home *** CD-ROM | disk | FTP | other *** search
-
-
- MODEMPOWER
-
-
-
-
- ... or how ç00l people say it
-
- M0ðemPoWa!
-
-
- ;-----------------
-
- SECTION Deny_all,DATA
-
- I WON'T ACCEPT ANY RESPONSIBILITY ABOUT DAMAGES CAUSED BY USING OR MISUSING
- THIS INFORMATION. YOU DO IT ALL AT YOUR OWN RISK.
-
- Actually, this (as all of my hardware projects) includes only the idea of
- the circuit. They're for people who do have some EXPERIENCE of constructing
-
-
- ;-----------------
-
- SECTION What,DATA
-
-
- So, whatta h*ll is this that you, on your weak moment, happened to download?
- Maybe I'll tell a story...
-
- Once upon a time :) there was a hacker. He was going to study in Oulu
- Institute of Technology. He knew, that he was going to get his internet access.
- So, he bought a modem, an external one, because he had no card slots in his
- beloved computer. But, with the modem came also evil AC/AC-adaptor, that made
- noises, mostly at the night. Bzzzzzzzzzzzzzzzzzzzzzzzzzzz.... all night long.
-
- Hacker was unhappy. He was tired every morning after being awake all night
- because of that evil AC/AC-adaptor. He got p*zzed off. One day, after eating
- a delicious, plenty of vitamins containing meal at his school's cafeteria
- (actually the food makes people having hallucinations :), he had an idea!
- He has a 200 watts power supply in his computer! Plenty of power to get a
- single modem running.
-
- So, he tried to create a circuit, that converts +12V into 12VAC. He deliberated
- upon it... night after day, day after night... until he discovered that a
- power-audio-IC could be used as oscillator and charge pump for -12V! He made
- first trial circuit based on a TDA2030 power IC. ... and it didn't work. IC
- either oscillated at about 200kHz or didn't oscillate at all. Poor little guy
- went to bed... sad... sooo sad.
-
- Next day he began thinking the circuit over. This time he put, not two, but
- three RC-timebases to feedback loop of the IC. It worked! It oscillated
- promptly at 50 Hz, output being nice square wave. Eagerly he continued
- building the charge pump and ±12V switching circuits. Later that night circuit
- was just waiting for a test. During the test, however, become quite clear,
- that the IC couldn't deliver enough voltage during +12V phase. So, the poor
- little guy went to bed... sad... but not SO sad.
-
- Next day he was thinking about ways to make the circuit capable of delivering
- more voltage, and if possibe, current. Like a thunderbolt, idea of putting
- a PNP transistor to connect original +12V to output instead of drawing it
- from the IC stroke his mind. He put the transistor in and... Yes! +12V at
- output was steady as rock, no matter how difficult load was connected.
-
- After making some optimizing and further tests he connected ß-version of
- the circuit to his computer and modem and went for a little ride into the
- Internet....
-
- Next day he made a little metal case for the circuit, reconstructed
- the circuit into the case and connected it in. After testing it to work he
- pulled out the BUZZZZZZZZZZZZZZZZZZing AC/AC-adaptor and abandoned it to
- a dark, cold corner of his cupboard and lived happily ever after.
-
-
- Hmm....
-
-
- khm....
-
-
-
- well.....
-
-
-
-
- OK! OK! That 'poor little guy' is ME!
-
- Yes! That AC/AC-adaptor BUZZZZing all night p*zzed me off!
- Yes! I DO have a 200 watts PC power supply in my Amiga!
- Yes! It took me some time to get this hack working!
- Yes! It's working right now... in fact, I uploaded this package to Aminet using
- my modem powered by MoDeMpOwA.
-
-
- ;--------------
-
- SECTION More_Serious_Text,DATA
-
-
- So? How does this.. thing work?
-
- <picture: ModemPower.pic1>
-
- Main component is, indeed, an audio-power-IC TDA2030. Its positive input is
- connected to half of input voltage. Negative input is connected to output
- via three phase-shifting RC timebases. When +12V is supplied to pin 5 of
- the IC, pin 1 (+ input) raises to +6V. Positive input is higher than negative,
- so, output will go high, to nearly +12V. From output current flows via 100k
- resistors to 100n capacitors. First capacitor takes a little while to charge,
- when it's charged, current flows thru second resistor to second cap and
- finally thru third resistor to third cap and to negative input of the IC.
- Negative input raises higher than positive input, so, output will go low,
- nearly to 0V. Negative input follows it after a while, and so on.
-
- <picture: ModemPower.pic2>
-
- Square wave in output pin of the IC will be fed into charge pump, that is
- 2200/16V electrolytic capacitor and 1N4001 diode. When square wave goes up,
- diode keeps cathode of the elco below 1 volt. About ten volts is charged into
- the elco. When square wave goes down, cathode of elco goes down to about -9V,
- that is connected to output of circuit via Q2. That happens, when square wave
- is low. When it's high, Q3 gets it base current and draws base current from
- Q1, that connects original +12V to output. At the same time, Q4 gets its base
- current and supplies base current for Q5, that prevents Q2 from leading.
- A 47n capacitor from output to ground is for suppressing RF disturbance.
-
- ;------------------------------
-
- SECTION Before_Building,DATA
-
- First! Make sure that your modem has either one of its power connector's pins
- connected to GROUND of modem's RS-232/eqv. connector. That is vital, because
- this circuit relies upon that. Otherways, some smoke will be generated.
-
- I'm using this ...thing with my Well XM-2814SAM external modem. It seems to
- work fine. I tested MoDeMPoWa with 1A positive half-wave load and 150mA
- negative half-wave load. Worked fine. I tested powered my modem by MoDeMPoWa
- and lowered input voltage. Problems occured when input voltage was as low as
- 8 volts. So, it seems to be enough headroom there...
-
- About constructing the circuit.. do it as you like. It's wise to add a 1.25A
- fuse to supply line of this circuit, in case something goes wrong (accidental
- short-circuit between output and ground, that's quite possible).
-
- Custom-made PCB would be nice, but a piece of vero board will do just fine.
-
- Those power transistors won't need any cooling, they are just switches. The IC
- needs a little cooling to prevent IRCing to stop because of overheating.
- If the circuit is placed to a metal case, just attach the IC to the case.
- IC's metallic part is connected to pin 3, that is, the ground.
-
- You do remember to connect GND of ModeMpOwA to the ground-connected pin in
- power connector of the modem, don't you? Otherways, the fuse recommended
- earlier comes in handy.
-
- Well, I think that's enough guidelines for you to construct the .. thing.
-
- Long live and prosper, sano savolaenen.
-
-
- Send snails to:
- Arto Niskanen
- Valtatie 8A8
- 90500 OULU
- ..and that's in FINLAND
-
- ..and electrones to:
- archy@rhea.otol.fi
-
-
-
- ;----------------
-
- SECTION ADDENDUM,DATA
-
- <picture: ModemPower.pic3>
-
- Just added some information about creating a -8VDC voltage with this... thing.
- Might be CøøL for those tower-builders.. as nowadays seems that PeaSea-
- Power'supplies' don't output any negative voltages. To make an Amiga work
- with that kind of... PeaSeaCr*p'Power', use e.g. this to make serial port(s)
- and audio output to work.
-
- Principle is simple, a diode is used to grab that -9V part of a square wave
- and an elco to filter it... This might supply a few hundred milliamps.
-
- Nzoi!
-
- 17.12.1995 Archy
-
-
-