How low can you go? - part II
Posted 31st August 2010 at 03:34 AM by abraxalito
Updated 1st September 2010 at 04:23 AM by abraxalito
Updated 1st September 2010 at 04:23 AM by abraxalito
Here are the schematics - the original output stage and the modified one.
On the first player I bought, I did all those mods in one hit. The third pic shows what they looked like - I took out some of the redundant connectors (guessing the same board is used in players which have more features like component video outputs) to make space for the electrolytics. Having replaced the board (just two screws, one on the rear and another in the base), then I listened. The improvement in bass performance brought about by increasing the value of the coupling caps was noticeable in listening - the cello became more full-bodied - and also on a frequency response plot I made using Audacity. I felt (but can't quite describe) that the soundstaging has improved - music feels more interesting to listen to somehow. I week or so later, I went out and bought a second one (at 99RMB, who's counting?) to remind myself of the original sound. On listening for the first time to a CD I'd just bought of Lang Lang playing Beethoven Piano Concertos I and IV, the unmodded player was trumped for involvement by the modded one. The sound just drew me in more, I noticed more going on in the recording, the orchestral instruments sounded less homogeneous, more distinct and the piano 'filled out' more in space.
Having made a frequency response measurement, I was keen to try dreaming up more measurements which might correlate with the sound quality change. Frequency response was measured simply by recording a chirp (Audacity has the facility to create these with flexibility over the frequency range) close to full output level. I wondered if the frequency response might change with level, so created a CD with chirps at levels descending in 10dB steps, down to -80dB. Recording these descending chirps from the modded player, the waveform in Audacity looked rather disappointing at the lower levels - a considerable increase in noise (mainly low frequency) was visible on the waveform, even at -50dB. A quick comparison with the unmodded player showed this noise increase to be considerably more modest - even the -60dB level looked respectable. Something was clearly amiss with my mods, but I just could not think what it might be.
On the first player I bought, I did all those mods in one hit. The third pic shows what they looked like - I took out some of the redundant connectors (guessing the same board is used in players which have more features like component video outputs) to make space for the electrolytics. Having replaced the board (just two screws, one on the rear and another in the base), then I listened. The improvement in bass performance brought about by increasing the value of the coupling caps was noticeable in listening - the cello became more full-bodied - and also on a frequency response plot I made using Audacity. I felt (but can't quite describe) that the soundstaging has improved - music feels more interesting to listen to somehow. I week or so later, I went out and bought a second one (at 99RMB, who's counting?) to remind myself of the original sound. On listening for the first time to a CD I'd just bought of Lang Lang playing Beethoven Piano Concertos I and IV, the unmodded player was trumped for involvement by the modded one. The sound just drew me in more, I noticed more going on in the recording, the orchestral instruments sounded less homogeneous, more distinct and the piano 'filled out' more in space.
Having made a frequency response measurement, I was keen to try dreaming up more measurements which might correlate with the sound quality change. Frequency response was measured simply by recording a chirp (Audacity has the facility to create these with flexibility over the frequency range) close to full output level. I wondered if the frequency response might change with level, so created a CD with chirps at levels descending in 10dB steps, down to -80dB. Recording these descending chirps from the modded player, the waveform in Audacity looked rather disappointing at the lower levels - a considerable increase in noise (mainly low frequency) was visible on the waveform, even at -50dB. A quick comparison with the unmodded player showed this noise increase to be considerably more modest - even the -60dB level looked respectable. Something was clearly amiss with my mods, but I just could not think what it might be.
Total Comments 0
