Archive for the Electronics Category

My Begali morse key misses some dash

Posted in Electronics on 8 December 2014 by iw0ffk

Recently my Begali “Simplex” key starts to miss some dash during transmission, very annoying. I cleaned the moving contacts with isopropyl and triple checked wires and solders, everything ok but the problem was still present.


The bad point was hidden inside the turret that holds the ball contact with the mobile lever.

I removed the contact and the spring and it’s evident the corrosion on the bolt



Maybe the different metal in contact favors the corrosion of this bolt.

Cleaned the dust gently with a small brush and applied a thin layer of vaseline grease.

Now it’s back on air


Useful cardboard rulers

Posted in Electronics on 11 November 2014 by iw0ffk

I really like smart rulers.

Slide or fixed, circular or rectangular they are all useful gadgets and have a place in my lab.

I’ve found a new one for my collection at the last European Microwave Week 2014 at the Rohde & Schwarz stand.

No pretty frills, it converts at a glance Watts to dBm and to Volt (Z=50 Ohm) and calculate a ratio, voltage or power, in dB


Another useful ruler is made by TK5EP with the great software Galva by F5BU

It’s a circular SWR/Return Loss/ Gain  calculator. Aligning the forward power on the external disk with the reflected power on the internal disk, we can see the SWR and the RL through the little window.


The file to print and cut out is on TK5EP’s website.

Following this link you reach  the Museum of Rulers with many scales.



Restoring a surplus military coaxial switch

Posted in Electronics on 4 November 2014 by iw0ffk

I bought years ago a rugged military electically-operated 2-way 50 Ohm coaxial rotary switch with N-female connectors, model SA-185/U (MIL-S-3928/1A) produced around the year 1965

This switch acts as a (slow) latching relay, the supply voltage is 28V DC with a 3-way connector. If the A-C pins are energized, the common RF input is connected to the first output. The second output is selected energizing the B-C pins. At the end of the commutation, the supply circuit is open, so one can leave the voltage applied without power absorption.

The switching time is about 500-700ms, it’s too slow to be used as a tx/rx relay, for example in a PA.

I don’t have any specs of this object, please write me at my email address, if you know something about it.

In my current configuration I use it for switch the antenna between local (TS-590S + tube PA) and remote (TS2000 + SS PA) 50 MHz stations. Unfortunately, the past few days I am noticing some failed changeover and a good contact is reached only after a few repetitions, so I decided to open the box to solve the problem.

A rotating stepped electromagnet is connected to a barrel that holds the RF contacts. When energized the barrel rotates for 180 degrees with a sound like a “machine gun”.

The problem was some dry grease on the contacts, easily solved with isopropanol and a Dremel with a felt wheel.

Now that it’s fully functional I will do some RF measures and I’ll post here the results.

TCXO “Made in China” for FT817 (FT857-FT897)

Posted in Electronics on 22 October 2014 by iw0ffk

I’ve installed a cheap TCXO on the little Yaesu rig and “measured” the performances. The TCXO has declared stability of 0.1 ppm.

The FT817 will be the IF (432 MHz) of a 3cm CW/SSB transverter, that is still WIP , so an increase stability , even if on the IF, can help to improve the performances of the whole system.

The TCXO comes without board, so it was quickly wired as a dead bug.

This slideshow requires JavaScript.

A kapton tape isolate the case from the pcb, a strong double sided adhesive tape and silver plated wire hold the oscillator firmly in place. Two capacitors, 10uF tantalium and 10nF are placed to the supply as on the original Yaesu TCXO-9 board.


For acquire the data, I used the great software Spectrum Lab by DL4YHF to capture the audio frequency peak of RWM standard frequency and time radio station during the minutes 00:00 to 07:59 and 30:00 to 37:59 of every hour, when the RWM sends a continuous carrier. The rig was tuned on 14.995 MHz USB so the demodulated peak should be (around) 1000Hz. The data have been sampled for about 12 hours: RED is the original oscillator and BLUE is the new TCXO


Negative voltage from LM2596 DC-DC Ebay module

Posted in Electronics on 19 September 2014 by iw0ffk
Some months ago I bought on Ebay 5x DC-DC adjustable switching converter  module for few euros.
As suggested in the datasheet, an additional LC output filter must be used to reach an acceptable voltage ripple.
For an ongoing project I need a source of negative voltage of12V @ 100mA, so I looked for a way to modify one of these modules.
The solution is very simple and it is not necessary to unsolder any component, it is sufficient to use the points of the input and output voltage as specified in the picture below
To obtain a negative voltage, the input/output GND must be connected to the left two pins of the trimmer (OUT+ signal on the PCB silkscreen)
Below the pictures of the -12v output with no load and with a current of 150mA:

The ripple is about 120mV, it can be reduced with a filter on the output, I used an inductor series of 1uH and a 220uF capacitor to GND:

The ripple is now reduced by 5 times, around 25mV also with a drain of 500mA.

NOTE: In the inverting regulator mode, the LM2596 needs at least 2A input for a correct bootstrap, see the datasheet for a delayed-startup trick, if you doesn’t have that current.

NOTE2: of course, if wired in inverting mode, the module should be installed isolated from the GND

Hope this helps someone looking an easy and cheap source of negative voltage with good current.


Adjusting SSB transmitted audio without bore your friends

Posted in Electronics on 5 August 2014 by iw0ffk

New transceivers have dozen of controls for conditioning our voice: mic level, proc level, bandwidth, equalization, etc. Microphones also often have volume and compressor level or they can switch between different caps.

Setting a good transmitted audio without a feedback is almost impossible, indeed many transceivers includes a “Monitor” function for hearing own voice when transmitting, but this is not sufficient because the voice processed by the rig is received in real time, mixing with the voice received directly by the ears.

Often I heard OMs testing on air with a friend in QSO like:

– Now the mic level is 29, proc on 45, EQ TX on bass-boost…how is the sound?

– Ok now the mic is 32, how?

and so on for long time…

The session does not end when the OM reach a good modulation, but when the friend gets tired and closes the QSO with a lie: OK, now it’s PERFECT, don’t touch anything anymore !!

The trick to achieve a good sound quality and timbre modulation as I like, and maybe not to my friends, is to use a tiny software: Audiodelay

This program has been written for synchronize the audio received by a local FM broadcast with the video received via satellite or via internet stream. It’s used to view i.e. a soccer match on tv with the comments of a preferred radio speaker, correcting the lag between audio and video.


For our purposes we need 2 rigs tuned on same QRG, the DUT goes on a dummy load and the receiver audio is connected to a PC with Audio Delay, rf attenuator activated, if needed.
We need to set 3-4 seconds of delay and we can hear our transmitted voice in the headphones after that time, without the QRM of our “not transmitted voice”.

schema audiodelay

For TS-590S I like for dxing, mic on 25, proc (max 10dB) and TX Equalization HB1.


PCB print with 20€ laminator and waxed paper

Posted in Electronics on 1 August 2014 by iw0ffk

After the damages of the last storm, I had to rebuild some circuits. This is a good opportunity to make printed circuit boards instead of the existing prototypal wired matrix board.

So far I printed ​​my pcbs with iron and sheets of glossy magazines, now I used waxed paper and a cheap laminator from Amazon
I use Eagle for drawing schematics and pcbs, but any CAD can do the job. Paths, via and pads must be larger than the default value and appropriate for hand drilling.
Usually I make double side circuits with the Top (or Bottom) layer with only horizontal or vertical wire jumpers, so the copper board is printed only on one face. For the component side I use an Ink Jet Adhesive Transparent Film for the silkscreen (optional) to be placed before the drilling step.
Here the complete procedure
1. Draw and double check the schematic and the routing. Next picture is an example of a mixed SMT and Trough-Hole technology.
In blue is the Bottom layer (wire jumpers). Red and green are traces pads and via to be printed on the copper board.
The SMT boards are printed on the Top layer, the Trough-hole boards on the Bottom. Remember to mirror the layer before printing the Top Layer.
2. Print the correct layers (Top or Bottom, Pads and Vias) to the waxed paper with a Laser Printer at best quality (ink jet are NOT good for this work)
With a piece of masking tape, fix the waxed paper on a paper sheet, then print on it at best quality.
stampa su carta forno
3. Carefully clean the copper board with cotton wool and acetone then stick on it the printed wax paper with the masking tape.
4. Turn on the laminator for maximum heat (position II), wait the warming time, then make 3 passes with the copper board
5. Now gently peel the wax paper and it’s done.
6. Etch the board with Ferric Chloride or Muriatic Acid and Hidrogen Peroxide (be careful…), drill and mount the components. Optionally, before drilling, one can print the silkscreen on adhesive trasparent paper.
It looks good.
 Happy homebrewing!