Using the AD844 as an I/V

Are these two from Denon´s professional line? Not sure what that really is? They were the top of the line back some time ago. There is known problems with the transport/lasers. I bought them thinking I would use them as dacs. I was luck that the both ended up being repairable. The 5000 is a two channel audio dvd player, the 9000 is 5.1 with all channels using PCM1704s. Nicely built with a few peculiarity, like the 5000 has no decoupling caps on the digital side of the dacs. Designed to be there but left off when built(?).
 
AD844 triple stack + BUF03

Hi George, Finally wired up the BUF03's off the TZ point of my triple stack (SOIC) I/V's. While I was waiting I switched to Cinemag transformers for the filter function instead of the Edcors. With the triple stack and running the bottom 844's output stage and the Cinemags, I was really quite happy. However I am listening to the BUF03's output and I must say this isn't a subtle change in my system. I am listening to a remastered CD set, Pink Floyd "The Wall". Dynamics are outstanding, clarity and detail. Bass is super. I will work on my Pedja discrete circuit next week. I am short a few Wima bypasses. I will report my experience using the BUF03's as a buffer. Dave
 
Hi George, Finally wired up the BUF03's off the TZ point of my triple stack (SOIC) I/V's. While I was waiting I switched to Cinemag transformers for the filter function instead of the Edcors. With the triple stack and running the bottom 844's output stage and the Cinemags, I was really quite happy. However I am listening to the BUF03's output and I must say this isn't a subtle change in my system. I am listening to a remastered CD set, Pink Floyd "The Wall". Dynamics are outstanding, clarity and detail. Bass is super. I will work on my Pedja discrete circuit next week. I am short a few Wima bypasses. I will report my experience using the BUF03's as a buffer. Dave

Yes Dave for such an old (very old buffer) the BUF03 is a stunner, and they are high biased and will burn your fingers even with a heat sink (and before anyone say it. No they're not occilating) +- 16vdc is the highest I would go.

BTW yesterday we finally had the A/B with 2 identical cdp's with PCM1704k's in them between my 2 stack and my friends 4 stack intedical cdp's both using the BUF03 as output buffer. The system to remind all was Wilson Watt Puppy 8's Halcro DM68's (100kohm input impedance) and my Lightspeed Attenuator as the passive pre.

Both were very very good but a touch different. 2 of us had the 2 stack winner by a hair and the other listener could'nt decide.
The 4 stack was a touch less delinated in the bottom end and a bit heavier/weightier, the 2 stack sounded like it went lower.
The 4 stack also did not have quite the jump factor that the 2 stack had in dynamics.
Two of us thought the 4 stack was just a very little grainy on female voice but that could have been the recording, yet we think we didn't hear it on the two stack.

So there you go with the PCM1704 dac with low -+1.2mA output.
1 x AD844 is very good, far better than any negative feedback opamp i/v
2 x AD844's stacked blows away a single big time
3 x stack (shrug of the shoulders)
4 x stack (shrug of the shoulders again)

Cheers George
 
Yes Dave for such an old (very old buffer) the BUF03 is a stunner, and they are high biased and will burn your fingers even with a heat sink (and before anyone say it. No they're not occilating) +- 16vdc is the highest I would go.

BTW yesterday we finally had the A/B with 2 identical cdp's with PCM1704k's in them between my 2 stack and my friends 4 stack intedical cdp's both using the BUF03 as output buffer. The system to remind all was Wilson Watt Puppy 8's Halcro DM68's (100kohm input impedance) and my Lightspeed Attenuator as the passive pre.

Both were very very good but a touch different. 2 of us had the 2 stack winner by a hair and the other listener could'nt decide.
The 4 stack was a touch less delinated in the bottom end and a bit heavier/weightier, the 2 stack sounded like it went lower.
The 4 stack also did not have quite the jump factor that the 2 stack had in dynamics.
Two of us thought the 4 stack was just a very little grainy on female voice but that could have been the recording, yet we think we didn't hear it on the two stack.

So there you go with the PCM1704 dac with low -+1.2mA output.
1 x AD844 is very good, far better than any negative feedback opamp i/v
2 x AD844's stacked blows away a single big time
3 x stack (shrug of the shoulders)
4 x stack (shrug of the shoulders again)

Cheers George

Hi George, For the TDA1541A I like the triple stack. It also gets the I/V input resistance down to about 16.7 Ohms which is close to the "ideal" 12 Ohms. I was also surprised by the holographic sound field. I have several more hours listening time now. I am a happy listener. Thanks for the information on the PCM1704's. I am considering building a dac based around that chip. Dave :D
 
I always wondered why nobody who bought to my place a dsd or sacd front end ever got it to sound as good as my current Redbook setup PCM1704K with the 2 x stacked AD844 I/V stage into the BUF03 buffer.
I guess Daniel Weiss of Weiss dacs with this white paper thinks the same.

Mono and Stereo High-End Audio Magazine: Daniel Weiss White paper on DSD

Cheers George

George
Im not silly enough to ever say anything is "better" , because its impossible to convince anyone that your perceptions are "better" than anothers, however
you might be pleasantly surprised at the sound of a I2S driven Sabre with an OP861 I/V , same concept as yours but an even "better" device than the AD844 , PCM or DSD, but then again you might not.
 
Hi Mick, yes I've got a couple of those 861's here waiting for me to do something with them, even have the second BUF03 for the buffer. I aggree they could be even better than the AD844's it just that it's sounding soooo good at the moment, I'm not rushed.
I even have a second identical 1704 cdp now to do it on, also I have a pair of OPA660, which may even be better.
One day I'll get around to it but for now I'm in heaven listening to to what I have.

Cheers George
 
Sorry, lack of time. Akai cd79 with LC on all power supply pins of dacs, conductive polymer for dac, Muse KZ for AD844, PCM63KY, 3x AD844 stacked, sounds better than Meridian G08.2

The sound is much more analogous and more fluid than nervous Meridian. The character of the sound is similar for both cd players. Next step will be with two AD844 stacked and additional i/v resistor and Sen/Cen. I want to keep the detail and naturalness but i want to get a little more PRaT (when the foot dancing in the armchair)
 
Hi Impuls, Sorry to say this about such an expensive unit but Meridian G08.2, bit stream delta sigma dacs "yuk", had one here for appraisal for a couple of weeks, interesting as bat s**t on redbook made me yawn. Nicely made though and boringly smooth.
The Akai CD79 with PCM63P-K should be better on redbook with just a little tweaking, and far better with the 844 i/v and a good output buffer. bet this makes you jump on dynamic sections compared to the Meridian.

Cheers George
 
I want to keep the detail and naturalness but i want to get a little more PRaT (when the foot dancing in the armchair)

IME stiffer power supplies are a good way to improve rhythm and pace. I just increased the capacitance I have across the PSU pins of my I/V preamp - from 6000uF to 12000uF and noticed more inclination to foot tap :) Seems low impedance down to as low a frequency as possible is important.
 
DSD as a format is technically broken, that's probably why. This is not a matter of opinion, rather hard math.

I have a few DSD recordings I have downloaded and they are fabulous. To me, what is even more important is the quality of the native recording, whether done in PCM or DSD (also could be called PWM or PDM). If they do the recording right, I can live with either formats.

BUT... the real problem is with delta-sigma low-bit DACs and their own attendant noise shaping. Today, most PCM of various bits are processed in delta-sigma DACs - unless you use older DACs and NOS-DACs.

Delta-Sigma DACs, what ever source material they use, need a particular form of post-DAC filtering - and I know of no manufacturer who does it and only a couple of DIY'ers that do.

Read from this post onward: http://www.diyaudio.com/forums/digital-source/222596-oppo-s-bdp105-discussions-upgrading-mods-70.html#post3743000

And: http://www.diyaudio.com/forums/digital-source/222596-oppo-s-bdp105-discussions-upgrading-mods-72.html#post3758596

Coris tries it: http://www.diyaudio.com/forums/digital-source/222596-oppo-s-bdp105-discussions-upgrading-mods-73.html#post3760127

Cheers, Joe
 
Last edited:
BUT... the real problem is with delta-sigma low-bit DACs and their own attendant noise shaping. Today, most PCM of various bits are processed in delta-sigma DACs - unless you use older DACs and NOS-DACs.

Its true that nowadays, the vast majority of PCM recordings are made through an S-D modulator. I rather suspect this fact goes a long way towards accounting for the rather extreme variability of digital recordings in terms of how much dynamics they portray. However it needn't be this way - various vendors have highly capable SAR ADC chips now which boast impressive specs and are reasonably priced. Thinking in particular of both ADI and LT but they're not the only ones, TI and Maxim also make decent-looking parts.
 
Its true that nowadays, the vast majority of PCM recordings are made through an S-D modulator.

I would like you to consider those links (and postings around them and subsequent posts) on the Oppo 105 thread where the 'single order dominant pole' used in the post-DAC circuit, is discussed. It is not just theory, now that too many have sampled (heard) the results and 100% agreement that it works, only that the theory and why it works is open to discussion.

In fact I think you have been there on that thread, pardon my memory.

Cheers, Joe
.
 
Why would you like me to consider them - do please persuade me they're relevant? I have no beef with your findings about fitting a capacitor to an S-D DAC. To me this isn't of interest (from my pov as designer) because I'm for the forseeable future sticking with multibit DACs in my designs.

If anyone comes up with a hypothesis for how it works, I shall certainly be interested in that.