last update : 21-02-1998
After the D/A converter you can find components for filtering, (to reduce ultrasonic garbage), buffering (to obtain low enough (typical 100-2k) output impedance) and muting. In almost any budget CD_player (below fl.1500 (~$900)) muting is performed with bipolar transistors. THIS REALLY SUCKS !!! More expensive players use relays for this job. Manufacturers apply muting to prevent "plopping sounds" while switching on/off. The audio circuit is temporarily short-circuited during on/off switching. When just playing music they do nothing at all! Nothing ?............., NO WAY, it alters the perceived sound. Why ? I remembered this explanation : a transistor has a small paricitic capacitance. The value is depending of the voltage from collector to emitter, i.e. your audio signal. The result : a signal dependent high-cut filter in your CD player. (ironic) Thank you so much : Marantz, Philips, Technics, Denon, Harman Kardon, Pioneer, Sony, Teac etc etc. Have they ever actually listened to their budget stuff ?! I can't imagine. Maybe they do apply muting transistors to give their more expansive one's a bigger lead and better sales. This would really suck !! (in my humble opinion)
I am not satisfied with this explanation : If you actually calculate the parasitic capacitance it is too small to hvae effect in the audio band. I hope to find a better more detalied explanation soon !
Check Audio en techniek for the source of this tweak.
The more expensive models use relays, so the manufactures are aware, but fail to pass the benefits to you.
I own a Marantz CD 52 MK2. After removing the muting transistors I noticed (warning : vague audiophile terms appear here) , a more fluently sound, less stress and more clarity, especially in female voices. Yes there are some plops to detect while switching on/off. This doesn't hurt my amp or speakers. (Unless : very high volume setting.). I am very pleased with this tweak. It is the best I ever encountered. It kept me from spending (too much) cash on audio en have left me more for music.
There has been some discusion on this bit in some newsgroups.
Goto : Deja News
and search with the keywords : muting transistors
or the keywords : CD Improvement Worked
the : view thread is very usefull
Simply remove the 2 mute transistors (at each channel). In my CD-player these are SMD's. You find them after removing the bottom plate : the very small blocks with 3 leads close to the output socket's. This modification also turned out very nice on my Technics ST-G550 FM/AM tuner. It really started to sing after removing the single mute-transistors : ! The dull, uninvolving, undetailed sound was gone. Switching channels (without muting) makes some tuning noises (just like with the "dial kind" of FM-tuners). I don't care, but it might be an inconvenience for some.
Have fun, get this little devils out, sit back and enjoy the music
There are several more options to get more of cheap CD-players. I just love to try them all, but I just short on time. I swapped the opamps for a pair of OP2604 's. And give these an own power suply. The result : A very crisp sound. Some music is hard to listen to, it is unforgiving with bad recordings. So there is improvement in well record stuff, but some CD's are not fun to listen. I am a bit ambiguous about this mod, and can not fully recommend it.
Why would somebody remove ceramic capacitors? This capacitors are excellent for decoupling of High Frequencies circuits, but they have nothing to do in or at the signal path. The value (of C) is voltage dependent. Therefore phase- and time shift problems can occur. Many amps have these cc parallel at he input to prevent HF entering the amp. This aim is oke , but replacing with styroflex capacitors is a sensible and musical thing to do !
Styroflex is also usefull (and available in 1% accuracy)
Sometimes a high value resistor is used after the last opamp, in my Marantz and most Philips you find 2 * 100ohm and this isn't a problem at all. Many R's are in the range of 400-1500ohms. This can give problems (loss of high frequencies) if you use long interconnects. Removing is not recommended, lowering to 200 ohms or so is oke. (secure for accidental shortcuiting the output)
As far as I'am concerned : using better opmaps a silly thing to do onless you take care of a decent power suply. A ne5532/4 is capable of musical results. The power suply rejection rate of an opamp is exellent. But several millivolts of HF on power suply lines is in my point of view a bad design. The high feedback of the opamps will reject it to level below the noise floor. But i feel real sorry for all these litle transistors slewing and being saturated a million time a second, so I prefer to prefent this.
I upgradeded all ceramic capacitors in the above schematic with styroflex 630V 1%.
I could not get 39pF so I tried 47pF instead. The result : It sounded horrible , lots of distortion especialy in the mid band. After a week of dispair (styroflex sucks or what ..?..) I put back new ceramic 39pF 's. Lucky the sound was ok again. (I left all the other 100pF 's styroflexes in place)
Now I wondered : if these minor changes in capacitors are so important, what should I do now. Why are these caps here anyway. Filtering I suppose (integrating with 39pF doesn't make sense. But these 39pF 's are directly connected tot the DA conertor, their is no R or L in front ! OK just the output impedance. Most (op) amps don't like C's at their ouput, so I decided simply to remove the 39pF's.
And guess what ....
I had to grip my chair to prevent falling ofF : a huge improvement, anybody can here this change.
A very clear open sound hit me. WHOW !!
Hmm you think : " this sounds good and very easy to me, I can do that". BUT I have to warn you : Reducing filtering can hurt if you have still the original op-amps in place. Oscillation can occur !! To be a bit more save : use a smaller (3.9pF????) value.
So after the mod's we have the next :
THe question mark indicates an LC filter circuit. The filterfrequency is is weird.
It is about 340 KHz. I suspect to find a ca. 11.3 MHz frequency.
With the values L=0.22uH an C=1nF you get 10.7 MHz !!! Is this a) a misprint or b) a design error ?
mail me !
return to home (page)
back to audio page click here !
THe question mark indicates an LC filter circuit. The filterfrequency is is weird. It is about 340 KHz. I suspect to find a ca. 11.3 MHz frequency.
With the values L=0.22uH an C=1nF you get 10.7 MHz !!! Is this a) a misprint or b) a design error ? Who knows?
mail me !
return to home (page)