Author Topic: Simple speaker hack produces poor sound quality  (Read 4815 times)

0 Members and 1 Guest are viewing this topic.

Offline EDPRTopic starter

  • Contributor
  • Posts: 11
Simple speaker hack produces poor sound quality
« on: September 11, 2016, 04:16:21 am »
My excitement was big when I found a big fresh pile of 15 projectors at (next to) my desk.  Just 3 days after my initial inquiry they were already there.  The purchasing department seemingly made a good deal somewhere.

The idea was a cool one; get some Apple TV's, some beamers, speakers and put them in our guest rooms.

So I started a test environment.  Hooking up the Apple TV, the beamer, the external speake.... oh dear... no audio output on the beamer, just the internal speaker.

The internal speaker featured rather impressive sound for its size (< 2").  But far from enough for a movie or game.

Tried to find out what my options were:
1) Return the beamers and get other ones.  This was absolutely not possible. 
2) Split the HDMI signal into seperate audio and video.  Possible, but not user friendly.
3) Open the projector, disconnect the speaker and create my own audio output.

I went for the 3rd option.  Opened up, soldered a 3.5mm audio socket and ready to test.  The sound was awful, there was a huge distortion even on lowest volume. 

The internal speaker says
Code: [Select]
MT
4 ohm   5W
534
and has 3 cables connected; red to +, black to -, and white to the ground of the speaker.

Is there an easy solution for getting decent sound out of this speaker cabling? 

 

Offline alsetalokin4017

  • Super Contributor
  • ***
  • Posts: 2055
  • Country: us
Re: Simple speaker hack produces poor sound quality
« Reply #1 on: September 11, 2016, 04:26:13 am »
What kind of external speaker are you trying to connect? Does it have a built-in amplifier or is it simply an ordinary passive speaker?
The easiest person to fool is yourself. -- Richard Feynman
 

Offline jonovid

  • Super Contributor
  • ***
  • Posts: 1425
  • Country: au
    • JONOVID
Re: Simple speaker hack produces poor sound quality
« Reply #2 on: September 11, 2016, 04:31:44 am »
image's ? take some of it & or schematics circuit your using?   :-//
« Last Edit: September 11, 2016, 04:39:54 am by jonovid »
Hobbyist with a basic knowledge of electronics
 

Offline EDPRTopic starter

  • Contributor
  • Posts: 11
Re: Simple speaker hack produces poor sound quality
« Reply #3 on: September 11, 2016, 04:36:10 am »
What kind of external speaker are you trying to connect? Does it have a built-in amplifier or is it simply an ordinary passive speaker?
The powered ones, with built-in amplifier.
 

Offline StillTrying

  • Super Contributor
  • ***
  • Posts: 2850
  • Country: se
  • Country: Broken Britain
Re: Simple speaker hack produces poor sound quality
« Reply #4 on: September 11, 2016, 04:37:22 am »
Yeah, you don't say whether you're connecting a speaker or an amp, if using an amp it might just be needing some resistive load.

has 3 cables connected; red to +, black to -, and white to the ground of the speaker

Could be bridge connected, class D, or both!
.  That took much longer than I thought it would.
 

Offline edy

  • Super Contributor
  • ***
  • Posts: 2385
  • Country: ca
    • DevHackMod Channel
Re: Simple speaker hack produces poor sound quality
« Reply #5 on: September 11, 2016, 05:14:56 am »
Could the audio going to the original small speaker be really amplified? So by the time you drive it into an external amp it is too loud and distorted? I would check the levels on the output. Otherwise you may have to see if you can tap the audio signal somewhere else on the board, if there is a separate amplifier chip then before it enters it. Otherwise if they are using a SOC that processes and amplifies the audio you won't be able to get it before amplification. I'm not sure how else you can maybe dampen the volume on the output, or have it go to un-powered external speakers. Also check if it improves if you lower the volume, if you have the option on the projector. If you have an oscilloscope handy you can see if you are getting clipped levels.
« Last Edit: September 11, 2016, 05:16:53 am by edy »
YouTube: www.devhackmod.com LBRY: https://lbry.tv/@winegaming:b Bandcamp Music Link
"Ye cannae change the laws of physics, captain" - Scotty
 

Offline EDPRTopic starter

  • Contributor
  • Posts: 11
Re: Simple speaker hack produces poor sound quality
« Reply #6 on: September 11, 2016, 05:31:49 am »
picture of where the speaker output leaves the pcb
 

Offline TheBay

  • Supporter
  • ****
  • Posts: 1016
  • Country: wales
Re: Simple speaker hack produces poor sound quality
« Reply #7 on: September 11, 2016, 05:50:46 am »
Have you just tried using the red and black wire, also it might need some form of attenuation as line in will be Millivolts.

The white looks like a chassis ground, unless that speaker has a tweeter on the front of it? Cant see anything from the picture to show its a dual voice coil. Does a copper wire connect the other side of the middle terminal?
 

Offline edy

  • Super Contributor
  • ***
  • Posts: 2385
  • Country: ca
    • DevHackMod Channel
Re: Simple speaker hack produces poor sound quality
« Reply #8 on: September 11, 2016, 05:55:17 am »
Interesting... the speaker has 3 wires, the white one seems to go to the "center" while the black and red go to the sides. How did you hook up these wires to your external audio jack? I haven't seen a speaker like this before, why are there 3 leads to it?
YouTube: www.devhackmod.com LBRY: https://lbry.tv/@winegaming:b Bandcamp Music Link
"Ye cannae change the laws of physics, captain" - Scotty
 

Offline EDPRTopic starter

  • Contributor
  • Posts: 11
Re: Simple speaker hack produces poor sound quality
« Reply #9 on: September 11, 2016, 05:58:28 am »
Have you just tried using the red and black wire, also it might need some form of attenuation as line in will be Millivolts.

The white looks like a chassis ground, unless that speaker has a tweeter on the front of it? Cant see anything from the picture to show its a dual voice coil. Does a copper wire connect the other side of the middle terminal?
No the white one it's just connected to the speaker chassis, no tweeter.  The speaker seems to work without the white connected also.
Perhaps interesting to mention: there is quite the difference in sound power if I connect the speaker with BLACK (to -) and RED (to +), or BLACK (or RED) (to -) and WHITE (to +).  Short circuiting one cable with any other stops sound all together, needing to reboot the beamer to get the sound back.



 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 21606
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: Simple speaker hack produces poor sound quality
« Reply #10 on: September 11, 2016, 06:07:20 am »
Yup, you're double-screwed...  :scared:

1. See the traces coming off the speaker connector?  Two ferrite beads, then a teensy black square?  Looks to be a 9 ball CSP.  Class D amplifier, unfiltered.  So, there's full amplitude switching noise on the output, which is balanced (bridged, as they say in amplifier lingo).

2. Amplitude is whatever's needed to drive that speaker.  Which... 5W, 4 ohms, so it only needs about 7V supply.  So expect signal levels on the order of 5V, not ~1V that an external speaker wants!

Solution?  Well, it's possible, but it's not pretty...

a. Keep a load resistor between the terminals, of maybe 4-10 ohms.  Just because.  The amp might have poor linearity into light loads, or an open-circuit protect or something, I don't know.
b. Attach a filter.  If the switching frequency is quite high, then an RC filter might be good enough.  Otherwise, you'll need an LC filter, and a matched impedance load (usually a series R to dampen the LC, since your attached amplifier probably won't have a useful input impedance).  Typical values might be 1k\$\Omega\$ + 8.2nF, or (120\$\Omega\$ + 1mH) + 68nF.  Ground is evidently white, and red and black are signal. Doesn't matter which one.  Leave the other alone (aside from the load resistor if necessary).
c. Add an attenuator.  For the RC filter, you can simply put a voltage divider after it.  For the LC filter, use resistors somewhat larger than the series resistor (i.e., for 1/10th voltage, use maybe 820 ohms in series, 120 ohms parallel to ground).
d. You only get mono, unless there's another one on the other side.  In which case, repeat the process.

How's that HDMI splitter looking now?...  :-DD

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Online Brumby

  • Supporter
  • ****
  • Posts: 12287
  • Country: au
Re: Simple speaker hack produces poor sound quality
« Reply #11 on: September 11, 2016, 06:13:40 am »
Yup, you're double-screwed...  :scared:

1. See the traces coming off the speaker connector?  Two ferrite beads, then a teensy black square?  Looks to be a 9 ball CSP.  Class D amplifier, unfiltered.  So, there's full amplitude switching noise on the output, which is balanced (bridged, as they say in amplifier lingo).

2. Amplitude is whatever's needed to drive that speaker.  Which... 5W, 4 ohms, so it only needs about 7V supply.  So expect signal levels on the order of 5V, not ~1V that an external speaker wants!

I was looking at the PCB too - and that makes more sense to me than anything else.

I agree with the approach on the solution.  Put a load on it, filter the audio and pray there aren't any earth reference issues.
 

Offline T3sl4co1l

  • Super Contributor
  • ***
  • Posts: 21606
  • Country: us
  • Expert, Analog Electronics, PCB Layout, EMC
    • Seven Transistor Labs
Re: Simple speaker hack produces poor sound quality
« Reply #12 on: September 11, 2016, 06:13:47 am »
Interesting... the speaker has 3 wires, the white one seems to go to the "center" while the black and red go to the sides. How did you hook up these wires to your external audio jack? I haven't seen a speaker like this before, why are there 3 leads to it?

EMC.  The speaker is driven with full square waves at some high frequency.  It's grounded to minimize radiation, and fed with balanced power.

Tim
Seven Transistor Labs, LLC
Electronic design, from concept to prototype.
Bringing a project to life?  Send me a message!
 

Online Brumby

  • Supporter
  • ****
  • Posts: 12287
  • Country: au
Re: Simple speaker hack produces poor sound quality
« Reply #13 on: September 11, 2016, 06:16:57 am »
From the symmetry on the PCB, it would seem the balanced power is on the red and white - so it seems curious to me that the white is the one connected to the speaker frame.....
 

Offline EDPRTopic starter

  • Contributor
  • Posts: 11
Re: Simple speaker hack produces poor sound quality
« Reply #14 on: September 11, 2016, 06:18:02 am »
From the symmetry on the PCB, it would seem the balanced power is on the red and white - so it seems curious to me that the white is the one connected to the speaker frame.....
FWIW it *is* working without the white cable connected also.
 

Online Brumby

  • Supporter
  • ****
  • Posts: 12287
  • Country: au
Re: Simple speaker hack produces poor sound quality
« Reply #15 on: September 11, 2016, 06:20:42 am »
From the symmetry on the PCB, it would seem the balanced power is on the red and white - so it seems curious to me that the white is the one connected to the speaker frame.....
FWIW it *is* working without the white cable connected also.

I would expect that.  Red is one power line and black is the 0V, so it will work.

Could the White be an antiphase signal to attenuate RFI?  Dumb idea.
 

Online SeanB

  • Super Contributor
  • ***
  • Posts: 16271
  • Country: za
Re: Simple speaker hack produces poor sound quality
« Reply #16 on: September 11, 2016, 07:26:40 am »
Red is the one drive, and black is the other (inverted) drive, so the white is to reduce radiated EMI from the speaker. Simply use a load resistor ( 22R 1W carbon film) to get it working properly, and then use a RC filter from one of the leads ( as Tim said, 1k and 8n2 will likely work) to ground, and then an attenuator to get the voltage low enough to feed the external amplifier.

Likely the class D amplifier runs at well over 100kHz, likely 1MHz, so the simple Rc filter will do for this.
 

Offline EDPRTopic starter

  • Contributor
  • Posts: 11
Re: Simple speaker hack produces poor sound quality
« Reply #17 on: September 30, 2016, 04:28:06 am »
I'd really like to proceed with building these converters but lack the skills for understanding the components and connections between them.  I'm not afraid to get my hands dirty and I do know how to operate a soldering iron.

Could somebody help me with a component list and simple drawing so I can make a test on a breadboard?

I understand this might be a a biggy and a lot of work.  1000x Thanks in advance.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf