31 January 2014

Playing with AD9850 board

While trying to understand all the pins of the AD9850 board (marked "SR-HC08"), I could not find the square wave output. A good sign was that QOUT2 was at +5 V and QOUT1 at 0 V. A simple adjustment of the onboard trimmer (duty-cycle control) activated the AD9850 internal comparator and there appeared the square wave.

Therefore output pins in my board are:
  • ZOUT2: sine wave through an LPF
  • ZOUT1: direct sine wave
  • QOUT2: square wave
  • QOUT1: square wave (180° shifted)
 If you don't need the square wave output disable it with a fully (C)CW rotation of the trimmer.

Musings about going Bluetooth

I occasionally receive a question about transporting bidirectional audio over Bluetooth and not only CAT controls. The answer is "technically possible", but there are two possible scenarios.


This is the case for handsfree operation, be it in on the move or just around the house. Using the VOX this is possible, but on the RTX side we need a Bluetooth "master" device. In order to establish a Bluetooth link there must be a master initiating the connection towards a slave. Since all in-ear handsfree earphones are "slaves", there is need for a master device connected to the transceiver: I have not found yet a master bidirectional audio Bluetooth module (maybe the one mentioned in an earlier post?!).
The discontinued Jabra A210 seems to be the only device with such characteristics, and it must be modified in order to work with a transceiver.


Since digimodes are generated in the computer/tablet/smartphone, we already have a master Bluetooth device. So theoretically any handsfree BT earphone would work. Besides difficulties in modifying such tiny electronics, I have been told that the audio encoding and compression operated by BT audio devices introduces unacceptable noise and phase distortion. The former affects the ability to copy weak signals at our antenna, the latter makes it impossible to work phase-based digital modes (PSK31, for example, while RTTY would probably work).

29 January 2014


The discussion and web-xcitement around the MINIMA transceiver designed by the BITX author added yet another project to my to-do list. Before buying an Si570 for MINIMA, I still had to play with an AD9850 DDS board I recently bought.

Wiring it up to an Arduino Nano clone was a matter of half-an-hour. A working Arduino sketch was easy to find thanks to the work done by AD7C: in an hour my DDS signal generator was up and running.

The code is simple and gets the job done. I did a few easy changes:
  • corrected the "MHz" label (was "Mhz")
  • delayed the store timing of the latest tuned frequency (2s, now 20s, not to stress the uC EEPROM)
  • added a visual feedback whether the latest frequency has been stored or not
Even to my non-calibrated frequency counter, the DDS frequency was off by ~ +50 Hz. This was corrected with a simple math that I will describe in the future.

To my surprise the whole circuit draws about 165 mA: that's a lot of current! That's 0.8 W or power consumption. Even if the AD9850 DDS offers a huge frequency agility in an HF RX/TX, that's at the expense of current consumption. I will not give away my QRP frequency XTALs just yet!

22 January 2014

First 23 cm QSO

The first 23 cm monthly activity contest date came too soon. It was my first opportunity to test on the air a 20W transverter and 23 element beam. The antenna went up easily and I used the TV coax already installed plus two BNC-N adapters: losses everywhere!

The 23 cm antenna on the balcony, beaming 60°
Apart being away for a long time from an SSB contest, I immediately felt restricted to the 3000 Hz bandwidth of the FT817, while for the last 12+ months I was able to see 2 MHz of RF spectrum.

The RTLSDR does offer an advantage. Even though I have probably not missed a QSO, I was thrown back to the good old days of search and pounce with the VFO.

I was able to speak with all the people I have heard last year. All signals were very strong, with the ODX being S9+60. Curious band!

Next month I will try to arrange more room in the shack/lab desk so that I can switch the antenna to the SDR when done with QSOs: at last I will be able to compare the two RX solutions.

15 January 2014

Fire protection in the shack?

The recent premature failure of the Gould oscilloscope PSU made me think of the danger of dealing with (old) electronics in our indoor workbenches.

The white smoke that emerged from the PSU, smelling of burned paper, was most probably caused by a small fire or something alike. Even though I was quick to remove AC from the device, that action does not extinguish a fire!

Sure, electronics are not too flammable, but the spark/flame could reach something else nearby (namely: paper).

How to protect my lab/flat/home/building from these events? How did you, reader, prepare yourself? Should I keep a small fire extinguisher in the shack? A flameproof blanket? Water?

11 January 2014

Baofeng UV-82L first impressions

So far I have owned for less than 24h the Baofeng UV-82L VHF/UHF handheld (firmware version B82S21). Here are my first impressions.

It does lack a DC socket: it must be recharged with its own base.

With battery and antenna it weights 250g. It feels good, you can't forget it in a pocket like the UV-3R. Easy to use for a left-handed (PTT with left thumb).

The operator must get used to the double PTT , one for each displayed line. If you want to avoid mistakes, just tune both lines to the same frequency.

Switching between VFO mode and Memory mode requires a power-off cycle.

Manual programming is not straighforward, as usual on Baofeng devices.

There is no signal strength indication. Even if it has a 5-bar symbol, it is either absent (squelch closed) or present (squelch open).

The loudspeaker sounds good and volume control does work.

It does V/V, V/U, U/V, U/U. One receiver active at a time.

The LED torch is pretty powerful :-)

I got positive modulation reports (Wide setting must be used here).

Last but not least, the charger base didn't turn either green or warm after 8 hours charging. I am now in the process of discharging the battery so that I can try again.

Washing a chip

Before passing along the Arduino-to-PIC programmer, I needed to test it. My only PIC16F84A chip was buried in salt leaked from AA cells in my old Morse keyer.

Since salt should be solubile in water, I gave it a try: I washed the chip in warm water and in 5 minutes all pins were free.

Most electronic components are waterproof (ICs, resistors, capacitors, diodes) and will survive a bath in clean water just fine. Without power applied to them, obviously. Let them dry (or dry them by hand) and they should work again as before.

This was the case with my PIC16F84A, that confirmed the Ardpicprog is a good modern alternative to cheap JDM-style programmers. I loaded a simple "LED Chaser" firmware, whose visual effect is "so 80's", like RS232 technology.

08 January 2014

Gould 4072 PSU - removal steps

In the end I deciced to have a look at the Gould 4072 DSO Power Supply Unit: since the scope was already going to be trashed, I cannot do much more damage. So, how to access the PSU?

First of all: you will remove a lot of different screws, be methodic and classify them. I was not.
Second: you will probably meet a lot of dust.

Needlessto say, remove the AC cable. Remove the top cover (5 screws). Remove the bottom cover (7 screws).

The PSU has the following connections to the PCB:
  • to the nearby fan (access from above)
  • to the CRT circuitry (access from above)
  • the metallic ribbon to the front panel on/off button (access from above, remove it from the button, don't remove the button)
  • to the bottom side with four fast-on's (access from below)
  • to the bottom side with a flat ribbon (next to fast-on's)
  • one ground wire to the chassis (I cut it and will solder back)
Ribbon cable to the CRT
and the ON/OFF strip.
Take a picture of everything before removing it.

The PSU slides towards the top of the DSO, so prepare all the cables accordingly.

Remove all screws on the back panel except for those holding the fan. Your aim is to be able to separate the back panel top side far enough to let the PSU slide out (up). The whole operation can be comfortably accomplished with at least three hands :) Pay attention not to damage the metallic on/off strip.

Once the PSU box is free there are still 4 screws that keep the lid in place.

Fast-on's, ribbon and ground
wires on the bottom side.

Edit 2014-02-11: I do not own anymore the Gould 4072.

06 January 2014

Arduino-based PIC programmer

My adventures in electronics recently took me to the old world of simple serial RS232 programmers for Microchip PIC microcontrollers (16F84A should sound familiar).

Most modern serial boards do not output full +/- 12V, but some voltage in between, which is still perfectly acceptable by the RS232 standard (+3 to +12V is a valid "0" logic level, -3 to -12V for "1"). But the lower voltage breaks those simple programmers.

While looking for a USB-based solution I came across an Arduino-based PIC programmer called Ardpicprog (currently hosted at github.io). The huge advantage of Arduino boards is that they offer a virtual COM port over USB, which works with modern operating systems (let's say from WinXP and newer).

The new circuit will cost 5 to 10 times less than commercial products (5-10€) and directly replace the JDM programmer in the current housing.

Writing the firmware to the uC will now require a different software, but for a once-a-year operation it is a satisfactory alternative.

04 January 2014

R.I.P. Gould 4072

Shortly after discovering an optional accessory on my Gould 4072 DSO, I decided to keep it running for a whole morning. I left the top cover off, since I cleaned boards from the dust accumulation.

After some 10 minutes of power on a buzzing sound of paper into a fan emereged from the PSU unit, followed shortly by a white smoke. The smell reminds me of my paper burning experiments as a child.

A possible sign of the upcoming problem could have been the unstable signal measured on the display, but since I had cleaned boards the problem could have been somewhere else.

I suspect a failure in a PSU capacitor, but I am not going to troubleshoot it.

R.I.P. Gould 4072.
Edit: (Single boards are available if someone needs them to repair their unit) Since I've got nothing to loose, I will try to reach the blown power supply unit and, in case, attempt a repair.

Edit 2014-02-11: I do not own anymore the Gould 4072.

A bare Bluetooth module for audio transceive

While shopping for serial (CAT) Bluetooth modules, I came across a not-so-documented not-so-popular circuit sold as "XS3868 Bluetooth Stereo Audio Module".

The heart of the circuit is the OVC3860 chip, whose datasheet is hard to find, but specs look promising.

As I understand it, the XS3868 module is somehow what you would find inside a large Bluetooth headset for a cellphone, featuring playback buttons, volume control and a Li-xx battery charger. It runs off a 3.7 V Lithium cell and it can be controlled with AT commands via an embedded serial port.

Theoretically, if the resulting audio signal is not too distorted, this circuit can be used to build a wireless audio interface between a computer (Bluetooth master) and an RTX. This is most interesting for digital modes.

First tests may complete in late February 2014.

At about 7 USD, this is a worthwhile experiment for the future of HAM radio, HI.