PMA said:nice job! But do not hijack my valuable thread
Sorry Boss.
How did you make your PCB?
Toner transfer, my first attempt at two a layer board, it worked quite well. It has three paralled chips for a particularly awkward 4 ohm nominal sub that has an impedance curve that swings all over the place.
For all the chipamp lovers, see what quality you can expect. At 1W/4 ohm, quite standard listening conditions:
http://web.telecom.cz/macura/549.gif
Designers of the circuit did not take care about crossover distortion at all.
http://web.telecom.cz/macura/549.gif
Designers of the circuit did not take care about crossover distortion at all.
I looked at the datasheet for the chip, it has this schematic towards the end wich shows useing a second chip as a slave giving twice the current output, I think they called it parallel output stage, too lazy to check it up right now again...
Would this at least in theory work for other chips like lm3875 etc?
Would this at least in theory work for other chips like lm3875 etc?
LMs have lower distortion. Anyway, for lower powers one also can see crossover distortion residuals.
Regarding the OPA549, my result shown here quite corresponds to the datasheet - they show 0.04% at 100mW, I measured 0.05% at 250mW. The number is not very important. The spectral content is. You can see infinite number of spectral lines - classical crossover distortion, reduced loopgain at higher freq. is unable to make correction. This is quite audible, even if the number says 0.05%.
Regarding the OPA549, my result shown here quite corresponds to the datasheet - they show 0.04% at 100mW, I measured 0.05% at 250mW. The number is not very important. The spectral content is. You can see infinite number of spectral lines - classical crossover distortion, reduced loopgain at higher freq. is unable to make correction. This is quite audible, even if the number says 0.05%.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.