ad: L-HROutlet
ad: l-rl
ad: Subscribe
F4HFS France flag France

Login is required for additional detail.


Email: Login required to view

Premium Subscriber Lookups: 2122

I've got my license at 16, in 1974, thank to the F6KAW radio-club. Then I was F6KAW operator for 2 years.

I met a young lady who became my wife and I gave up with radio-ham until 2013. I successfully fought against a cancer and I realized that I should not keep any regret of any kind. So I recalled the old radio-ham memories. I asked for a Call Sign, and I began building.

For me Radio-ham is not just making QSO's... It is experiencing radio. I want to share my time between hamming and making. But so many things changed in 40 years...

I began by building a SoftRock 3 bands Trx and a 20W PA. I was a little disappointed because the internal filters were not well working and the transmitted signals don't match the FCC and the French rules. Mainly the local oscillator power is -40 dBc, while it should be < -50. I understood that I should get a commercial Trx that I would not modify to keep it working, while I can play making, building something else. I bought an Yaesu FT-897D, and a Genesis G11 all bands.

By the meantime I began to build an antenna. I live in a place where I can't get an external antenna. So, I built a Magnetic Loop in my attic. I wanted it fully automated, and I first wanted to build everything myself. It has been a 3 years work. I just finished the first truly working version today. I still have some work to do for improvements, but it's usable.

I built my own digital interface to connect on a sound card and I began some QSOs in digital modes.

By the use of the SoftRock receiver I saw how fun it was to see a large part of the radio band. But now, using the FT-897D and FLDigi, I just see a 2.5 KHz bandwith anf it's frustrating.

I have other projects : a 300 W PA. As the antenna is just above our heads, 300 W is largely enough; a 30-40m magloop; a second 10-20m magloop to use in array with the first one. I expect to earn at least 3 dB. and I will buy a direct sampling SDR Trx : the Icom 7300 or the Elad SDR-DUO. I did not choose yet.

June 08 2016

I just finished my Genesis G11 and I made 2 digital QSOs. A 10W PA with a magnetic loop in the attic don't make it easy to be heard. I surely have to work the 300W PA.

If you read this far, you know a bit more about my radio-ham life. Perhaps will we meet in a QSO ? Or perhaps it is already done ! Best 73, Alain

March, 13, 2017

I made several mistakes building the 300W PA and changed several times the PA FET pair :)... I grown impatient to be heard, so I eventually bought a RM Italy BLA 350 and I don't regret it. It seems very good. I just put my home made PA aside and I will take it back later on. I really want to finish it... let's see...

Now that I've got a working PA, I just love using my G11 SDR Trx. You really can see a wide scan of the band, what's going on, pile ups. As my antennas are omni-directional, the magic features are the software variable filters. They are sharp as knives and you can really isolate the signal you want, not being polluted by another strong signal just beside. I'm making roughly 80 to 100 QSO a month, mainly on week ends like many of us, I guess, for those who have to work for a living. Therefore, my FT897D is quite silent. I plan to use it soon on portable operations with an MFJ-1625 as a portable antenna... Next year I'll buy an IC7610. It looks really promizing.

I made my second Mag Loop for 20 to 40 meters. It is 2 meters diameter and it's much better on 20 m than my previous 10-20 m. The difference is really seeable. I was able to contact Indonesia, Abu Dhabi, the Dominican Republic, Canada, USA, and others with this loop at 7 meters height in my attic... I kept the first one and I "park" it on 25 Mhz when I am hamming on 20 or 40 meters with the "big" one, and so, the co-existence seems OK. I don't think I will ever be able to contact the pacific. Perhaps after the 2020 propagation gap ?

I improved my 2 Mag Loops by suppressing the network cable which I used to remotely pilot the capacitors. I replaced it by a 2.4 GHz wifi module and the noise just reduced by 5 to 6 dB ! Ethernet network cables and switches are really noisy ! The wifi modules just don't care about 300 W HF !

That's it for today. Next time I'll add some shots.

73, C U soon on my SDR screen ! Alain F4HFS

November, 14th, 2017

I finally stopped using the G11. It had a strange behavior in transmission. Idon't know if it was from the Asus STXII sound card or the G11 itself. I am using an RS-HFIQ from HobbyPCB.com instead with an external Sound Blaster XFi HD. The performances are not lower and it's more reliable. I probably got/done something wrong with the G11.

Here is a complete description of my station.

Diagram 1

I'm going to explain all the parts of the diagram 1. But it is only about hardware and some important parts are software, too. The software organization is described in the diagrams 2 and 3.

Diagram 2 : The Software organization for the Yaesu FT-897D

Diagram 3 : The software organization for the RS-HFIQ

The Magnetic Loops

I wanted my antennas automatic. So I just chose the frequency with the TRx and the antenna automatically follows. I almost got it.

The learning took me 3 years. I learned how to simulate a magnetic loop with an antenna simulator and then I saw that the performances are much higher with an horizontal loop, and the horizontal magnetic loop is quite omni-directional. I did not want to use a rotative antenna in the attic. So the choice was obvious for a horizontal magnetic loop.

I made several attempts, with air capacitors, with barbecue reducers and step motors, and I saw that the tuning was very sensitive. It does change with some hundredths of pico-farad. The barbecue reducers have got big loose positioning, going one way or the other. I experimented the PCB capacitors :they are only for receiving antennas. The HF power make them heat up to burn. Eventually, I ended up to the Variable Vacuum Capacitor (VVC). The capacitor is the very more delicate piece of the magnetic loop. I took a VVC from the USSR army, that you can buy on ebay at roughly $100. By this way you solve all the capacitor problems : the reducer, the precise capacitor, the voltage ability...

Then I studied the environment influence on the antenna tuning. If you got a VVC, you just have to consider temperature and atmospheric pressure. If you are using an air capacitor, you should consider the humidity as well. I built a thermal cabinet around the antenna in my attic with a thermal regulation. With that, I studied what the relation is between the temperature and the tuning frequency. Of course, the device was totally dependent on the atmospheric pressure. But I was able to move the temperature much quicker than the atmospheric pressure moved. So, globally I did not consider the atmospheric pressure. I won't detail it here, but my conclusions are that the tuning frequency varies linearly with the temperature. Yeah, I know : I am lucky.

Once this was set, I had to calibrate the antennas on all the bands, based on this simple variation law.

So what is the principle of the magnetic loop management ? With the antenna, you get the step motor control and a thermometer. Of course you have to use the same thermometer for the antenna calibration and the antenna use. The magnetic loop manager is a piece of software that calculates the capacitor position, based on the TRx frequency and the antenna temperature. That's it. Almost easy...

Ok, I wrote that I almost got the full automatic antenna. Based on the atmospheric pressure, the capacitor tuning made by the magnetic loop manager might be slightly wrong. Good or accurate enough for receiving, though. So, when you make a big frequency change, you have a "fine tune" button in the magnetic loop manager that you use just before transmitting. The fine tune can be made manually or automatically. It's reliable. If you don't do the fine tuning, the SWR is under 1.5, so not that bad.

Let's get to the impedance adaptation. With a magnetic loop, you may use 3 different couplings : an induction loop, a toroïd coupling, and a gamma match. The induction loop seemed easier to build. It's a symmetric coupling connected to an asymmetric feeder. So I added a choke ( the famous 8 8 doughnut : 8 turns of 8'' diameter of the feeder coaxial ) to avoid getting the coaxial feeder a part of the antenna. By this way, the induction loop and the main loop are a kind of HF transformer. When you use the loop on a wide range of frequencies (from C=lambda/4 to C=lambda/8, where C is the loop circumference and lambda is the wave length), the impedance of the main loop, at its middle, varies from 2 ohms to 0,1 ohm. So does the impedance of the induction loop in the same proportions. No way to get 50 Ohms on the wide range. So you have to chose where you want to get your 50 ohms. You adjust the induction loop on the chosen band and then you use an antenna tuner for the others. If you are lucky enough, you might get several bands with a good SWR. I tuned the 20-10 meters loop to get a good SWR for 15 meters, 12 meters and the beginning of the 10 meters band. For the 40 - 20 meters loop, I choose the 50 Ohms induction loop on 40 meters. And it works pretty good. What about the antenna tuner ?

Picture 1 : The two antennas in the attic

The antenna tuner

Not much to say or write. It's a classic MFJ-941E. I just attached the relay on the box instead of the TRx SO239. So, the TRx antennas and the SWR meter device are connected to the antenna tuner/relay. Thus, the relay is not loose on the table.

I don't know if an automated antenna tuner could be possible as the capacitor tuning is not always exactly the same, based on the temperature and the atmospheric pressure. It could be giving some slightly different tunings but perhaps still good enough. I've got an AT-897 Plus from LDG with my FT-897D so I'll give it a try.

The SWR-meter

I made the SWR meter from an MFJ-213. I suppressed all the buttons, the display card, I took it off from the box, and I took the processor off as well. I connected a VM167 from Velleman instead to pilot the AD9851 DDS and to read the forward and reflected powers, both real and complex. It was working quite good but the VM167 was a noisy tool and Velleman stopped it. I took the VM167 off and I replaced it by an arduino nano and the measurement was much more stable, but I never was able to pilot the AD9851. I finally gave up, I suppressed the onboard DDS generator, and I added an external HF generator based on a AD9854 DDS. It's a complete device bought on ebay in China for around $60. The SWR circuitry of the MFJ-213 is designed for a 0 dBm output power and the generator gives -1.17 dBm. So, sometimes the SWR lacks some accuracy. I plan to make a simple amplifier based on a 2N2222, and, later, retry with another MFJ-213. I also saw other designs from the FoxDelta team or from Rig Expert that look promising and much cheaper. I will have to look at them.

The Generator - SWR meter is a "black" box, blind, only with PSU and USB cables. You mandatory need software with it. I made the software on a windows PC. It could be a notebook, but I don't know how to write a software for IOS or Androïd. Moreover, I just don't want. I don't stand these small screens, even worse when they are LCD displays with some few colors and few pixels. Ok, the small tools are easier to climb an antenna tower, but I don't have such for now. I made 5 programs :

  • a complete HF generator with all the features of the AD9854
  • a SWR meter
  • a magnetic loop antenna analyzer
  • a (classic) antenna analyzer
  • and of course the magnetic loop manager that I describe further.

The magnetic loop analyzer is a classic antenna analyzer plus the variable capacitor control so you can draw the line of the lowest SWR for all the capacitor positions you want.

Then from the SWR meter we go to the TRxs, through the manual antenna switch. It's 3 positions switch as I have already planed the next ELAD FDM duo.

The Yaesu FT-897D

Not much to write on this one either. I installed and used a YF122CN filter and after one year it was broken. I took it off and discovered that using the embedded DSP filter was quite easy and much more efficient, specifically for Digital.

The PA RM Italy BLA 350

I just made a simple and classic change in the PA to get 2 separated RX and TX chains.

I have cut the track between the two Normally Closed contacts of the Tx relay and connected a RG174 coaxial terminated by a BNC socket. for each of the two contacts. If you bind the two BNC, then you can normally operate the PA. If you don't, the "TRx" SO239 becomes the "Tx" SO239 and one of the 2 additional BNC is the "RX" socket.


As I wrote, this small Trx has performances that could compete much more expensive rigs. I made some modifications (originally prepared on the board, this is a good design !) to get it working my way :

  • I took off the SMC leds to put 3mm ones, with leads, on the front plate. I attached them with some silicon mastic.
  • I made the main HF BNC the Tx output. I installed a female SMA planned on the board for an external RF. I cut the board straps for the RX circuit and connected it to the SMA by a thin 50 ohms coaxial, RG174.
  • I used the J3 aux for the TX commands for both the driver and the external PA
  • In the original design, the box is the heatsink. As it is not anymore, I glued a big piece of aluminum heat sink on the final HF transistor.

With this hardware modifications, I had to make some software modifications as well in the software within the arduino :

  • I suppressed all the code about the synchronization with a HardRock 50.
  • I added some code to get the TX command on both the two outputs on the J3 aux.
  • I don't know why, but I had to change the clipping values. Some were too low.

That's it for the TRx itself.

With this kind of SDR technology, the main problems are the unwanted signals : the image signal and the local oscillator. With the RS-HFIQ it's not a problem. You can tune your Tx to get the LO at -50dBc and the image signal at -80dBc. Of course it depends a lot on the TRx software. I am using the famous HDSDR. It's both simple and intuitive, and featurefull.  With HDSDR you have to split a total band in several pieces and have a tune set for every band split. For instance, the 40m band is split in 2 pieces, though the 10 meter is split in 14.

With HDSDR + RS-HFIQ you can adjust the output power level. You can get the LO at -50 dBc if the ouptu power is at 5 W. For my driver, I don't need 5 W out, but 0.5. If you decrease the power output, the LO stays at the same level. So, at 0.5 W, the LO is at -40 dBc instead of -50. As the aim is to get to an external PA, between 300 and 600 W in my case, -40 dBc at 500 W makes a LO at 50 mW. So instead of decreasing the output power by the Tx software, I made it by a 9 dB attenuator, thus the unwanted LO stays at -50dBc !

Picture 2 : the RS-HFIQ set

Diagram 4 : the RS-HFIQ set

As the RS-HFIQ is fully filtered, Tx and Rx, and the PA (RM Italy BLA350) as well, I did not put any LPF in the driver. It is made after a design from Andy Hunter G6LBQ.

Diagram 5 : the 15 W driver

The sound cards

I am using 3 different sound cards :

  • external USB Sound Blaster XFi HD, model SB1240 for connecting the SDR RS-HFIQ
  • Internal PCI-Express Asus Xonar DG SI for connecting the digital signals from the FT-897D
  • external Sound Blaster X G5 for the audio in (micro) and out (headset and Loud Speakers)

For connecting an "IF" SDR, you must get a very good Sound Card : it is the heart of your receiver. You want a low noise level ? So must be your sound card. I first tried an internal Asus STXII that is, theoretically, the best sound card for this kind of use. But the card installed within the PC got a small humming noise that made decoding digital signals impossible. Then, the best I found was the SB XFi HD at a much more reasonable price (around $75). I'll try the STXII in an external USB/PCI express adapter. Of course, before that, I've got a good PC PSU from Corsair.

I tried to get 2 SB XFi HD. One for the SDR TRx and one for the digital signals from the FT-897D. Not possible. The driver is made so that you cannot get the 2 same cards on the same PC. It just does not work.

Decoding the audio digital signals ( 100 to 3000 Hz with the FT-897D) does not require a very good sound card. The Asus Xonar DG SI seems to be doing a good job at it.

For the true audio, I am using the SB X G5, just because I have it. I hoped that the software equalizer could be used in input, with the microphone, but no, it's only for output. So for compressing my audio input, I plan to get an MFJ-654.

You don't want to mess up with the sampling rates. As the audio is quite complex with the 3 sound cards + the virtual audio cables, you have to set all the sampling rates at 96 KHz. Thus you never get any incompatibility between your different audio devices and programs, and it gives you a 96 KHz bandwidth for the SDR operation; it's quite enough.

Software and programs

For Digital decoding and traffic management I am using the classic HRD suite.

For the virtual serial links I am using the very good VSP and VSP Manager by K5FR.

For the virtual audio cables, I am using the VAC from VB Audio Software.

For SDR operation, I am using HDSDR coupled to omnirig for piloting the RS-HFIQ.

I won't make any comment on these good and reliable programs, well known in the radio-ham world.

Then I need to get my 2 magnetic loop antennas as automatic as possible. For that, I am using my magnetic loop manager.

The magnetic loop manager

Operating the magnetic loop manager, you first have to choose the magnetic loop to operate. Then the software connects to the related control module via the Wifi network. It reads the antenna temperature, the controller version given by the arduino, and the capacitor position kept in the arduino memory.

You can choose the rig interface you prefer : HRD or Winfldigi.

Then, the software reads the TRx frequency, and position itself in manual mode.

When in manual mode, you can type the frequency you want and ask the antenna for tuning on this frequency. The software calculates the target capacitor position, based on the antenna temperature and ask the controller for this position. You can ask for a temperature refresh, for a "manual" capacitor move, for a SWR measurement, or for a manual or an automated tuning based on the SWR measurement.

When in automatic mode, the software permanently pools the rig interface (HRD or Winfldigi) to get the TRx frequency and automatically changes the capacitor position accordingly. When I am operating the FT-897D, I do use the automatic mode, though when I am operating the SDR RS-HFIQ, I do use the manual mode because you can jump from a frequency to another just as you see the signals and it could be a bit "jerky"... So I first choose the station I want to work, and then I turn the magnetic loop manager in automated mode.

With the magnetic loop manager you also can command the audio switching device.


At 100 meters from the house there is a powerplant transformer. This gives me a noise at S 5. Otherwise I would have better results.

However, I'm able to contact all the European countries and I even have got some interesting confirmed contacts : Azores Islands, Canada, Dominican republic, Gabon, India, Indonesia, Puerto Rico, Asiatic Russia, Lebanon, Israel, Saudi Arabia, The Gambia, and the farest in USA was the Montana. They are not very far DX, but let's say middle range ? Not that bad for a piece of copper pipe in the attic hugh ?

Let me know if you are interested and I could give much more details about the magnetic loop building and about the programs I made.

8531574 Last modified: 2017-12-23 20:36:00, 31071 bytes

Login Required

Login is required for additional detail.

You must be logged in to file a report on this page

Please login now...

Public Logbook data is temporarily not available for this user
ad: giga-db
Copyright © 2018 by QRZ.COM
Wed Jan 17 18:09:35 2018 UTC
CPU: 0.063 sec 87479 bytes mp