30 January 2009

Optics 1 : Electronics 0, tale of a partial failure

Here follows the report of a partial failure of an optical one-way contact over a distance of about 9 km (5.5 mi).

TX location: my parent's flat in the suburbs of Torino, NW Italy.
RX location: parking lot of Superga hill, 9km East of TX

In between us, right downhill the river Po and then 8 km of city with its heating fumes, car pollution, humid air that created the usual urban haze. Lights up to 10km were visible and the moon was shining on far away mountains.




TX equipment: 10mm hi-bri red led MCW driven at an average of 80mA
RX equipment: OPT201 followed by 488Hz bandpass mounted on a 90mm telescope

On the RX side we also had an EEE netbook running Argo for QRSS work.

Aligning the RX was tricky. Street lights helped to identify the area but not the actual spot. I used a pair of binoculars and cellphone with earphone to talk to the TX site (no HAM there!). No high power lamps on TX side. Then an idea: TX site is right under the landing path to our airport, and an airplane was approaching. When it was right on top of the building we got a "now!" signal.

After was just a matter of aiming the telescope and search for the blinking LED. It was sending 6 dashes, pause, cycle in slow MCW morse. Car stop lights traveling around were catching out attention all over.

Simone, the optics owner and expert, could spot the TX in a couple of minutes. Hooray! In the telescope the 10mm LED was clearly visible but as an average star would look at the naked eye! Buried in yellow street lights and reflections.

We mounted the receiver, adjusted beaming, ..., but nothing. Not even the 50 Hz hum was audible. We replaced earphones with the computer DSP software but nothing. Sometimes 50 Hz & friends were visible in the normal scroll speed, but no 488 Hz line.

Now I've got to work on improvements. I need a larger and/or more powerful source. Those 50dB over background noise we had over the 200m path were not enough for this DX. I will add more LEDs, and maybe put them in the focus of a beam expander (Fresnel lens or parabolic mirror).
I will build an aiming system for the TX, even if it has a rather broad beam.

I will build a k3pgp receiver. I have 5x BPW34 detectors. I'll probably follow Yves's circuit. Any idea for a 2N2482 replacement?

We will wait for a windy day to repeat the experiment.


26 January 2009

PTO, taking apart

Sorry I kept you waiting for further pictures of the PTO. This is it, removed from the FM 88-108 car receiver.


A small piston drives three cores in and out inductors. There are two adjustments. One marked "IF", the green can. The second looks like a trimcap, marked "OSC".

I am slowly digging my way to the circuit behind PT inductors. Then I should be able to draw a circuit diagram.

Heavier paddle

I do not own a professionally made morse paddle. Or IAMBIC key, if you prefer to call it that way.

I have used an ultimate low-cost ultra-portable paddle for all my A1A (CW) contacts, but it was also ultra-light. I had to use boths hands: one for keying and one to keep the paddle still.

It is a fine compromise when sitting on a mountaintop, but not in the shack. I asked around and those fine looking paddles weight 1kg (2.2 pounds) or more. Heaviest materials for a proper base are: platinum, gold, tungsten... marbles. They all weight more than 2 kg per cubical dm.

I excluded those noble metals and opted for a scrap piece of marble from the local reworking plant. I got three pieces of what looks like granite for a total of 2.7kg (approx 0,7 + 1 + 1 kg).

Before drilling I wanted to try the concept. Here's the result:


"JFK" in SP3 land was the first contact. Then few more rolled in, for some finally relaxing 2xQRP contacts.

Now, a new problem: how do I drill this piece of stone?

More 30m QRP adventures

On Saturday, 2009-01-24, 30m were wide open in my shack. I had troubles finding a clear spot to place my CQs: up to 10130 kHz I could hear some operators!

I could manage 6 contacts, our of which 3 were 2xQRP and two of these with GQRP members.

I was testing the new base on my freshly homebrewed paddle. So this is how the "shack" looked like...


If you have IK1ZYW on your log in early 2009, then this is probably where your signal has landed to. The balcony antenna was shows few posts back.

I wonder what the OP understood when I announced "NOW QRT FOR XYL". I had my daughter calling to play with her.

23 January 2009

21st Century PTO

I was quite puzzled when I bought a used 2005 car in 2007 and it came with a peculiar el-cheapo radio. It was an Audiola SCD-857N for which there's no mention on the Web, with the following specs:
  • FM broadcast band only, very insensitive
  • analog-like tuning (yes, a knob!) but digital frequency display
  • obviously no memories
  • CD player for original media only, not even a CD-R
At the time being, January 2009, Audiola/Majestic still sell devices with similar receiver specifications in the Italian market at 50 euro entry level price tag.

You cannot imagine my engineer's happiness when the front panel hook broke and I had to replace the whole device: I could finally peek into the tuning mechanism! The replacement is an old FM/MW/tape Daewoo car radio that I kept aside 9 years ago because of its good receiver performance (so at least 14 y.o. device). Tape vs original CDs doesn't make much difference to my listening comfort! The replacement process will be documented in another post.

The Audiola radio was built in 2004 according to its outer labels. The CD player module is stamped 12/2003. Inside I found two large pieces of thin cardboard as insulators between boards. 3 ICs are SMT, the rest are through hole R/C/L. But the tuning thing...


there you see it, at the rightmost side of the chassis. Here's a zoom view:


First thing, the black disk. it is a sort of clutch that transmits rotation from the knob to the PTO. It is needed because PTO axis and knob axis are not aligned. Cool.
Inside the metallic box there are 6 cylinders, of which only two are used, and a plastic plate pushes inside them a metallic rod. Behind them there is a shielded circuit that I haven't opened yet. It has 6 wires plus ground, a couple of holes for adjustment (there's an IF-can). One line comes from the antenna plug and another goes near the display driver chip (frequency counter?). Other 4 lines (grey wires) land nearby a couple of ceramic filters.

I have been able to remove the whole block, but not taken a picture.

What is behind the variable L? A passive circuit? Active? Could I reuse this input stage for 2m or 4m? Has anyone played with similar things?

In any case I was surprised to find such an arrangment in 21st century instead of a mechanically simpler varicap or a PLL.

22 January 2009

FT817 keypad - updated circuit diagram

A keypad user has reported a non destructive problem when using the on/off switch on the +13.8V line.

It shows as a faint LED when the keypad is off, and blinks when keys are pressed.

This is probably caused by the internals of the uC that let some voltage through from input pins. And TXD line is at +5V in stand-by, as long as

I have investigated a BJT/FET buffer on the TXD line, but didn't solve the problem.

So the simplest solution is to use a DPST switch that disconnects both +13.8V and TXD lines. This mod should be possible on all already built keypads.

16 January 2009

Breathing a LED down to zero


Finally I got something broken to fix. Actually this time was something to improve.

I helped the amateur astronomer of my optical experiments with a circuit to give a "breath" effect to the eyepiece grid LED.

He built the circuit off a diagram I found on the web right away on the perfboard/veroboard. The LED throbbed/breathed, but not all the way to "off".

All similar circuit uses probably do not need to turn off the LED, but he did. The op-amp has a virtual ground at Vcc/2: what about retouching it?

Pulling it higher, say at 2/3*Vcc, does the trick. This requires that the two 100k (or 47k) resistors become 100k on the Vcc side and 47k on the GND side. Or something like that. A 100k linear trimmer allows for finer adjustment.

[Picture: nice wiring, isn't it?]

07 January 2009

HAM radio: childish fun?

I can't stop thinking of this episode few days ago.

Once I fed breakfast to XXYL (2y4m old) I left her playing and went into the shack. YL was still sleeping. XXYL came over and joined me while I was listening to who was on the band. She did some drawings and left to her mother.
While I worked UE9, mum had asked what was I doing. And XXYL replied: "he is playing" [with his toys]. XXYL understood my fun and YL understood where I was.

Hopefully one day she will join me behind the key.

Winter 30m experiments

These past winter holidays have brought a positive spike in my on-the-air activity. After many months of HF inactivity, thanks to the availability of a comfortable support, I assembled what I called "2009cedevita30" antenna.

It is a base loaded radiator supported by a 4m long fiberglass pole on the balcony, at the 8th floor (approx 25m a.g.l.). The coil has been computed with the help of loadcoil/vertload programs and wound on a empty Cedevita can: 9 turns on a 6cm dia. support. The counterpoise is the balcony fence (mine and unsuspecting neighbor's).

I chose to develop a 30m antenna because I had built a SW-30+ kit, and obviously needed an aerial to try it out. But once the antenna was tuned I got carried away, used it and forgot the SW-30+!

First picture shows the antenna the morning of 26/12/08 after a Siberian snow storm. Second picture was taken few days later after a snowfall, with frozen drops hanging on the pole. No damage to the pole was noticed afterwards.

First few days I could listen after sunset and the band was pretty much dead. But few RTTY/PSK31 signals were there, so I prepared the FT817 for digital modes. On the workbench/shack it was a tight fit, wires running everywhere and nothing really interesting on the band. After one successful contact with OM-land I noticed some hum/ground problems in my transmission so I returned to the CW key.

5 months had passed since my last CW-active days. I could then operate after lunch when 30m were open all over Europe.

In 6 days I could manage about 20 European QSOs, all with 5W CW, with the worst report 549 against his 579. Some 2xQRP, 1 SOTA activation too. W/VE stations were clearly audible on the greyline but didn't manage to work any. Best DX probably Asiatic Russia one morning. Not bad for a balcony fisherman!

The E4 activation did ruin a bit the fun since many operators were busy on the pile-up.

Being able to work CW daily I noticed an increase in my copying proficiency. I still cannot send code and answer someone talking to me, and now the fun is over...

What about the SW-30+? I did find the time to connect it to the antenna and tune around. It has a nice narrow filter and I could hear everything the FT-817 did. I will try to setup a video recording of it and hopefully start using it!