27 December 2010
All my diagrams show the contrast pin going directly to ground: that's how my display works. Any voltage level higher than that makes the contrast too light and display unreadable.
Other displays need an intermediate voltage, so you better check out how yours works before soldering it in place! The datasheet might be a good starting point.
These displays will be used on my FT-817 Frequency Reader and Interactive Frequency Reader projects, so I will have to be careful too. ;-)
10 December 2010
The location has a landline, but I have not found useful references online on how to interface with it (is that considered a pre-Internet era technology?!). So, let's go mobile.
Take a cellphone that allows to customize the ring tone depending on the caller:
- associate a fixed tone sound to one calling number: this will be the "ON" triggering event;
- associate a different fixed tone sound to another calling number as "OFF" triggering event;
- associate a third, no-sound, ringtone to all other numbers (default).
Next, build some simple hardware to detect those two tones and throw them into a simple set/reset system that drives the relay. Or make it with a microcontroller. The latter solution would allow some feedback on the relay state if the cellphone can be configured to auto-answer a call: let the circuit emit two different sounds depending on the relay state (or Morse code!). Maybe don't let it beep forever, so neighbors won't get annoyed.
Note that this "interface" can be arranged without harming a cellphone, because the audio coupling can occur over the air (yes, the location is quiet and empty).
Let me look for some sound recognition sample code for AVRs... :-)
29 November 2010
Recently I have re-engineered my homebuilt joule-thief after a too heavy shake that broke one wire. It was a good chance to let my older daughter choose for a new container, and a new color.
An empty play-doh can laying around was chosen as the new night light. The lid holds just right a 1xAA container and a small SPST slide-switch. Wires hold up the LED.
The 5mm while LED gives a quite directional light beam, but the transparentish can allows the light to be diffused around.
The Joule-Doh night light generates enough brightness in a dark 4x5m room so that you can walk around without hitting abandoned toys and tuck your child(ren) up.
For those not familiar with the "Joule thief", it is an oscillating circuit that allows to light a LED out of a 1.5V battery (be it AA, AAA or button). It usually works down to 0.8V, so it is a very good way to drain all the juice from those batteries that won't work anymore in normal toys. This one has been running already for 5 nights off the same "officially discharged" AA battery.
12 November 2010
NE602 datasheet reccommends a 200 to 300 mV p-p drive from an external LO.
An application note shows this arrangment with a resistive divider for a TTL signal: 510 and 30 ohm, fed through a 0.1uF capacitor. This results in 277 mVpp for a 0-5V square signal.
Simple, isn't it?
Since I ignore the output switng of the canned oscillator I intend to use, I will try with a 1k multiturn trimmer and tune for an acceptable level.
08 November 2010
So I produced 5'30" worth of video-instructions for the external frequency readout for transverter users. The audio is in Italian and I will add English subtitles.
Current features were summarized in another post. Please send feedback via email.
05 November 2010
The component count is as low as possible. You may choose to take the supply from another source than the FT8x7 ACC port, such as a dedicated 9V battery.
You cannot leave out the two N.O. pushbuttons "[Up]" and "[Enter]" ("[Next]" on the diagram), but you can skip the LCD backlight control (mind the total supply current does not exceed 100mA, better less!).
This circuit is also the base for the keypad+display device I am still working on. The keyboard extension chip (PCF8574) will be connected to pins 4-5-6, so leave room for it on your board... in case you'll be interested in that device too.
"Release Candidate 1" ("RC1") firmware is available on request, until I publish it on my website.
03 November 2010
I have kept a two-button user control interface and changed the LO programming method.
Since the survey (and following emails) showed that it is possible to have 100 Hz accuracy on 47 GHz transverters (WOW!), the display now allows 0.1 kHz resolution both for LO and VFO.
The LO value, 16 values actually, are inserted digit-by-digit through the two buttons when in "config mode".
Updated characteristics are:
- based on ATmega48/88/168 chip clocked at 11.052 MHz
- 2*16 character display (HD44780)
- 2 buttons interaction
- "config" and "normal" operation modes
- 16 LO values stored in EEPROM (survive poweroff, can be fewer)
- LO value down to 100 Hz
- LO ranges from 0 to 214'748'364'700 Hz (that's 214.7 GHz)
- VFO/output frequency computed to 100 Hz resolution
- support for both Out= LO+IF and Out=LO-IF
- LO value selection by button press (1 > 2 > ... > 15 > 16 > 1 > ...)
- VFO A/B toggle on second button
I am working on a circuit diagram to be published soon. Who built the ATtiny2313-based display can assemble a simple adapter board for ATmega chips.
26 October 2010
After I have fought with the tiny2313 to fit all requested features for the external display (for transverters), and lost ... after I have searched up and down the Net for the ATtiny4313 PDIP and realized that it is probably not yet in production ... I gave up.
The choice was whether to maintain backward compatibility with the stand-alone keypad based on ATtiny2313 OR establish a new standard, compatible with the "remote control" to-be.
While the ATmega chip can cost 3 times the ATtiny uC, it gives me much more freedom to produce something not only satisfactory, but fitting all the requirements. And leaves headroom for future extensions. Moreover a PCB drawn and produced for the display can be done for the keypad+display at the same time, just leave out components when assemblying.
So, ATmega48/88/168 is the final choice. Probably an ATmega48 will suffice.
In any case I will not throw away the ATtiny2313 external display firmware for those who have built the circuit published in the last month (October 2010). It will be available on request.
In any case the 2*16 LCD display based on HD44780 controller will not change.
21 October 2010
I have arranged an online survey in three questions. It's anonymous but will help me gather a broader overview on the microwave "land".
Please spend two minutes with my survey. I can leave it running for 10 days, so until October 31st, 2010.
I will publish aggregate results here once the survey is over.
19 October 2010
- Design and test an ATmegaNM stand-alone fuse restorer
- Design and build an HF/4m transverter (TX-only)
- Assemble a keypad+display prototype on perfboard
11 October 2010
Now you need two N.O. pushbuttons and one DPST switch.
If you don't want the A/B function, leave out the pushbutton but do insert the 10k pull-up resistor. Same for the dual LO, just put a pull-up resistor between pin 12 and 5Vcc if you won't be using a transverter with input into the UHF region.
10 October 2010
The new firmware allows L.O. values for 10 GHz (and more) transverters. The programming occurs as announced before, reading the display frequency and interpreting it as GHz instead of MHz.
But two L.O. frequencies can be stored, one to be used when the 817 (857, 897) is tuned to HF/VHF and another for UHF. According to a microwave experimenter, with this option it is possible to use 10 and 24 GHz transverters. UHF is recognized automatically by the display, and L.O. value varied accordingly.
Since automatic band recognition is enabled, there is also an output pin switching high from HF/VHF to UHF in order to switch transverters! One radio, one external display, two transverters! All within an ATtiny2313.
Here is a preview on the schematic diagram. Firmware will be published on the usual website. But it's available on request since now.
PS: this firmware works on FT817, FT857 and FT897, and their ND versions.
23 September 2010
A preliminary version of the circuit layout is working on breadboard. It includes an ATmega168, a PCF8574A TWI I/O expander, matrix keypad and LCD display.
I have been able to read the keypad using interrupts and display keypress on the LCD. I have been able, with another piece of firmware, to read data from the FT817 and display it on the LCD.
Now it is "just" matter of putting both codes together and working on a firmware logic that combines a continuous frequency display, keypad input and a sort-of menu driven function selection.
20 September 2010
I took a wall-wart power supply, said to output 9Vdc, 1A max. I measured it under load to be at 10.3Vdc. So with my surplus power resistors summing up to 12.7 ohm, I could drive the LED at about (10.3-5.9)/12.7 = 0.34A (measured 0.32A).
The result is shown in a short youtube video, composed of few pictures.
As a practical result, the sink area is very well lit even at just 2W input. The cold white light gives an unreal look to the inox sink, but colors are not distorted.
The heatsink has been equipped with two thin magnetic strips (superglued), so that the installation is not permanent.
16 September 2010
14 September 2010
I know from their datasheet that they have a rather small angle and need heat dissipation.
The shop in town offered me a cold white 3W LED for 10€: too much for an experiment. The Net comes handy, with a 5W LED at 3.5€ shipped from China.
It's from "DX", part number sku.4516.
The LED is soldered on a star for easier mounting on heatsink, and should work at 7V, 0.7A. My sample has a forward voltage of 5.9V only (out of specs?!), but it's really like having a star in your room, even at lower current drives!
Heat. After the initial warmup, the LED+heatsink reaches about 40°C at room temperature of 25°C. Wow, and I'm running it at half power!
In the next post I will show a possible application of this LED, and the comparison with two halogen spotlights in my kitchen.
02 September 2010
The problem is evident with moving objects: they leave a "shadow" behind them. It's more or less the same effect of scrolling text on slower (~= cheaper) LCD TV/screens, that becomes unreadable.
With a colleague we tried to imagine the technological culprit, be it the slow CCD sensor or the videocamera processor or even the SD card speed.
Speaking of the FS200 only, it has three recording quality settings: LP, SP and XP. They are equivalent to 3, 6 and 9 Mbps. Is this the data rate from the CCD to the MPEG processor, or the top writing speed to the SD card? Probably it's the latter since the user's manual states that the FS200 uses VBR encoding.
In order to get closer to a plausible answer I need to create a relative speed between the camera and an object that is repeatible an infinite number of times. The speed must always be the same, so that it's possible to evaluate artifacts in the recorded output (on a CRT screen, if possible!).
Excluding all human being movements (non-repeatible speed) and those controlled by a human being (like a car), I came up with these constant-speed sources readily available at home:
- an R/C car at full steam after the initial acceleration distance, within few runs of fully charged batteries
- the swinging weight of a pendulum clock within a short timeframe (say, 1 hour or so)
The experiment will be as follows:
- mount the camera on a tripod and point it to the moving target path
- record the moving target at each recording quality settings
- change the SD card to a faster/slower one and repeat the three recordings
- compare on a fast screen
01 September 2010
It started with a suggestion on the 817keypad reflector when I asked for feedback on the simple remote display last Spring. Now I have a working code filling 100% of ATtiny2313's program memory with the following features:
- read the transverter local oscillator frequency from the 817 with 10 kHz resolution (in case it's not set to x.000000 MHz...)
- store the transverter L.O. in EEPROM for power-off retention
- show the output frequency with 1 kHz resolution
- all frequencies are displayed in kHz (sorry, too much code needed for a nicer format)
30 July 2010
I have composed a firmware with common functions to 817, 857 and 897 that is available on request. Key/function assignments are as follows:
- A VFO A/B
- B VFO A=B
- C SPLIT ON/OFF
- D MANUAL TUNE
- # Onboard Memories
- * Mode change
- 0 direct dial
- 1 direct dial
- 3 Lazy PTT
- 4 direct dial
- 6 RIT
- 7 Repeater +/-/no
- 9 LOCK
05 July 2010
Its name is "Design Spark PCB", look it up with your favorite search engine. It is sponsored by a worldwide known component seller.
I am giving it a try with the circuit-to-be of my 817 keypad evolution.
23 June 2010
The current, working, version is 22.214.171.124 .
Is it the end?!
18 June 2010
Mine liked to crash when opening a youtube video, or when something happened on the audio output. Earlier on there was a problem on the network card driver, that was apparently solved with a driver update (available through official channels and offered by Toshiba's software onboard).
Driver patches for the soundcard were not available. Instead there was a BIOS update to be applied. Follow carefully instructions and you'll probably solve this BSOD cause. It's all on Toshiba website.
14 June 2010
I own a parallel port "serial download" programmer, which works perfectly fine both for ATtiny2313 and ATmega168 (mind different chip size and pin assignments, of course). I could unset the RSTDISBL fuse but then I was locked out of chip reprogramming.
Why? Why!! Well, it turns out that it's not stated in the datasheet, but once RESET is disabled, serial downloading is not possible anymore. You need either a High Voltage Programmer or an ISP or a JTAG cable (the latter two might not be supported by the chip itself).
Now I have two mega168, perfectly sane but that I cannot reprogram unless I build a HVP. At the time being I have chosen to order new mega168 and invest my time on the code instead of building the HVP.
Well, actually I have re-built my parallel port programmer with a DIP28 socket and a longer cable. Pics will follow in a future post.
Note, for search engines ;-), this observation applies to Atmel's ATtiny2313, ATmega48, ATmega88, ATmega128, ATmega328 chips and variations. To others probably
08 June 2010
Then I played with an LCD, to make a remote display.
Now I am working on a keypad+LCD to make an "Advanced Frequency Readout". :-)
If it works out, the uP will be an ATmega88 or ATmega168 chip and the keypad will have to be reduced to 3x4. A prototype is already assembled and it waits for a first firmware relase.
04 June 2010
- quick band/mode change
- split control
Band/mode combinations are for 9 bands with two modes, summing up to 18 combinations. Since the standard firmware allowed 15 (or 16 if static) onboard memories, I wrote the code to use all the available Flash memory and give the operator 25 dynamic locations.
These 25 memories have to be programmed once by the operator, and they are retained after keypad power-off. They are organized in two banks, accessed from the main menu with key "#" and key "D".
The same firmware also carries A/B, A=B, split on/off, direct dial and mode set. Unused keys are locked too, so that they don't produce invalid frequency input.
More static memories could be built in, but you would loose the ability to reprogram them once out on the field.
25 May 2010
17 May 2010
So does 7030 / 16.
How about 7040 / 16 then?! It gives 440 kHz, that is even closer to the I.F. of 455 kHz.
I should add this feature to my xtalfind.pl utility.
I also tried out the receiver and with my finger as antenna a DRM transmission was clearly audible on 9.8 MHz band. So I'll proceed with this IF downconverter project.
10 May 2010
14060 kHz / 32 = 439.375 kHz
455 - 439.375 = 15.626 kHz
So, with a binary counter/divider I can get a LO acceptably close to 12 kHz. I know that DReaM software will decode a DRM signal centered at 15.6 kHz.
Where did I store that 74HC4060 ... ?!
03 May 2010
Download it here.
Please note. Article submissions are welcome for future editions of the bulletin. Even if not in Italian language, they can be translated by some native speaker.
25 April 2010
- volume control slider too sensitive
- audio volume too loud
- strong hum on audio
- loss of presets and clock unless 2xAA cells are fitted in
The radio has plenty of room inside and all components are through-hole. At a first glance the probable cause of hum is evident: 15VAC and audio line to speaker run together to/from the main board.
I have to build a (circa) 455 kHz to 12 kHz downconverter.
The IF system chip is BA4237L; the AM IF filter is marked SFU450B (thus 10 kHz BW at 450 kHz). First decision to be taken: should I pick the IF before or after the filter?
If I pick it after the filter I need a L.O. at + or - 12 kHz, so 462 or 438 kHz: how to make a stable one? (I will use my xtalfind tool to see if any of my surplus XTALs fall close to those values)
If I pick it before the filter I can make a stable 455 kHz L.O. and retune the receiver L.O. so that the resulting difference is 12 kHz. Would the incoming signal be too wide?
Let experiments begin!
19 April 2010
It was probably caused by a detail of the GPS protocol information, where it carries a "week" information with 1024 possible values. That's almost 20 years. See the wikipedia page about GPS for further info.
Garmin has released a firmware update (currently v2.90 for the Geko 201) that corrects the issue.
14 April 2010
- the IC706MKiiG "as-is" is not suitable for weak signal work on VHF and UHF
- aircraft scatter is easily observed at 2m too
The picture shows both the IC706 drift during a two-minute WSPR transmission and aircraft scatter on two airplanes (at circa 2 minutes interval, their "usual" distance on the same path).
The enormous drift was present at each transmission slot (1:4 ratio), rendering unuseful the WSPR signal. This was on 2m. A compressed drift amount was later observed on 10m as well, but the signal could be decoded correctly.
Another IC706 problem, not shown in the screenshot, was the misalignment of the device and probably some difference on TX/RX frequency too: the IC706 station had to retune 400 Hz higher to appear into my WSPR receiver window.
It's always good to know your transceiver peculiarities!
09 April 2010
The microcontroller board is exactly the one of my FT817 keypad. I only needed to add a connector to source +5V and GND to power the display.
Depending on firmware size, some single-function buttons could be added around the display, such as VFO A/B, split, mode roll.
07 April 2010
Yes, it does work. First I checked with an antenna analyzer the new setup (using dir2 and dir3 as dir1 and dir2 respectively) and I measured a nice 1.1:1 SWR at 144.3 MHz.
Then I mounted it outside and tried some real-life contacts. SWR was 1.2:1, probably increased because of the metallic fence nearby the antenna.
Either propagation was better than one month ago or the theoretical 1dB extra gain on my other 4el beam did help a bit: best DX at 227km with 50W SSB. Not bad for a beam on a balcony!
Greg assumed that given my keypad code it would not be a big problem to implement such a gadget, and he was right.
I have the code ready for real-hardware testing. Now I need to get hold of a LCD display.
I will let YL choose the color :-)
03 April 2010
Months have passed and I have come across the broken boom again. I had already planned a longer beam for my portable VHF contests (7 elements, 3m long) so I will not replace the boom.
Instead I simulated what would happen if I shortened the 5el to 4el with MMANA-GAL. The result is encouraging because without changing element lengths I can:
- install the 3rd director as 2nd director at 1.465m from the reflector
- have about 7.7dBi gain and 15dB F/B (not the best)
- keep a good match to the original 28 ohm
Not bad! Well, this operation moves the center of gravity a bit behind, so I have to re-position the boom-to-mast adapter or add a little weight in front of the antenna.
I hope I can get it up and running for next Tuesday's monthly VHF activity day.
Note: the original 5 el. is DK7ZB's VHF Yagi 28 ohm with 2m boom.
27 February 2010
Might have been a coincidence, but I noticed cleaner signals than 8 months ago, and especially cleaner macros! Nobody told me about their computer and software. Only one sent me the various club codes. But ... nobody told me anything about their operating conditions, which I would have liked to know!
Thanks to the 7 guys that answered my QRP calls.
18 February 2010
So, does the Yaesu FT817/FT817ND output 5 Watts when powered with the internal battery pack? Does it output 5W when the "LiiI" icon is flashing?
I did the measurements and posted a video on youtube.
The setup includes a 26-30 MHz wattmeter (10W f.s.) and a dummy load, stock 1400 mAh internal battery.
I measured the output power at 5W and 2.5W level with the internal battery and external PSU. The sequence in the video is battery, external and then battery again.
The result? Even if the icon is flashing, the radio is capable of outputting 5W.
08 February 2010
If you have a similar chaos in your computers and you have easy access to broadband (ADSL or more), you might like an online folder fully integrated with your Microsoft and Linux (Ubuntu) operating system.
Search for "dropbox" in your favorite search engine. They offer 2GB of online storage, backed-up. Files will be downloaded (sync'ed) to all your installations. So you have them online and on all your computers.
If you complete few steps you can receive 4x 250MB, for a total of 3GB of storage.
You can get your first extra 250MB if you subscribe through this link (I get 250MB too).
03 February 2010
The transformer core is unknown, salvaged from some electronic equipment that needed an RF-choke to prevent RF radiation.
As a rule of thumb I would assume all those cores are good candidates for LW/MW use. I used 24 turns on the loop side and 4 turns on the 50 ohm side.
31 January 2010
The transformer requires approximately a 6:1 turns ratio, for a 36:1 impedance step down.
From the bag of unknown toroidal cores I picked a small one, black, the size of most familiar T37-x. It was used as an RF choke in some computer equipment. With a thin wire I wound 24 turns on the primary (loop side), so that I would need 4 turns on the secondary (RX side). The DVM LC meter measured 900uH from those 24 turns: high, as expected! The material should be appropriate for the intended frequency range!
The result on the receiver is astonishing. The previously observed lower Q on the high end of the tuning range is fixed: the tuning is razor sharp on the whole range. The new range is 150 to 570 kHz. The transformer does not seem to be lossy in LW/MW as observed signals are equal if better than before.
I still have not understood where the figure-8 radiation pattern lies: orthogonal or on the loop plane? Sources say othogonal, but I observed it on the loop plane with aeronautical NDBs. I mean, if on the loop plane was built a yagi, I orient it for the best signal. Anyone?
28 January 2010
The first hosted on blip.tv is about Russian Arctic Islands in 2001. It is one hour long.
The second is from 1976 on Google Video, 7J1RL Okino Tori-shima, full of music background from 70's.
Both are very instructive for anyone considering a field activity of any sort, even the apparently simple task of operating from the nearby park.
Use the search engine on both websites to locate these videos.
I will resume listening only once I have done some changes to my setup, like installing the antenna outside or building the loop preamplifier.
25 January 2010
CFL light bulbs seem to have no ill effect on the background noise, except for some birdies that can be heard occasionally. Also SMPS in the shack don't make a difference.
The loop still has no impedance matching transformer, so the high Z goes directly into the FT-817 antenna socket. That's probably the next improvement planned.
22 January 2010
As antenna I used my multi-turn 1m square loop inside the shack, turned so that I minimized the received noise. Moreover the loop tunes about 230 to 450 kHz, with a very low Q at the high end, so 500 kHz is a bit border line for its current configuration.
Nevertheless, buried in the noise I could receive PA0A about 900km away. Go figure on the NDB band I am able to read only 3 signals, the furthest being just 60km from my antenna!
So the loop is not that bad. The location needs improvement. I will move the loop outside, on the balcony. Then with a warmer season I might try to move in a less RF-polluted environment.
21 January 2010
I played with it maybe 10 years ago but I still had all parts laying around. It is composed of:
- 4 meters of computer flat cable, 26 wires (probably), soldered "shifted-by-1"
- two wood sticks 1 meter each
- one variable capacitor (whatever you have at hand, even a trimcap will do)
The loop will be 25 turns in parallel with the capacitor, coupled with one full turn loop to the receiver. Probably this configuration has a high impedance, about 2kohm (how to measure it? my MFJ259 doesn't go that low), so I might also try a step-down transformer with a surplus toroidal core salvaged from somewhere.
If the loop will not pick too much noise I will try to add an active amplifier in between the antenna and the receiver.
18 January 2010
I had once bought them online and I was relatively satisfied with their quality. Then I had to re-supply at the local store and paid the incredible amount of 1.5€/cell. They had even more expensive ones!
Out of 4 batteries, one was dead. Other 3 lasted much less than those that came with the toy.
During a desperate need of one of those toys last night, I got hold of my DVM and measured voltages across those "exhausted" cells.
40% of them that had been marked as "unsuitable" for toys were at 0.5V or below. Way below: they showed polarity reversal! The remaining cells were measured to be 1.2 to 1.4V and, not surprisingly, when mixed with good cells, brought a couple of toys back to life.
You may have to do some interactions until you find a working trio, but don't throw them to recycle bin right away because the fault might be in just one cell, not in all of them!
Apparently they should have capacity in excess of 150mAh, I think that a simple testing clamp can be arranged with a resistor mounted on a clothes peg...
04 January 2010
Ho pubblicato il calendario IAC 2010 su un file separato per non creare problemi a chi avesse già utilizzato il calendario dei contest da "weekend".
Lo trovate sul mio sito: www.paolocravero.tk
Il mio calendario dei contest si può consultare online su google o sottoscriverlo (iCalendar).