CS4397 dac

[AndersN]

To big pdf.

[AndersN]

At last

[Ekke]

Too low quality.

[AndersN]

New pdf.

[yiannis]

Hi everybody. I have a CS4397 dac with the following mods: All electrolytics replaced with oscons at the digital section. Bypass the analog stage by removing C27,28,30,31 and connecting two scr mkp caps directly from the cs4397 (AOUTR-, AOUTL-) to the output RCAs. It was working great for weeks, but suddenly the following problem appeared: I get strong noise (like an impulse signal) when changing to the next track of a cd. During a track's playback everything is ok. Changing to the next track gives a strong noise, and the woofer gets displaced out to its limits, like dc is passing to the speakers for a very short period of time, which does not make much sense since i have dc blocking capacitors at the outputs.

While this noise is heard, i measure the chip's output dc falling from 2.5V to approx 2.0V dc and then it gets back to normal 2.5V. This is when measuring directly the output pins (AOUTR- for example). When i measure AFTER the dc blocking capacitors that i have added, the voltage is 0V dc (as expected) but when the noise happens the analog voltmeter indicates a negative voltage, like a negative impulse signal. Check the following video to see what i mean. I measure the voltage on my rca outputs (= after the dc blocking capacitor) and the needle is on 0V. I press the "next track" button and the needle moves to negative voltage for a while.

http://www.youtube.com/watch?v=Be4j2biZPgU

I measure the voltage on C3 and it remains stable when the noise happens. I measure on C25 and the voltage drops when i change track. If i change tracks many times continuously (pressing the next track button 2-3 times per second) the voltage drops more each time i change the track (like something sucks much current instantly) and when i stop changing tracks the voltage gets back to normal. I increased the C25 to 150uF and the voltage drop is quite smaller but the noise is still quite strong and the woofers pop out to their Xmax

I re soldered most of the connections (input rca, power supplies etc) with no improvement. The same happens when i connect the dac to my computer and play a sound file. At the beginning and the end of the file the "tsak" noise happens and the woofers stick out. It seems like when the spdif is initiated or stopped the noise is heard.


Any ideas what could be wrong?

[yiannis]

and this is when measuring directly the chip's output: The voltage is stable at 2.5V (needle stays at the right corner, maximum of the scale). I press the play button on the cd and 3 impulse noises happen, and the voltage drops down to 2V for three times, and then the track starts playing.

http://www.youtube.com/watch?v=5FeucajR6is

[fandersen]

Hi all. Lots of good info on modifying this dac!

I was wondering if there are any recommendations for a replacement power supply for this dac, preferably diy?

[janos_904]

Hi DIY-ers, happy to post, that the sound of my mini DAC is improved further (see #417 ). The foil GE-s were replaced to 0.33 uF Mundorf Silver/Oil. It makes the total project pricier, but the song of mini’s turns to another league! I have very good results using Russian military silver mica KSZG caps right now: 100 nF/500 V. I found them very close (or better) to Mundorfs at lesser price! There is question showed up about the values of C25, C16. I replaced Yageo at C25 to Blackgates (2x10 uF) and left C16 (OsCon 150 uF). There was a nice, clear sound, and! deeper bass. But, during long listening I found the sound tiresome (if it is the correct word for). Lampizator’s advice was chosen; I put 100 uF OsCon at C25. I know, the value is much higher than the factory recommendation. The sound is very relaxing, nice, there is a bit less bass but easy to listen for 2-3 hours. It is not slowed down, as was mentioned earlier. Did you find similar results?

[BlackKnight]

Hi from a newcomer.
I hope that someone will give me a helping hand. After soldering a new C8 capacitor I noticed that there is no voltage on the soldering pads. The voltage should be the same as on pin 26 on CS4397 chip, right ?

[janos_904]

Hi, I am experienced using Mundorf Silver/Oil and KSG capacitors further. I think, despite of high respect of Mundorf S/O, they colors the sound a lot, KCG-s do as well, maybe not that much. There is a very important capacitor, namely C9 (DC blocking capacitor) just in SERIES in the signal path. I got some Russian military SSG-3 (CCГ-3) 126 nF/350 V Silver Mica’s from Vilnius. Because of their size, I made a new rout to receiver chip via a BNC terminal installed (do not forget there is a 75 ohm between signal to ground just before silver mica). I put 0,252 nF CCГ-3 silver Mica’s (C9). The result is surprisingly good (stunning). I highly recommend trying this tweak. János (to get CCГ-3, search on US e-bay)