Automating a manual antenna tuner

MFJ Versatuner with servo’s, a stepper and a Raspberry Pi

Automatic antenna tuners are expensive, manual antenna tuners are slow. So it has been a long time I dreamed of making my own automatic system.

As this year I was following a raspberry pi class and we needed to make our own project I decided to finally try to make this automatic tuner.

Summary:
– Forward and reflected power are measured with a Boxa Swr bidirectional coupler combined with an ADC
– SWR is calculated in the raspberry pi and displayed on an LCD
– When SWR is to high an alarm is sounded in CW
– after pushing a button the capacitors and the coil in the tuner are reset to 0. For the capacitors I use 2 servos, for the coil I use a stepper motor. All mechanical connections are made using 3d-printed parts. The reference point for the stepper motor is made using an infrared port.
– next is the tuning operation:
— the ‘transmitter’ capacitor is sweeped over it’s full range and swr is measured every degree. Then this capacitor is returned to it’s best position
— next idem for the ‘antenna’ capacitor
— sweeping operation is repeated for maximum 4 times, if the SWR is below 1,5:1 then tuning is stopped. If SWR is above this limit then the coil is moved to the next setting and the sweeping is repeated
– built in securities:
— all alarms on the LCD and in morse code
— alarm when SWR is to high
— alarm when tx-power is to high for tuning
— coil/stepper can not be switched when there is rf-power

Remarks:
– no tuners have been harmed in any way for this operation!
– Was the raspberry pi the best choice for this? This could probably have been done using some microcontroller instead but the class I followed wat not arduino or esp 32 πŸ™‚

A 30w dummy load really can’t handle 100w

Or at least that’s what I found out a few days ago… I was doing some tests on a bidirectional coupler (more about that project in a future post) and I didn’t pay enough attention so I cranked up the power of my transmitter to 100w. And, of course, after a few seconds I got the smell of burned electronics in my shack.
That’s when I decided to go through my junkboxes and find the resistor I once bought to build myself a decent dummyload.
These are the results:

The resistor is an RFP 250-50 TC so this should handle up to about 250w.
SWR between 160m and 6m is max 1.04:1. I have not tested any higher frequencies because my antenna analyser only goes up to 54 MHz.
The heatsink is salvaged from an old telephone switchboard (made by Bell Telephone in 1980).

Antenna switch quality?

What does a ham do when the weather is bad and there is nothingDaiwa CS-201 exciting going on on the bands?

For some tests I needed a coax switch and I found my old Daiwa CS-201 in my drawer. I vaguely remembered there was something wrong with it the last time that I used it.

As soon as I connected a coaxial cable I noticed one of the N-type connectors on the switch was loose. So it was time to get out the screwdriver.

This is what I noticed:

inside Look at the contact at the red arrow! Due to the top right coax connector rotating when connecting a cable the contact was barely touching the post it should touch. Not visible on the picture was that the bottom contact had the same problem but of course this was rotated in the other direction.

And, this is a switch rated for 1 kW!

I could pretty easily fix the problem by unscrewing the little black screws and rotating the coax connectors back to where they were supposed to be. But I guess this will only last until I have connected and disconnected the coaxial cables a few times again.

Soon back on the air!

Finally my OB11-3 and new rotator made it up the tower.

For now no QSOs have been made because of course there’s still a lot of work to do like running new coax cables and the rotator control cable up to the first floor shack and putting the wires for 30-40-80 back up.

Time permitting I’ll try to be on the air in a week or 2 max. I’d like to test in the DMC Rtty contest in 2 weeks from now.

Here are some pictures:

Tower without antennas
My tower, like a virgin!

going up
Going up!

ready to turn
OB11-3 ready to turn

almost there
OB11-3 almost there

finished
Finished! On top you can also see my Fritzel UFB-12 (12 & 17m Rotatable dipole)

Finished
Ready to get some RF flowing!

I wish to thank Tom from Optibeam for the advice and help provided in choosing the ideal antenna configuration for my needs!

End of this contest…

In the last 2 weeks I have been working hard to get my band decoder up and running and to prepare for next contest: the ARRL RTTY Roundup (in which I placed first in M/S LP non W/VE in 2013).

Yesterday I made some QSOs and everything worked all right.

This morning I click on a dxspot on 12m and have my logging program turn the rotator… a few seconds later: no more signal on the receiver.

Something has gone wrong on the Big Boy Rotator and, of course, before I noticed it had overturned and broken the coax cables for my WARC rotary dipole.

The 10-15-20 beam is working OK and the rotator seems to turn but with no correct indication of the direction. I guess this contest has already ended for me this time.

Next big RTTY contest on my list is WPX in a months time…

So, todo:
– find someone to climb the 15m tower twice
– fix the rotator or find a new one
– fix the broken coax

Anyone some money to spare to help me buy a SteppIR DB-18E so I can replace the good old FB-33 while we’re climbing the tower anyhow?

Inrad HF Triplexer

My new contest secret weapon worked very well… for exactly 1 QSO.

For the moment I have only these (permanent) antennas for contesting: a Fritzel FB33 3-element tribander at 16m and homemade inverted V dipoles for 40 and 80m at 15m.

So for SO2R I have always used an extra Fritzel GPA404 vertical for 10-15-20-40m. But to avoid the hassle of installing this antenna before the contest and taking it down after the contest I recently decided to purchase an Inrad HF Triplexer so I could use the FB33 on 2 bands simultaneously.

Last week I had the specs of the Triplexer tested by Luc ON7KZ and all seemed to match the specs as in the Inrad web page.

I installed the triplexer between my Sixpak and the FB33 the day before the contest but I didn’t have enough time to thoroughly test it. (I did check all connections twice though!)

Once the SCC Rtty championship started I made a S&P QSO on 20m and during the QSO I saw the SWR on 20m going up to alarming levels :-(.

Continue reading Inrad HF Triplexer

PL259 made in China

Today I installed a few PL259 connectors of this type:

PL 259 Connector (picture from Wikimedia)

I should have known better… the ‘barrel’ is too long so the plugs can not be tightened on most SO239 sockets… Only 25% of the price of a ‘European-made’ (???) plug but what are they really worth?

5 days left to find a good solution before next contest (SCC RTTY Championship).

I guess I’ll have to try and file/sand off some material and see if these connectors can be saved.

Again a lesson learned!

Most of the time I use N-connectors for everything, even for frequencies below 30 MHz but of course there is still some gear that is manufactured with SO239.

=> update: I grinded down the barrel 1,5mm. Now the barrel is ok but the pin is too long… {SIGH}

My current project

Ever tried to manually switch your antennas and bandpass filters 30 hours into a contest? I have been doing it for years. πŸ™‚ (And I have been complaining about it for ages too)

But the solution is near:

band decoder 001 band decoder 002

This is the prototype for my new universal band decoder.

The hardware is an Arduino (in the pictures is a Mega, it will be replaced by a UNO r3 in the finished product) and 2 4×4 Driver Shields from Logos Electromechanical.

The goal is to switch my SixPak and my 2 Dunestar 600 band pass filters for SO2R.

The data comes from the pc (over the usb port) using the Open Two Radio Switching Protocol (OTRSP) which is supported by N1MM and DXLab logging software.

For now this is a work in progress,Β  so come back here soon to follow the evolution!

Writer’s block?

A lot of things have happened since my last post here but I didn’t really feel like writing anything about them lately.

A small summary:

  • we did the Region 1 SSB fieldday with the club station ON6CK/p (and came in second behind the OQ8A/p team)
  • the belgian guys conquered Rockall, congrats to the team!
  • the british guys decided not to go to Rockall. I don’t really understand the reasons, if all those who wanted to climb the Everest would have cancelled their plans because someone was there before them then only a few people would have done it. Moreover the pileups from Rockall will not be less for the second team going there than for the first one. And I still need EU189!
  • I finally ordered a new transceiver!
  • I modified my RTTY setup to do FSK instead of AFSK (with the FT990)
  • I made a portable setup to take on holiday

The last 3 of the list will probably be the subjects of some future posts!

So what makes me write this in the last hours of CQ WW SSB? I’m at the qth of my girlfriend and to tired to set up the portable station. Wedding parties are probably more tiring than 24hrs of contesting πŸ™‚ Moreover we have to leave for an other party in less than an hour.

My next contest is in 3 weeks from now… still some work to do (the 4 rotary switches problem, remember?)

 

Designing antennas using a genetic algorithm?

A few days ago I found a nice piece of software on the internet: BoxCar 2D. This is a way to design a car using a genetic algorithm: a population of cars is created at random and then tested on a track chosen by the user. Then the best cars ‘mate’ to get a next generation. Together with a set mutation rate the cars improve over the generations.

Warning: watching this is addictive, it’s far better than television!

My programming skills are rather limited but I wondered if someone maybe could design antennas using one of the (mini)NEC engines and this genetic system?

Update: it seems that this is not such a new idea, at least for developing antennas. A little googling for ‘genetic algorithm antenna’ gave me some 160000 results πŸ™‚

SuperNec seems to be great software for this but at $6080 it seems I’ll stick to Eznec and MMANA for a while.