I've found that if you grossly overdrive the board, it shuts down (requiring a power cycle). It also has an output short circuit detection circuit (which shuts it down as well), I'm sure it's not a "thermal issue", since the chip runs cool even at full power.

Yes! All the headers are on a 0.1" grid. One wouldn't normally break out the PTH gain resistor replacement holes, but if you need to, they are on-grid as well.

This amp shouldn't have a problem driving either speaker (or even much larger ones). Is it very quiet or is there no sound at all? In case your 2" speaker is defective, can you try a different one or measure any resistance across it? Contact our tech support department, they'll be happy to help you out.

I've found that any reasonable amp-to-speaker wire length radiates a ton of RF noise, detectable on AM and SW radio receivers. I made a little filter board using two 68uH series inductors and a parallel 1.0 uF capacitor on a tiny board, mounted directly to the amp board. This allows the use of longer speaker wires without spraying the whole room with RFI... and it doesn't effect the audio quality one bit.

It should work at "any" distance, but because the speaker signal is a high-power medium-frequency signal, longer runs could cause interference in nearby sensitive devices (radios, etc.). In an ideal world each speaker would have an amp close to it, and the input side would be the long one. For whichever side is the long one, do try to use shielded or twisted wire (CAT5 is cheap and pretty good for this) to minimize transmitted or received interference. Let us know how it works!

Yes, it works great with the big transducer!

Working on getting that fixed now! Sorry about that.

Whine is sometimes a ground loop problem. You could try connecting the "-" input to ground and see if that helps.

Other than that, if your input is a PWM signal, it may be interfering with the amplifier's own PWM frequency, in which case you should try low-pass filtering the input. There are filter calculators out there that could give you a better answer, but you can start with 100 ohms and 0.1uF, which will pass everything below 10kHz. Low-pass filtering the output is harder because it is a high-power signal and needs high-power filtering components. Good luck!

The board is 0.8" x 0.8". The holes are on 0.1" centers; the two input/output rows are 0.7" apart. We highly recommend that you install Eagle if you're working with our boards. The free version will open all of our files and answer questions like this and many more.

Any answers would be rough because it depends on the specific audio and what constitutes a reasonable volume. For a worst-case analysis you could use Ohm's law: 3 volts into 15 ohms will be 200mA (ouch), which would drain your battery very quickly (15 minutes). In practice you should get more time than that because you'd never be driving the speaker all the way, but I doubt you'll get more than an hour out of the coin cell.

This amplifier will work well for that transducer. But it can't make audio louder without a source of power, so you'll need to supply one. You might be able to tap off the batteries or wall wart in the DSi. Or you could use a separate battery holder like this one.

You should be OK at lower volume levels, but you won't be able to crank it up all the way without overloading the 5V regulator on the Uno.

The longer answer is that the Uno regulator can supply about 500mA max, and can thermally support 1W before it gets too hot and shuts down (it will turn on again when it cools off, so you won't hurt it). Ohm's law says that 5V into a 4 ohm coil will draw 1.25A = 6.25W, which is a bit much for the Uno. However that's for a constant load; audio tends to be more bursty, so you should still be able to get things working. If you run into serious problems, an external 5V regulator like the 7805 will give you 1.5A if it's heatsinked. Have fun!

It's 0.84 x 0.84 inches for the board size. You can also find the exact measurements of all the components in the Eagle files.

Oops, sorry about that! Here at SFE we will sometimes design footprints in our own libraries, and only make them public after they're proven. I'll move it to the big repo today. Also note that there's a ULP (User Language Program) called "exp-project-lbr.ulp" or "exp-lbrs.ulp" depending on the Eagle version, that will create a new library from all the parts in a project. You can then edit parts or copy them to other libraries as you wish. This is very handy if you can't access the library that was used to make the project.

You're not likely to burn out our 0.5W speaker with this amplifier and "normal" audio duty cycles (the effective wattage is going to be quite a bit lower than the maximum). But you're right that it's not going to be terribly loud by itself. Some ideas for increasing the volume: Using larger speakers is the most effective solution (we've had a lot of fun with used automotive speakers), though this can be hard if space is limited. You can increase the number of speakers as long as the total resistance is between 4 and 16 ohms. And small speakers will sound much better if you can add some sort of enclosed cavity behind the speaker(s). I hope this helps, good luck and send us pictures!

You've got me... I know that apple equipment uses four-conductor jacks, which might interfere with a normal stereo plug, but the iPhone has the same jack and you say it's working fine... Could it be a ground-loop problem, since the iPhone is battery powered and isolated?

Exactly right, those vias help heat get from the top of the board where the amplifier chip is, to the bottom of the board for extra heat-radiating surface area, as well as making sure the ground plane is solid as this chip can use quite a bit of power. It's so efficient it typically doesn't even run warm, much less hot.

They worked for me just now; these were made with Eagle 6 (earlier versions won't open the new format). If that's not the problem try downloading them again, it may have just been a glitch in the Matrix.

EDIT: Kamiquasi beat me to it. ;)

Could it be that you're using an older version of Eagle? I get that error in Eagle 5.11.0, but 6.1.0, 6.2.0 and 6.4.0 all load it fine. If this does appear to be the issue and you're not able to upgrade, I could give making a 5.11 version a shot. I really recommend upgrading, though.

Yes, it can push more power at 5V than 3.3V. (Ohms Law is a harsh mistress.)

It sounds like you have everything hooked up correctly. At this point you have two options; make your source louder, or increase the gain on the amplifier.

The VS1053 used in the Music Instrument Shield does have a volume control register that can make it louder, but unfortunately you can't access that register from the Music Instrument Shield (the MP3 Player Shield, based on the same chip, has a more complex interface that does allow this). So you're correct, in your case changing the MIDI volume is about all you can do to make the source louder.

If that's not enough, you can increase the gain on the amplifier board by adding your own gain resistors. The Quickstart Guide at the top of this page has instructions. This will make the amplifier considerably louder and I think will solve your problem. Let us know if you have further questions!

The best thing to do is to find a pinout diagram for your jack (is it a Sparkfun part?). You could also test it with a multimeter to figure out which pins are which:

Without anything plugged in, two pins will be connected (0 Ohms). Those are the switched pins. The other pin will be your ground.

Now plug in headphones and measure between ground and each of the other two pins. One pair should read 50-ish Ohms; that pin will be your audio input. The other pair should read as disconnected, that will be the pin that goes to the amplifier input.

Many headphone jacks have internal switches that are connected when nothing is plugged into the jack, and disconnect when you plug in headphones. This lets you cut off the amplifier when you plug in headphones.

If you look at the datasheet for the above jack, pin 1 is the ground, and pins 2 and 5 are the L and R inputs. It doesn't explain this very well, but pins 3 and 4 are the switched outputs to your amplifier (2 and 3 are paired, as are 4 and 5). Connect your mono audio signal between 1 (GND) and 2 (signal), then connect 1 (GND) and 3 (switched signal) to your amp input.

You should be able to connect either the shutdown or the input (ground the second input) directly to an Arduino, no series resistor needed.

Interesting question, and not stupid at all. According to the datasheet, the switching frequency of the amplifier is 250kHz, which I suspect (but I could be wrong) is too high for an XBee (I believe the max serial rate of an XBee is 115200bps, and the maximum virtual wire rate is 1kHz). And as you suspected there are other devils in the details, such as the fact that this chip sends out slightly different waveforms to the + and - outputs to improve efficiency; ideally you'd want to send both signals, which would double the required bandwidth.

A more efficient wireless system might use an ADC on the input side to rapidly sample the audio waveform, send the binary data across the wireless link, and use a DAC to restore it on the other end. You'd want to sample at twice the highest audio frequency (20kHz x 2 = 40kHz), which the XBee can theoretically handle.

If you try any of this, best of luck and let us know how it goes!

Sorry for the late reply. We've successfully tested this board with 4 Ohm and 8 Ohm speakers (and various combinations in series and parallel), so it should work fine with a 6 Ohm speaker.

Yes, and yes!

I guitar is about -20 dB, which is not line level (0 dB), you'd need a pre-amp stage.

If the guitar output is line-level audio, and you want to drive a speaker, then yes it should work well.

There are two things that affect how loud your project will be; the power of the amp, and the size of the speaker. If you hook up a 6" or 8" speaker to this amp, it will be fairly loud (especially if you put smaller gain resistors into the amp to increase the volume), but that speaker might be larger than you want for something that goes on a handlebar. Give it a shot, and let us know how it works!

It would work, but there are headphone amplifier circuits out there that are better suited for that task. This board is primarily designed to drive larger loudspeakers.

Unfortunately no, the output of this board is designed only to be connected to a speaker. But is there a reason you couldn't connect your line-level source directly to your AUX input?

SDN* stands for "shutdown", if you leave it disconnected or connect it to a digital high signal, the amp will function normally. If you connect it to ground or a digital low signal, it will turn off the amplifier and LED to save power.

Thanks for the feedback! SFE leans towards having no mounting holes on our smallest boards, because it reduces costs (PCB real estate is surprisingly expensive), and also on the theory that most people will be plugging these into a breadboard or larger circuit board. It's also common for people to stick boards down with double-sided foam tape, hot glue, etc. That said, we're listening, and take your comments into account when designing new products. Thanks again!

That's exactly how it's done. The LED will indicate whether the board is active or shut down, so you'll know if you're doing it right.

Sure! The MP3 Shield has a "Speaker" header with -, R, L pins. Connect the - pin on the Shield to the - in pin on the amp. Connect either R or L on the Shield to the + in pin on the amp. That's it for the audio connections. You'll also want to run power to the amp (5V if you have it, 3.3V if you don't), and connect the amp to a speaker (at the out + and - pins). That should be all you need, ask if anything's unclear.

0.8" (20.3mm) square. If you're ever curious, you can always get exact measurements of our boards by opening the Eagle files.

I just tried it, it works perfectly and is VERY LOUD.

The LED follows the shutdown signal, so it's off when the board is shutdown. A very small amount of power is bled through the pullup resistor, but if you're really into nA, you can remove those components.

We could have made it smaller, but thought that the PTH resistor pads were a good feature to have. And just how small do you need it? ;)

We do mess up from time to time, but contrary to popular belief, we do closely read the datasheets. This board was routed with those recommendations in mind, balanced with meeting the additional requirements of the design (supporting PTH pads for gain-resistor replacement and a header for an optional external volume control pot). The diversions are minimal, and in any case the board is less than an inch square. Finally the proof is in the pudding, as the board performs extremely well. Try it, we think you'll like it.