I'm looking for help making a small powered speaker

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
I'm looking for some direction/advice.

My poor old computer needs some speakers as the old ones suffered some sort of magic smoke releasing failure and I've got a ton of old scrap electronics kicking around as I really hate to throw away anything with a pcb.

So, I figured this would be a great opportunity for me to get some education about audio equipment and the construction thereof.

I found this ridiculously simple thing on the interwebs 8 Watt Audio Amp

I've got an old atx psu and few crappy misc speakers from the smoldering remains of my old computer speakers and a dead television. I figure I can connect the speakers in a parallel/series circuit to get around 6 watts or so at 8 ohms.

The first question I have is can I substitute the lm383 (which I'd have to buy online) called for in the project with a lm324n I found in an old emergency light?
 
No. The LM324 is a low-power quad op-amp, used in low level amplification. The LM383 was designed to drive speaker loads. It is also an obsolete part. I suggest you find a current amplifier IC and follow the manufacturer's design recomendations to avoid disapointment and endless frustration. E
 

docd

Member
2013-01-25 8:27 pm
Labrador
Alright, thanks for the heads up. How about the ECG1232 and TDA2002 that were mentioned in the guide as acceptable substitutes? I couldn't find those on digikey either.

I found a post mentioning a the LM1875 and LM675T, however. They look similar, would one of those work in that circuit?



Concept art of the powered speaker in question:
cod3.jpg
 
Regardless of which chip is used inside
I'd follow a form for function idea.
The way most Pc speakers are built doesn't fit with usual hi-fi gear.
The volume knob ( with the relative stereo potentiometer) could be
omitted, tough useful for fast and handy action, as it can be controlled with the Pc .
The other thing is that one box is master and the other is slave, so one amplifier would see a 10 cm long wire and the other about 1,5 m .
And , when power rises >10W, the heatsinks associated with the power devices
need to be put to direct air, as also the power transformers need
thermic 'compensation' :confused:
So there's the power cord.
:mad:
That's why separated pieces of equipment guarantee some primary and safe norm for connecting them.
 

docd

Member
2013-01-25 8:27 pm
Labrador
Where are the speakers ? In the branches ? :p

why not...hanging from the ceiling !?

lol yeah, I'm not much of an artist. They're hiding behind the gray mush I wanted to look like mesh.

LM324s are best used as multi-pin thumb tacks. There is not one redeeming quality of those chips, except as thumb tacks. Of course I could be wrong.


How about the TDA2030, would it work in the aforementioned circuit?


edit:

Regardless of which chip is used inside
I'd follow a form for function idea.
The way most Pc speakers are built doesn't fit with usual hi-fi gear.
The volume knob ( with the relative stereo potentiometer) could be
omitted, tough useful for fast and handy action, as it can be controlled with the Pc .
The other thing is that one box is master and the other is slave, so one amplifier would see a 10 cm long wire and the other about 1,5 m .
And , when power rises >10W, the heatsinks associated with the power devices
need to be put to direct air, as also the power transformers need
thermic 'compensation' :confused:
So there's the power cord.
:mad:
That's why separated pieces of equipment guarantee some primary and safe norm for connecting them.

Some of this makes sense to me, I'll stick a little red led where I wanted to put the volume pot to give him an eye. I'm certain I can scrounge a few aluminum heatsinks, some old computer fans, and some thermal grease to cool whatever chip I end up. I planned to house all the speakers in the one fish shaped wooden box (yeah that part is probably not happening lol).
 
Last edited:
Although tda2030 and lm1875 have same pinout, and are better amps, they won't work with that circuit, The 383 and the tda2003 are designed to have output at half supply voltage with no external components, while the other chips would require external circuitry to do that. Also, IMO the 1875 doesn't work well on just 12v. My 2 cents.
 
Fully agree, if you have 12V, use a TDA2003.
But follow the schematic suggested in the TDA2003 datasheet, which adds a couple extra parts (Zobel network for stability) and suggest a PCB layout.
Even if you build it on perfboard, you can roughly use it as a guide.
You'll need to bolt it to a heatsink, which may be as simple as a piece of flat aluminum .
NOTE: THE PIN NUMBERS SHOWN IN YOUR "8W" SCHEMATIC ARE WRONG.
FOLLOW WHAT THE TDA2003 DATASHEET SUGGESTS.
http://www.st.com/internet/com/TECHNICAL_RESOURCES/TECHNICAL_LITERATURE/DATASHEET/CD00000123.pdf
The components labelled Rx and Cx are not needed.
 
Yes, TDA2003"A" is probably an "improved" or whatever version.
Must be fully equivalent or they would have used a different number.
Use the schematic in Figure 16 in the ST datasheet.
As you see, Cx and Rx are greyed out because they are optional and really nobody uses them.
In Fig 17 you have a PCB or layout suggestion.
I'm using something very similar for background music with my PC.
100000000000 times better sound than the typical "200W (PMPO) PC Speaker" with the 2" cheesy speakers.
 
There is not one redeeming quality of those chips,
Maybe not as audio preamps, but they are a Godsend in Industrial Electronics.
I also use them a lot (really LM358, which is the Dual version) in my amps in various sensing circuits; thermal muting, thermal tracking variable speed fans, controlling relays to avoid turn on thumps and protecting speakers from DC, etc.
And they are excellent as DC servos and autobiasing circuits for MosFets.
EDIT: just in case you think "all that can be done by any regular Op Amp", the answer is no: they are happy with the 5v *single rail* supply available for the Logic chips, input voltage can reach 0V (in a single supply situation) , in fact even 0.6V *below* 0 V , are low power by design, and output can reach *almost^* 0V under the same constraints.
Very useful little chaps.
Did I mention that $1 buys 5 of them?
 
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.