Dirt cheap radio hacking!

[Th3_uN1Qu3]

I've been playing with an AM transmitter these days, built with nothing more than a NAND gate, a few resistors, a few caps, one transistor and a tunable coil. The NAND gate is the oscillator, and the transistor with the LC circuit provides modulation, filtering and power gain. The gimmick proved to be quite stable, working for 4+ hours on end without drifting at all as per my Yamaha PLL tuner. Better than a transistor LC oscillator, and certainly WAY better than a 555. I experimented with carrier current transmission (coupling the transmitter output to the neutral of the AC power line) and was able to achieve good coverage around the house. That's quite a feat for the puny 10mW that this thing puts out! Heck, that's even within legal limits for AM.

I no longer have an oscilloscope and my multimeter only goes up to 50MHz, so it wouldn't be wise to try to hit the FM band (plus i have a car modulator i can use for that anyway), but i could aim for shortwave. Trouble is, i had given away my last "world band" radio 2 years ago! Since i was out shopping anyway, i remembered a store at the bus stop which sold cheap crap like watches, clocks and radios.

For $7.69, i got myself a pocket radio with LCD display, world band, AND alarm clock! They even had radios with a USB/SD card MP3 player built in for $15 , but they were kinda big.



The radio is branded "Interational". Knock-off... Surprisingly, i actually found out who makes them, it's a Chinese company called Kaide/Kchibo, and they even have a website, including PDF manuals. This radio has "KK-848" written on it as the model, but the manual that came with it (in Engrish of course) was for "KK-9704" which appears to be an earlier model with 7 SW bands, mine has 8. I even found a review of a "KK-9702" here.

Surprisingly, it actually sounds decent on headphones, it's got some bass to it. The volume knob goes way up into overdrive, as usual for those Chinese radios. Reception is stable, but the LCD readout is .2MHz off on FM. Just like the last one i had. Looks like there's hacking to do. I also agree with that reviewer, the band indicator and tune LEDs are way too bright and just waste power. I don't want those shining in my face and eating up my battery, so i'll tone them down. But hey look, there's even a button for LCD backlighting! That's a new one... Unfortunately it's a lame little green LED in the top-left corner, which doesn't really do anything. At least they tried.

I'll be back with pics and hacks.

[kc8adu]

roflol!
interational.
yes pics please.

[Th3_uN1Qu3]

Ask and ye shall receive. First 2 pics show its initial state - as you can see, the soldering quality is generally atrocious. Not that it's a surprise...

The main radio IC is a clone of the TA2003. Also visible is the reason for the powerful sound - a TDA2822 in bridged mode operates as the audio amplifier. The inductor visible just under the radio IC was the one i adjusted to align the displayed frequency with the real one.

First close-up is of the swaps for the LED series resistors. The originals were 220 ohm each - which means that on a full battery, each LED ate over 10mA! Considering that two of them are on at all times, this is quite a waste of battery. I settled for 2k2 for the red Tune and FM LEDs which resolves to just over 1mA each, and 1k for the green AM led, ~3mA, because it is dimmer than the red ones for the same current. The Tune LED did not have its series resistor close to it and i couldn't be bothered to trace it, so i cut the PCB trace and bridged across with a 2k2. This would be 2k4 series resistance for the Tune LED, but the difference in brightness from the FM one is insignificant.

Second close-up is the coupling capacitor swap for the TDA2822 and for the speaker output. Upon closer inspection, the 2822 had a 10k resistor to ground just like in the datasheet, but the coupling cap was just 220n! This was a cutoff frequency of 72Hz which had room for improvement. More bass? Yes please. I put a 1u lytic instead, i found one small enough to fit the cramped space (the speaker magnet comes right over that area). Now the cutoff frequency is 16Hz, more than enough for FM, highpass filters in broadcast equipment are usually set to 30Hz. Now this baby packs some serious punch on headphones.

Of course, the lowering of the cutoff frequency of the amp raised a new issue - the tiny internal speaker would get overdriven from all the bass. 2x 100uF in parallel as the coupling capacitor for the speaker took care of that - the tone remained exactly the same, while the output gets louder AND cleaner than with the speaker wired directly. I also measured about a 1/3 reduction in the current drawn by the speaker. This probably uses less power than even the stock config - the speaker is 4 ohms, so 200uF in series with that gives a cutoff frequency of 200Hz, which is more than one octave higher than the original. The original 72Hz frequency was likely chosen as a compromise between bass response on headphones and battery life on speaker.

Then i put the thing back together, and Murphy got me. The LCD display went incomplete on me. I tried everything to no avail, it would go incomplete when the unit was assembled. Since the LCD is connected via one of those carbon/plastic ribbon things, it sure seemed like the end of it, didn't it? Well... a bit of thinking out of the box fixed it. The ribbon HAD to be stuck in place by heat, there was no other way. Here's where a thermally controlled soldering iron comes in handy. I turned the iron to low, so as not to burn thru the ribbon. Then using a wide tip i pressed on the intermittent connection for a few seconds, and hey presto, it's now making solid contact. I sure wish i hadn't thrown out that nice scientific calculator which had the very same issue.

There are two other mods i did today, that i didn't take pics of. The headphone output was wired in a really weird way - they had a trace running near where the ground pin of the headphone jack was, so they left that unconnected and instead wired the output to the L and R pins. That way the headphones end up in series, but also out of phase, so when listening with both cans it sounded interesting to say the least. I changed the wiring to normal mono config with both channels in parallel. With that, i also get 4x the power output, as if it wasn't loud enough already. Mo' bass fo' yo' face.
I also added a coupling capacitor from the antenna terminal to the headphone jack, because i carried it with me today and the reception sucked when on the move. We'll see tomorrow if this did any good.

[Tube_Dude]

Hi, Th3_uN1Qu3

I really appreciate your creativity. Keep up the good work!

Cheers.

[goodpsusearch]

This is genius! Well done!

[momaka]

Th3_uN1Qu3, your threads are always full of cool stuff to read . Thanks for posting this.

[shovenose]

Creative and Interesting!!