Hacking the Spectrum Analyzer Siglent SSA 3021X

Posted in Electronics on 29 January 2017 by iw0ffk

ssa3021Last year  in the laboratory where I work we bought two Siglent SSA 3021X 2.1 GHz Spectrum Analyzer for generic/every day use. It’s a nice cheap object with all the functions of a modern digital instrument: colors, memory and math for traces, ethernet access, usb drive support etc.
After a while I found some bug in the software (v 7.03), as example if a screenshot image is saved in the internal memory with spaces in the filename, then it’s impossible to delete….
So I googled for a firmware update and casually found a thread on EEVBlog’s Forum where a guy shared the root’s user/password to log in the instrument with a telnet client. :-)

The big brother of the SSA3021X is the SSA3032X 3.2 GHz SA that has the same hardware of the SSA3021X. The instrument works on Linux, Busybox is running, and it’s an easy job to edit a configuration file and upgrade the SSA3021X to SS3032X with all the options enabled (tracking generator included) and extra BW filters 1Hz and 3MHz (!!!)

This works with installed firmware version 7.07, probably Siglent has corrected this bug on the newer firmware releases, but for sure I don’t need further updates on these units…



10GHz update + first “DX”

Posted in VHF on 15 December 2016 by iw0ffk

The annoying difference of frequency between TX and RX has been solved by stabilizing the supply of the local oscillator with a DC-DC converter. There was a small voltage drop due to the high current of the TX-RX coaxial relays (600 mA…)

First test with Silvano I0LVA, QRB about 50 km easily worked in SSB also with minimum power (50mW).

The surprise was during the Vecchiacchi Memorial Day of 4 December, Cesare I6XCK in JN63QO QRB  of about 215 km with the mountains in the middle, worked in CW directly, no RS. I received I6XCK with a clear 529 for the whole duration of the sked, but I had to wait a bit and repeat the report/number/locator several times to complete the QSO probably due to about 10dB of difference of RF pwr. I think a PA is needed…


I’m always free for any reasonable sked on 3cm.

10 GHz, state of play

Posted in VHF with tags on 17 November 2016 by iw0ffk

Over the past few months I acquired some parts needed to complete the Eyal Gal transverter. I had to build the filter placed just after the TX mixer that has been done by using 15mm pipe caps.  The local oscillator feeds the two mixers (TX and RX) with a 6dB directional coupler and a buffer amplifier.

The sequencer has been made with a simple circuit (IK5ZWU) and adapted to the Eyal Gal. The signal “TX Inhibit” is used to avoid any RF coming from the FT-817 during the changeover. In RX the Eyal Gal’s amplifier is disabled and his gain is set to the minimum.



The “ALC” level sets the gain of the output amplifier. With the FT817 set to maximum output power on 432 MHz, about 4 Watts (total input attenuation is 37dB), the trimmer R11 is regulated to obtain little more than 1 Watt on 10G. By lowering the FT817 driving power to 2W, 1W, 500mW, the 10G power will be lowered to 600, 250 and 50 mW

Up to now I had only one SSB QSO, with IK0EQJ, and I discovered an annoying difference of some kHz between the RX and TX frequency. Probably I have to separate the power supply of the local oscillator from the transverter.

First steps on 3cm band

Posted in VHF on 16 September 2016 by iw0ffk

Two years ago in 2014 I bought from an Israeli surplus store an Eyal-Gal unit, model 6075 with range 10.0 – 10.85 GHz. This module has a LNA and a mixer for the RX path and an amplifier with huge gain and discrete power for the TX. It needs few other parts to become a complete transverter: local oscillator, mixer, bp filter and two relays for the TX/RX changeover. For the local oscillator I used a 100 MHz OCXO with a crystal at 99.36 MHz, manufactured by Eisch-Kafka Electronics, as a reference for a PLL-DRO to obtain 9936 MHz. This permits to convert 10368 MHz to 432 MHz and use the FT817 as IF.

Here the module tested in RX and TX in the LTG lab:


Finally this summer I had the time to install the dish on the roof, an old tv-sat 85cm dish and a feedhorn.

I still need some parts to complete the TX section, but at least I can receive and testing the band.

For now I received 4 beacons directly IQ0RM/B JN61fw, IW0REF/B JN62hj, IQ0AH/B JN40qw, IT9CIT/B JM67lx and IQ5FI/B in rain scatter.

Kenwood TS-590 official ALC mod

Posted in Electronics with tags on 12 September 2016 by iw0ffk


Thanks to the yahoogroup of this transceiver, finally is available the mod to eliminate the dangerous power spike generated at the start of every transmission. The power spike can reach two times the power selected for few milliseconds, in the best-case this can trigger the input protections of several linear amplifiers.

Kenwood performs the mod for free also on the units out of warranty, but here in Italy the official Kenwood lab has asked me the purchase bill. I don’t have that bill because my rig was bought used from another OM and it already was out of warranty.

Here the doc: ts590-alc-fix

Picture from DK5TX website



Sporadic-E summer 2016

Posted in VHF on 17 July 2016 by iw0ffk


Till now this summer is a bit short of multi-hop openings on 6m for my location especially toward USA and Caribbean area. Fortunately there was a good opening to Asia during the CQ WW VHF contest. The Thai hams are allowed to transmit on 50 MHz only during this contest, 16 and 17 July, so working the Thailand was a good shot.

Remarkable 50 MHz QSOs:

E2X in OK06 DXCC #196

HS0AC in OK04
E21EIC in OK03
E20QVD in NK99

DU7/PA0HIP in PK10

Worked also a new country distant only 1000 km to my station, so I had to wait a short ES skip:
5A1AL in JM62 DXCC #195

I was not much active on 144 MHz this year, but thanks to the great Android application ESsense  by PE1NWL linked to the DXrobot, I was warned promptly for the funny 18 June ES opening.

I had QSO with about 25 Ukrainian OMs on SSB and CW but the nice part was when the few kHz around 144.300 became crowded by neighbors callers so moved to 145.550 on FM mode and worked a good number of OMs transmitting with a vertical antennas and mobile transceivers.

Another USB controlled 230V socket

Posted in Electronics on 16 May 2016 by iw0ffk

The easiest way to switch on-off a 230V socket using an USB port is to buy an already made device, but if one has some hardware in the drawer and some spare time in a rainy Sunday, he can try to build it by himself.

The scope of this job is remotely switch on and off some instrument in the hamradio shack when I’m away, using the USB port of  an always-powered pc server.

Disclaimer: be careful, here we play with the main voltage. Proceed at your own risk and try to don’t kill anyone or burn something.

I used an USB-TTL adapter with the PL2303 chip, 0.65 euro included shipping on a web store (how they can do that?).


The pins on the right side are +3.3V, +5V,  TXData,  RXData, Ground.
TXData and RXData are useless for this project, but their tracks on the PCB can be used to route different signals, on their path there are also two surface mounted bonus leds to monitor the status of the lines.
A little mod is needed. By using a magnifying glass and a cutter, the TXD and RXD pins have to be cut on the PL2303 chip. I have cut them from the top, keeping the blade close to the body of the chip then, with the soldering iron, it’s easy to remove the pins from the pads and connect 1-2 and 3-5.


After this mod, on the output plug  with the labels “TXD” and “RXD”, there will be two RS232 control signals, DTR and RTS, that can be raised/lowered at will.

This project uses only the RTS signal.

A mosfet (BSS138) is used to drive an opto-isolator to energize the 24V coil of a 2×8 amps relay.
The 24V are obtained directly from the mains, well I don’t like it so much, but this voltage is needed only for the relay’s coil, it isn’t worth a transformer, anyway in the little junction box there was not room for anything else…
This circuit is *LIVE*, no one  must be able to touch it when powered, it has to be placed in a plastic box or carefully isolated.Danger-Live-Electrical-equischematic_powusb

Most of the parts (capacitors, zener, resistors etc.) have been recycled from an old steam iron control board.

When connected to my Linux server, the PL2303 adapter is mapped as /dev/ttyUSBx, where x is a number depending the inserting order in the USB hub. I have a number of PL2303 connected to the pc, so it’s  highly recommended to create a static link to the port, whatever be the first PL2303 connected. The file /etc/udev/rules.d/serusb.rules has to be edited/created to contain a string like:

SUBSYSTEMS==”usb”, KERNEL==”ttyUSB*”, KERNELS==”1-2.1″, SYMLINK+=”ttyPOW”

This maps the port always to /dev/ttyPOW .
The number 1-2.1 is the position of the serial adapter on the USB hub. These RS232 converters have all the same serial number, so the only way to identify them is to insert always in the same USB port, easy. The hub position of /dev/ttyUSB0 can be obtained with the command:

udevadm info –attribute-walk –path=/sys/bus/usb-serial/devices/ttyUSB0

To move  DTS and RTS, for now, I use a minimal Python script created to test the signals, I think I will insert it as a cgi in the local webserver to permit any browser  to power on the instruments.
Here the script: dtrrts