| How I Fixed Some Multimedia Speakers | Geek |
It's no secret that I struck it lucky, and I wish to share all I learned.
We had some misbehaving self-powered multimedia speakers, connected to a computer. They had been working fine, but had been left on for a while, which was an unusual thing for us to do. They were turned off at some point, but when they came back on, they suffered from a huge amount of mains hum and the signal sounded very faint.
Later on, the hum was replaced by a slow phasing sound and none of the sound from the computer was coming out at all.
It was up to me to fix these naughty speakers, so I took them to my garage.
What was wrong with these speakers? Would they ever work again? (If you've read the title, you'll not be guessing too much on that one.) How do you go about finding what's wrong and fixing it.
Take the speakers to pieces. Then, look at each of the pieces as if you know what sort of thing might be wrong with a circuit board, transformer and collection of assorted wires. Then scratch your head a lot. Then inspect the circuit board closely - looking for microscopic traces of either burning or cracked tracks/solder joints.
Admit it. You have no idea what's wrong with them... PLUS you cannot really test them just by looking. It's time to locate the multimeter.
Test the multimeter - it makes a noise when you touch its probes together - that's a start. Then, despite the temptation, don't bother testing each resistor to make sure that it's still conducting... they're not really going to go wrong. I occupied myself for a while in this activity and it's pointless.
In a vain attempt to reverse engineer the circuit board, start back with the power supply - is there a voltage across the power lines when the unit is on?
Its wee LED was coming on - a good start - and there was definitely 12volt across the diode bridge/rectifier thing. So, 12 volts coming in and 12 volts across the smoothing capacitor after the rectifier... we have power.
More general poking about. In the end, having tested continuity all over the place and decided that everything was connected, and having checked that the potentiometer, used as a volume knob, also did its job, there's only one thing to ask? What could be the single point of failure?
You see, neither left nor right speaker was working. This actually narrowed it down. It could only be the power supply, one of the controls... or...
Mmmm - what's this chip in the middle of the circuit board? TDA 2822 - it must be the amplifier chip - some sort of Op-amp perhaps. Could this be the thing that fried? or could the problem be in the myriad other components - capacitors and resistors dotted around?
A bit of a guess here, but I decided that the TDA 2822 chip had fried and that I would be able to replace it and fix the speakers. It seems as good a thing to try as any. I tried to sketch a wee circuit diagram, showing how the chip related to the inputs, outputs and common ground of the circuit board.
Then, it was onto the internet. I quickly found that google gave hits for the TDA 2822, but that was not enough. So, I searched Maplin's website. I reasoned that the only thing which would be better than knowing that the chip was not proprietary to the speaker manufacturers (a company called "Generic Made In China") would be knowing that I could go out and buy one. I was in luck - £1.50 for a TDA 2822M - I wrote down the Maplin part number for it.
Maplin's site suggested an operating voltage of 9 volts. This worried me - either the old chip had fried because the voltage was too high (perhaps the manufacturer upping the input for extra power to increase the price) or for some unknown reason - I wanted to look in more detail. I found a data sheet. I have included the data sheet on this site because I can't remember where I got it on the net. The TDA 2822 comes in two forms - an 8 pin and 16 pin version - it was the eight pin I needed. The data sheet included a description of the 8 pin, showing how its pins' purposes matched those I'd guessed from reverse-engineering the circuit board, and also showing an operating voltage up to 15 volts.
To be more exact, the TDA 2822 is a fixed gain stereo amplifier, whose documentation showed a demonstration of how to use it, at around 9-12 volts, as a way of driving speakers. The circuit diagram in the data sheet matched the design of the circuit board in my speakers.
So, I drove to Maplin, paid them £1.63 - I bought a wee DIL socket from them, so I didn't have to burn my chip out while soldering it directly to the board. Then it was back into my garage.
Well, although I made a little mess of two of the tracks, which came away from the board because I could not remove the solder from them easily enough, I managed to solder the DIL socket into place and connect it in all 8 places it required connections - jumper wire is a real boon to the bodger.
Plug in chip, turn on speakers...
... it bloody worked! I couldn't believe it!
It's okay to be the sort of geek who wanders into Maplin and asks for a specific chip, by Maplin part number, if you're then able to go home and fix something.
I wonder how long the speakers will continue to work? I really had no idea what I was doing.
02 April 2002
Ashley Frieze