SparkFun Forums 

Where electronics enthusiasts find answers.

General project discussion / help
Did you make a robotic coffee pot which implements HTCPCP and decafs unauthorized users? Show it off here!
By snowman815
#172139
Edit: 6/22 6:28am (Added drawing of wiring)

Hello and thanks for taking the time to read this,

I was looking to increase the audio volume and clarity in my project so I just swapped out:

1x - TPA2016 3w amplifier from Adafruit (https://www.adafruit.com/products/1712)
2x - 3" 4 Ohm 3 Watt speakers from Adafruit (https://www.adafruit.com/product/1314)

in my project for:

1x MAX9744 20w amplifier from Adafruit(https://www.adafruit.com/products/1752)
2x - 20W 4 Ohm Full Range Speakers from Adafruit (https://www.adafruit.com/product/1732)

Nothing else in my project has changed other than to code to control the gain of the amplifier which is very simple. You just write what gain you want via i2c,

Strangely though, I'm now getting LESS volume out of the 20w speakers than I was from the 3w speakers with the gain set to max (60). I'm getting almost no volume at all until up around 30 or so too.

The audio signal is coming from Sparkfun's Musical instrument shield. (https://www.sparkfun.com/products/10587) It's a shield but I'm not using it as one since it only uses 4 pins. I soldered headers only onto the pins that are actually used and tied those to where they belong on my micro controller (teensy3.1).

I'm also getting a fairly loud high pitched hum from the speakers. I assumed this is a grounding or interference issue so I tried using the analog ground on the amp without any improvement. I'll post more info about what I've tried as I do more tests.

I tried removing the Musical Instrument Shield and just connecting an iPod via the headphone jack to the amp and it worked beautifully. No hum, uncomfortably loud. So it isn't a problem with the amp in general.

Does anyone have any ideas as to why I would be getting less volume and more noise out of this 20w amplifier then I did with a 3w amplifier? I'm pretty new to audio projects so even the most basic of advice would be very welcome. Thanks again for taking the time to read this. Your help will be much appreciated.

I've attached a picture of how everything's wired. It's pretty simple.

Thanks,
Nick
You do not have the required permissions to view the files attached to this post.
Last edited by snowman815 on Sun Jun 22, 2014 6:32 am, edited 1 time in total.
By Mee_n_Mac
#172145
snowman815 wrote:Also, I have a picture drawn up of how I have everything wired, is it possible to upload images on these forums or is it external only?
If you use the full editor you'll notice an upload tab below the entry box. Click on it and do the obvious to U/L a pic.
By Mee_n_Mac
#172173
First thing I'd do (if you haven't already) is get an idea of the relative efficiencies of the 2 sets of speakers. Using some setup, old or new or whatever, set the volume and compare the sound level from one set to the other.

Was the Sparkfun Musical instrument shield in both the noisy (20W) and not-noisy (3W) setups ?

Was the power supply the same in each ? I assume the batteries depicted are really the 12V switcher ?

Did you wire up the analog volume control on the 20W amp ?
By snowman815
#172175
"First thing I'd do (if you haven't already) is get an idea of the relative efficiencies of the 2 sets of speakers. Using some setup, old or new or whatever, set the volume and compare the sound level from one set to the other. "

- At max volume with a single note being played from the midi shield the 20w speakers produce slightly less sound than the 3w speakers. Everything except the speakers remained unchanged (12v supply, 20w amp.

"Was the Sparkfun Musical instrument shield in both the noisy (20W) and not-noisy (3W) setups ?"

- I actually just solved the noise issue but I'm not sure why it works at all now. I removed the jumper from the ground of the amp going to the controller and the noise disappeared. I didn't that that I would still be able to communicate with the amp via i2c without a ground connection.

"Was the power supply the same in each ? I assume the batteries depicted are really the 12V switcher ?"

- In my previous 3w setup, the power supply was 5v. WIth the new 20w setup, I'm using a 12v 60w "laptop brick" style power supply.

"Did you wire up the analog volume control on the 20W amp ?"

- My project demands digital control of the amp so I haven't tried using it while set up for analog control. Do you think trying that would help?


I've also asked for help over on the Arduino forums and one of the suggestions was to try reading the AC voltage of the output of the midi shield with the amp disconnected. With the midi shield playing a single note at max volume every 20 milliseconds I measured the AC voltage on the shields output at around 200 mV. I tried reading DC voltage too (just because) and got around 200 mV as well.

I haven't heard back yet after posting my findings about an hour ago but it seems like the reason I'm getting roughly the same volume from the 20w set up is that both amps have the same amount of max gain, 30db. I only looked at 3w vs 20w and thought it would be louder like a noob.

A suggested solution was to use an op amp to "pre amplify" the signal to the amp. This makes sense now but op amps are something I'm not familiar at all with and at a glance there seems to be a large variety of types of op amps available. Is there any that you would recommend for this application?

Thanks a bunch for your help too by the way. It's greatly appreciated.
By Mee_n_Mac
#172177
[quote="Mee'n'Mac""]"First thing I'd do (if you haven't already) is get an idea of the relative efficiencies of the 2 sets of speakers. Using some setup, old or new or whatever, set the volume and compare the sound level from one set to the other. "
snowman815 wrote:- At max volume with a single note being played from the midi shield the 20w speakers produce slightly less sound than the 3w speakers. Everything except the speakers remained unchanged (12v supply, 20w amp.
[/quote]
OK, so the older speakers are a tad more efficient and therefore louder at most volume setting that the new ones. They may, or may not, clip before the new speakers and that may limit their max volume ... if you can drive the speakers that hard.
snowman815 wrote:- I actually just solved the noise issue but I'm not sure why it works at all now. I removed the jumper from the ground of the amp going to the controller and the noise disappeared. I didn't that that I would still be able to communicate with the amp via i2c without a ground connection.
I'll bet that your audio L/R cable also carried a ground. Thus you had 2 ground paths, often but not always, a bad idea. I'd be wary of sharing digital circuit grounds w/analog circuit grounds. It's too easy for digital noise to sneak into audible paths, even w/careful designs. If what you have now works ... keep it and declare victory !
snowman815 wrote: I only looked at 3w vs 20w and thought it would be louder like a noob.
Max power output should correlate to perceived loudness but only if the signal level is high enough to drive the output to max levels. As you've seen, equal levels mean equal loudness ... w/the understanding that your old setup was a bit louder at any given level due to the speakers.
Comparing the wattages (assuming, ugh, that they have any real meaning) suggests there is but a little increase in true loudness if the new speakers were to be driven at the max level by the amp (assuming the speakers don't choke before that). 20W is some 8.2 dB "louder" than 3W. IIRC a 3dB increase is just barely discernible as being louder. So 8.2 dB won't be that much louder, and perhaps not that noticeable given the relative speaker efficiencies.
What you may notice is a lack of distortion at the higher volumes. When amps clip, they introduce "noise" at higher frequencies. This may be mistaken for even more volume, as human hearing is both more sensitive to vocal frequencies as well as being generally logarithmic wrt to loudness.

Then again if all you listen to is rap ... you may never hear any difference.