After hearing fellow DIYers in this forum ranting about PCM56 (especially Bernhard), I decided to give it a try. I'm convinced that the channel switching using 74HCU04 is more than adequate, anything more that that requires a jig to hold my head in the chair.
SPDIF processing is done using twistedpearaudio's WM8805 kit, which basically takes care of jitter issues (to my standards). Looking at datasheet, the PCM56 has relatively poor power supply rejection so independant regulators at every input is a MUST. Since 8 regulators are required for for stereo, my laziness calls for the 7xx5 series.
Output current is converted using 240R resistor, and amplified with OPA627 (with RFI well taken care).
The midrange, trebles details is absolutely stunning. Can say something like TDA1541 after the DEM reclocking & stuffs..
Running without the MSB trim, it does the Sheffield disc track 40 quite well, something that even WM8740 and PCM1798 struggles with.
BUT... Would like to gather opinion from the brave souls who has tried this DAC (esp Bernhard):
Bass seems missing. Or is it that the neutrality of the DAC makes it seems missing?
Im using +/-5v. Any improvements going for higher supply voltage?
Is opamp IV a better match to this chip? Or any output stage that compliments well?
Cheers,
Colin
SPDIF processing is done using twistedpearaudio's WM8805 kit, which basically takes care of jitter issues (to my standards). Looking at datasheet, the PCM56 has relatively poor power supply rejection so independant regulators at every input is a MUST. Since 8 regulators are required for for stereo, my laziness calls for the 7xx5 series.
Output current is converted using 240R resistor, and amplified with OPA627 (with RFI well taken care).
The midrange, trebles details is absolutely stunning. Can say something like TDA1541 after the DEM reclocking & stuffs..
Running without the MSB trim, it does the Sheffield disc track 40 quite well, something that even WM8740 and PCM1798 struggles with.
BUT... Would like to gather opinion from the brave souls who has tried this DAC (esp Bernhard):
Bass seems missing. Or is it that the neutrality of the DAC makes it seems missing?
Im using +/-5v. Any improvements going for higher supply voltage?
Is opamp IV a better match to this chip? Or any output stage that compliments well?
Cheers,
Colin
You could say bye bye to the opamp and use 1k for I/V instead.
I have never had the impression of missing bass.
With a scope or true rms multimeter and a test CD you could check frequency response.
Also for nonos I suggest filtering.
http://www.diyaudio.com/forums/digital-line-level/165707-cs8416-pcm56p-4.html#post2430026
This can not be used with other than 200 ohms Ri/v.
Been tweaking with the chip for some time now, logically speaking the bass cannot go missing. Its possible that the bass is just clean.... super duper clinically surgically clean... Takes some time to get used to.
Bernhard, I went through the passive IV but the limitation is that the DAC has to drive something down the chain.. It is too much to ask for the DAC to drive passive EQ into signal cable into a preamp volume pot... I prefer to mount a beefy driver right in my source to drive a passive preamp.
Current implementation is using OPA861 current buffer-> passive IV-> OPA627 buffer. Imaging and tone neutrality is excellent. This is truly an under-rated DAC chip.
Bernhard, I went through the passive IV but the limitation is that the DAC has to drive something down the chain.. It is too much to ask for the DAC to drive passive EQ into signal cable into a preamp volume pot... I prefer to mount a beefy driver right in my source to drive a passive preamp.
Current implementation is using OPA861 current buffer-> passive IV-> OPA627 buffer. Imaging and tone neutrality is excellent. This is truly an under-rated DAC chip.
Hi Erin,
I think the bass is there but just too clean... refer to my previous post, I might replace the OPA627 buffer with Jfet discrete buffer. This might introduce some 2nd harmonics back into the music..
I totally agree that it is super clean. But..... my TDA1541 goes deeper and sounds better to my ears. But they are different PCB's etc. totally different boards and output stage. Its not a fair comparison. I will work on my PCM58 DAC when I get some time. I know it has a lot of potential. I think the output stage on mine is letting it down.
I doubt thats the sole issue (lack of 2H.) I like that you are using a passive I/V + active for gain and buffer. You may need a lower I/V resistor. Have you tried a phono preamp-type opamp stage ? But I really think your issue may be powersupply/capacitor choice.
Sometimes you can help the all the 7805's by preceding them with 2 big transformers feeding two sets of big LM317/LM337 regulators stages set to 12V's, then driving the 7xx5's ( so you have a digital +-12V and an analog +-12V.) Keeping the digital supply separate from the analog all the way to the transformer.
Using higher IV resistor tends to make the subsequent amplifier stage easier to design and build, can achieve good sound with relatively little effort.
But with low IV resistor,it demands tremendous task for the subsequent gain stage. Requiring gain of more than 150... any slight layout mistake can render the sound to be totally UN-listenable. My current implementation uses cascode, with shunt regulated supplies.
But with low IV resistor,it demands tremendous task for the subsequent gain stage. Requiring gain of more than 150... any slight layout mistake can render the sound to be totally UN-listenable. My current implementation uses cascode, with shunt regulated supplies.
Even with the troublesome PCM1704 I've found good results with a 30 ohm IV and + jfet gain (w/nfb) +diamond buffer, 2H right at -96db (there are some tricks to the decoupling caps.)
Basically just a passive I/v with a good phono stage, it works just fine as it did will LPs for years, the trick is finding how much resistance the I-out can handle and not using any feedback to the I-out pin. NFB is fine after the passive I/V just not before. NFB is bad on the I-out pin even when it is in the "current form." I am going to try another NFB free current steering I/V next week but my experience is they just can't keep up with an R2R's I-out sound dry and most just don't measure well. All jmho take it with a grain of salt.
In reading this thread, I was going to make that suggestion but I then saw that Regal had already done so. So, while I have not experimented with the PCM56, I have experimented with the AD1865 using this same passive I/V circuit. What I found with AD1865 is that lower I/V resistor values generally improved the sound, and in particular, brought better subjective focus to the bass.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.