Creative Sound Blaster 16s with incorrectly capped TEA2025 amplifiers

[Logistics]

Re: Creative Sound Blaster 16s with incorrectly capped TEA2025 amplifiers

For whatever reason, I never noticed this thread, and I even made a thread of my own, dedicated to refreshing the SB16. Obviously, I took things a step farther, in some ways, with a refreshed motherboard and PSU to stiffen the power delivery and increaae stability of the system in general.

But I wanted to address some minor things, which were commented on in the thread:

One user commented that you must use passive speakers with the SB16's output. However, as with most amplifiers, this is not entirely true. You may also use headphones, but you do need to be wary of your volume-levels as the output-power is quite high on the Speaker-Out. As far as impedance is concerned, the data sheet only scales to 16-Ohm IIRC, but I have used 24-Ohm headphones on my refreshed SB16, and I can barely turn the volume up within Windows when running directly off the card. (Maybe the low-impedance caps)

There seems to be confusion as to why one would use bipolar capacitors in the signal-path--this is because audio signals exhibit A/C-like characteristics which are better reflected through bipolars.

The way I understand it, larger you in value you go in capacitor, the farther down you generally bring the ceiling as concerns frequency handling. On inputs, often a careful combination of capacitors are used, one bypassing another--the combination often consisting of 1.0uF and 0.1uF, and sometimes the addition of an even smaller capacitor. But placement must be sequential or you will defeat the purpose of one or more.

Matthew

[gustinmi]

Re: Creative Sound Blaster 16s with incorrectly capped TEA2025 amplifiers

First of all, the datasheet for tea2025b stated, that NP caps should be used only on input (allowing source to be referenced to ground). So there you need 0.22uF NP cap.

I noticed that the cap that changes everyting is the small serial capacitors that goes from ground into PIN 2 (OUT-1), 0.15 uF.

If I used ceramic or polyester capacitor, the chip started to draw 4 AMPS of current (max should be 1.5), and the distortion was huge.

It only worked when I used small ceramic cap (104) or polarized cap. The last one worked the best, and chip consumption was 0.9 AMPS.

Go figure!

BTW, according to Make: Electronics book, you can make your own bipolar caps with joining 2 electrolyte together. They should be double in capacitance. Minus poles are joined. It works well for input capacitor I mentioned in beginning.

[Logistics]

Re: Creative Sound Blaster 16s with incorrectly capped TEA2025 amplifiers

Just real quick, because I'm out eating pho, I think you interpreted what the data sheet wrong. It doesn't say that NP caps should be used only on the inputs, it says that the capacitors used on the inputs should only be NP. This says nothing of the caps elsewhere on the board.

Edit: I'm interested in your observations of the effects of that 0.15uF cap. Why does changing the capacitor type yield such an effect? And since the data sheet cautions against using a smaller value, I wonder what effect a larger value has besides stability.