TDA2050 Mixer-Amp -- Help?

picbuck

Member
2008-05-31 12:50 pm
Would anybody care to take a look at this mixer-amp schematic? Any comments about what's going to blow up will be greatly appreciated.

PS: I'm looking at the other TDA2050 postings, thanks to those who posted.
 

Attachments

  • TDA2050_d_3b_iview.jpg
    TDA2050_d_3b_iview.jpg
    112.3 KB · Views: 294

picbuck

Member
2008-05-31 12:50 pm
As simple as that? Well then...thanks JMFahey!

ERRATA: just for the record, R5 and R6 on the schematic are shown as 2K, but the correct value for each is 10K.

The wrong values were left in when I was trying--and failing--to convert the whole thing to use NE5532s. Actually 2K would work fine (as far as I know), but the feedback network resistor values everywhere else are 10K, so what the heck. And thanks again.
 
Well, if you wish a "correction" (although it can be skipped safely), add a 10uF cap (pointing either way is the same) between R5 and R8 , and try 3 values for C3: original 1 nF, or 2n2F or 4n7F .
I think 1 nF will work "too high" and sound like it's not doing much.
Increasing its value will make it more aggressive, includig some mid highs.
I guess you want to use it for some kind of live sound, and there it's useful (bad rooms, poor speakers or mics, etc.)

But, as is, there's nothing *wrong* with your design.
 

picbuck

Member
2008-05-31 12:50 pm
Thanks for your comments, JMFahey. I need all the correction I can get.

I lifted the Baxandall network pretty much in one piece from the well known Rod Elliot, at Elliot Sound Products. But on his site the otherwise-identical schematic appears in different places with different values for what I call C3. I'll be breadboarding things section by section and I'll certainly use your suggestions.

Regarding a 10uF capacitor on the input side of R5, I kinda wondered about that.

As much as an experiment as anything, part of my goal here is to keep phase shift to a minimum. As in few-to-no capacitors in the signal path. Offset voltage be hanged, basically I just wonder if it will matter.

LTspice (which I figured out how to use in a very basic way) tells me that phase shift here is less than two-degrees. That's until the TDA2050 and C5, which throws everything into a bucket, but that's a different story (think headphones).

But then I started to wonder. The op amps themselves have internal capacitors--unlike the ideal op amps in LTspice. So I've started to say to myself: what diff? So back to the breadboard and again I'll use your suggestion.

More broadly, I kinda-sorta get the idea that you question whether tone controls are necessary? I'm aware that the present-day custom is to leave them off.

Not a custom I can agree with. Whenever I'm on my soap box I shout, "Amplifiers do not produce sound, speakers produce sound." Even the best of speakers have something of a hit-or-miss factor built in, and room acoustics can change everything. So I feel at least some ability to compensate for such variables is needed. I don't say I'm right, I just say that's my view.

Even more broadly, this current project is for my gf, to replace the utterly horrid computer speakers (redundant) she uses now. But it's also a warmup project.

My brother is a sometime-busker (street performer) in Key West, where he lives. If you see him wave, he's the guy with a buzzard on his shoulder.

What he obviously needs (still in my view) is to be able to run equipment off the plus-or-minus 24 (or more) volts available from series-connected car batteries. This I plan to implement with amp(s) running the mighty LM3886, the even mightier TDA7294, or similar. He plays both banjo (all treble) and 12-string (large instrument with strong bass, don't forget the sub-harmonics), and performing outdoors...he might need a 10-band equalizer.

None of which I actually know anything about. I use to know a bit about tubes (which are not valves unless you live in England, for crying out loud), but these teensy little black chips are a different world. So all the correction I can get, and again thanks very much for your comments.