Rigged up a headphone amplifier, trying to reduce noise

[Mrx3750]

As a few of you might know, I was looking to make a headphone amplifier. I thought about it and figured for about the same amount of money I could just get a cheap pair of USB computer speakers, remove its circuit board, and attach a headphone connector, in essence making a USB powered headphone amplifier.

So far it's been a success, in that it functions. However there's a high amount of background noise.

The pics I attached are a bit blurry, it's hard photographing a board that's so small. I can tell you it uses a TDA2822 IC for its amplifier. The pair of pink and white wires are left and right outputs, while that white wire connected to one of the legs of the power switch is serving as ground for the output.

From what I've read, others have made headphone amps using this IC with no issue, so I'm guessing either my noise issue is due to crap board design, or possibly a noisy power source. (The background noise did sound different when I plugged it into my USB phone charger.)

[budm]

Re: Rigged up a headphone amplifier, trying to reduce noise

Did you put 47 ~150 Ohms resistor in series with each headphone transducer?

[Mrx3750]

Re: Rigged up a headphone amplifier, trying to reduce noise

I haven't modified the board at all yet, except to rig up a ground for the output, seeing as I could not find out where it normally grounded.

Are you saying there should be a couple resistors in series with the output leads?

[lti]

Re: Rigged up a headphone amplifier, trying to reduce noise

That looks like a mono amp, not stereo. There was no output ground because a single speaker is connected across both output pins. Look at the "bridge" schematic in the datasheet. You'll need a second identical amp for stereo.

Also, the TDA2822 chips (actually, the TDA2822 is a larger chip with a ton of unused pins and the TDA2822M is what you have) are known for large amounts of output noise and no power supply ripple rejection. Any ripple from the power supply gets amplified and sent to the speakers.

[budm]

Re: Rigged up a headphone amplifier, trying to reduce noise

I have seen the really cheaply made PC speaker that is actually mono with the Left and Right inputs are tied together to just single amplifier that drives 2 speakers, but box and the manual claim that it is stereo PC speakers.

[Mrx3750]

Re: Rigged up a headphone amplifier, trying to reduce noise

Huh you guys are right. (I only tested it for a few seconds) Just did a stereo test, and it's not stereo.

Welp, that's got me pretty pissed off. I've seen some cheap shit in my time, but this takes the cake.


You know, I have an idea. This may sound crazy, but I do still have those LM324 IC's. It'd be a mess of wires, but I might try rigging one up to this board just to see what happens.

Thinking about it more, I could route the stereo inputs to the LM324, likely with a 4 wire cable, then connect the headphone jack to the output pins, and just see what happens.

[ben7]

Re: Rigged up a headphone amplifier, trying to reduce noise

Huh you guys are right. (I only tested it for a few seconds) Just did a stereo test, and it's not stereo.

Welp, that's got me pretty pissed off. I've seen some cheap shit in my time, but this takes the cake.


You know, I have an idea. This may sound crazy, but I do still have those LM324 IC's. It'd be a mess of wires, but I might try rigging one up to this board just to see what happens.

Thinking about it more, I could route the stereo inputs to the LM324, likely with a 4 wire cable, then connect the headphone jack to the output pins, and just see what happens.

Another thing that may piss you off: TDA2822 are not hi-fi, and they are noisy. So yeah, you really need a different amplifier chip which is actually made for driving headphones.

[Mrx3750]

Re: Rigged up a headphone amplifier, trying to reduce noise

Yep. Scrapped the idea. Pretty pissed because I can't return the speakers, and $12 ain't worth raising a stink over.

Still maybe I should tell the store that those speakers aren't stereo.

In the meantime, what would be some good IC's to look out for? I run across junk PC speakers every now and then.

[lti]

Re: Rigged up a headphone amplifier, trying to reduce noise

You could remove the chip and put it on another board with the circuit matching the stereo circuit in the datasheet. It will be kind of hard to remove the chip, but you can then put it in a DIP adapter or just solder thin wires to each pin and run them through the solder holes on the new board like I did in my headphone amp thread. However, it will still have a noisy output. Putting resistors in series with the output might help. I might try that with the cheap computer speakers I have that use the same chip.

Speakers with a headphone output will probably use the same chip or some other low-power amp like a TEA2025B. They could also use a dedicated headphone amp chip like a TDA1308 or one of its Chinese clones.

People do try to use op-amps as headphone amps, but they don't work very well.

[Mrx3750]

Re: Rigged up a headphone amplifier, trying to reduce noise

I threw the board away after trying to rig up an LM324 IC. better and saner just to try again with another, better built, speaker board.

(Seriously I have seen some cheap junk, but mono trying to pass off as stereo with fake tweeters is a new low)

[lti]

Re: Rigged up a headphone amplifier, trying to reduce noise

(Seriously I have seen some cheap junk, but mono trying to pass off as stereo with fake tweeters is a new low)

That sounds like something Cyber Acoustics would do, except they would try to tell you that the speakers are worth $25 and add a bunch of high-brightness blue LEDs as power indicators.

I don't understand how some of this cheap Chinese stuff works. I saw a $1 Chinese radio that had one IC (an SC1088), one transistor (with no markings), three inductors, two buttons, and a bunch of tiny ceramic caps (the largest value was 0.047uF). I have no idea how it worked, but it did. The circuit was even more confusing than those "CMoy" headphone amp circuits. Maybe someone can post some of those circuits so people with experience with op-amps can .