by security where I need ten passes to have a good result.

I is true that this process is better designed for small boards and mine where middle sized.

After the learning phase, I must admit that it works well and that we can obtain very thin tracks.

The big default of this Pulsar process is the difficulty to have a good accuracy in the registration of two sided circuits. That default comes from the fact that we must put the printed picture one the copper board without seeing it. Then it is very difficult without industrial machine to position the second side in front of the first one with more than 0.5 mm (20 mils) of accuracy.

This lake of accuracy is combining with a certain amount of X,Y scale reproducibility distortion coming from the printer and from the laminator. I am suspecting the laminator to make some distortion by spreading not evenly the toner on the copper board while pressuring it against the board. That fact must be taken into account when designing the PCB.

-It is very frustrating to see how poor is the choice of boxes on the market and how expensive they are compared to the cost of aluminum sheets, so I have invested in a multifunction machine which will be paid back at the tenth box. I have bought a 12’’ Metal machine 3 in 1 : Shear, brake and roll from Craftex. This machine was advised to me last year by a great electronics and radio amateur, Jacques VE2EMM who sadly disappeared in 2010 after a long illness.

FUNCTIONNALITIES

To test this whole process while learning it, I decided to build the first stone of my low band antenna system project that is the direction and switching control keyboard. The idea is to have one the same keyboard 4 sets of 8 keys controlling the 8 directions of 4 antennas and one set of push on/off keys for controlling the switching or various mixing, swapping and diagnostic functions. Using a PIC in that project will permit to modify ‘’in the minute’’ the functionalities when needed, the imagination being the only limit.

Each key has it’s led lighted if pushed or selected. The outputs of the keyboard will permit different choices to control the future antenna system : 20 binary open collector sink outputs plus one TX/RX serial output with separate RS232 (for a computer connection) and RS422-485 plugs. The 20 binary output are for 3bits X 4 antennas (3bits=8 directions)+8 push-button output.

The RS422 is a symmetric electrical protocol which permits 1000 feet twisted pair (Cat5) serial transmissions on a multipoint bus in noisy environments. The idea is to dialog with the antenna controllers located at the feet of the antennas with only one twisted pair of CAT 5 wire. The advantage is lower command wire cost, possibility of distant diagnostic and of course, more plasticity in the antenna pattern.

REALISATION

I have designed and realised everything including the metal box, I have attached some pictures that are speaking by themselves.

Just a point: Half of the array of the processor board is for protections: shorts of outputs, reverse voltage of power input and of course RF protection: each output including serial one has its choc coil and capacitor, even the 12V power inputs (plus and minus), and all the common (ground) outputs have they 1mH choc: I will not want the common mode to travel from one antenna controller to the other and by the way to the antenna radials etc...We are on 160m. This way, the PCB ground is connected to the metal box at only one point and the metal box itself is connected to the station ground at only one point everything else is as RF isolated as possible.

Gilles VA2EW VE2TZT



Powered by Online-Montreal