Russ / jgazal,
Thanks for the enlightenment as to the external clock. This brings me to a few other questions as I prepare to build my Ultimate DAC!
1. In dual mono mode do the clocks sync up some how or is just one clocked elected as a master and used for both Buffalo DAC boards?
2. In powering the DAC I am considering a battery based solution, has anyone tired this with success? It seems to me a lot of high end DACs are engineered this way. No dealing with any Nasties that may come off the mains. If I do this could I just bypass the AC rectification on the Placid or Placid BP? If I didn't end up going with a pure DC source like a battery what are peoples opinions on the optimal supply for the Buff II DAC?
Best Regards,
Theo
Thanks for the enlightenment as to the external clock. This brings me to a few other questions as I prepare to build my Ultimate DAC!
1. In dual mono mode do the clocks sync up some how or is just one clocked elected as a master and used for both Buffalo DAC boards?
2. In powering the DAC I am considering a battery based solution, has anyone tired this with success? It seems to me a lot of high end DACs are engineered this way. No dealing with any Nasties that may come off the mains. If I do this could I just bypass the AC rectification on the Placid or Placid BP? If I didn't end up going with a pure DC source like a battery what are peoples opinions on the optimal supply for the Buff II DAC?
Best Regards,
Theo
mmmm and what about digital dynamic range? With digital attenuation you reduce the highest digital figure to a low one. May someone correct me if I am heading towards the wrong direction 
Without digital attenuation, the highest digital number a sample can achieve is, with 16 bits of resolution,
1111111111111111 or 65536 steps.
Reducing volume by half provides you with a peak signal varying from 0 to 32768, so half resolution or 15 bits.
then divide it again to obtain a 25% volume from the original stream, and you get even less detail.
Is that alright? Or does this kind of attenuation work by reducing output voltage after having already made the digital to analog conversion?
Thanks, regards
Regi
Without digital attenuation, the highest digital number a sample can achieve is, with 16 bits of resolution,
1111111111111111 or 65536 steps.
Reducing volume by half provides you with a peak signal varying from 0 to 32768, so half resolution or 15 bits.
then divide it again to obtain a 25% volume from the original stream, and you get even less detail.
Is that alright? Or does this kind of attenuation work by reducing output voltage after having already made the digital to analog conversion?
Thanks, regards
Regi
I bet that this should work really nice, specially if you use the regulator, the shunts, etc...not just a direct connection from the batteries. Some people like to direct connect from batteries, but others say that there are nasty things in using it direct or just bypassing with little poly caps at the IC end.
Remove the bridge and just connect + and ground (and -).
batteries..digital volume...
I am running my B-II DAC with a battery supply for the DAC board. I used a LiFePO4 battery pack (9.6 volt nominal) form Batteryspace.com. These packs have built in management. Because the B-II needs 5-5.5 VDC, I used a Dexa 5 volt regulator with the batteries, and to assure low output impedance, I put 4 pieces of 1000 uF Cerafines at the output. On my next build I will probably just try a shunt regulated supply from a separate transformer for the digital supply. Most people suggest that really good shunt regulation is at least as good as a battery supply, so I am aiming to find out.
I believe the digital volume control on the ESS 9018 runs at 32 bits. My understanding is that 1 bit of attenuation is 6 dB, so for 24 bit files you get 8 bits of "free" attenuation (48 dB), and for 16 bit files you get 16 bits of "free" attenuation (96 dB). So I do not think data truncation with the digital volume control will result in any audible loss. On the other hand, although it makes no sense, my system sounds better running through a preamp (more dynamic, more presence-and not the kind of presence produced by noise/distortion) than run direct from the Buffalo, so I prefer to use my preamp.
I am running my B-II DAC with a battery supply for the DAC board. I used a LiFePO4 battery pack (9.6 volt nominal) form Batteryspace.com. These packs have built in management. Because the B-II needs 5-5.5 VDC, I used a Dexa 5 volt regulator with the batteries, and to assure low output impedance, I put 4 pieces of 1000 uF Cerafines at the output. On my next build I will probably just try a shunt regulated supply from a separate transformer for the digital supply. Most people suggest that really good shunt regulation is at least as good as a battery supply, so I am aiming to find out.
I believe the digital volume control on the ESS 9018 runs at 32 bits. My understanding is that 1 bit of attenuation is 6 dB, so for 24 bit files you get 8 bits of "free" attenuation (48 dB), and for 16 bit files you get 16 bits of "free" attenuation (96 dB). So I do not think data truncation with the digital volume control will result in any audible loss. On the other hand, although it makes no sense, my system sounds better running through a preamp (more dynamic, more presence-and not the kind of presence produced by noise/distortion) than run direct from the Buffalo, so I prefer to use my preamp.
Caution: layman thinking...
I have already asked that:
And the answer was:
I do believe that being an asyncronous DAC, the starting point/moment to read the digital archive will be tightly defined by your i2s source (the same for both channels).
From then onwards, at 80Mhz or 100Mhz (we are talking about 80/100 Mhz!), you will not notice any difference between channels.
As I see it, jitter interfers in conversion adding noise floor.
Stereophonic coherence is well sustained by that i2s first reference and the digital archive it is not long enough to notice some deviation between the separate clocks.
Please, engineers of this thread, confirm if such guesses are correct.
I have already asked that:
And the answer was:
With clocks running at 80mhz your not going to notice if they are not in sync.
The I2S clocks will be in sync, well because they are the same. That's all that matters.
Don't make it more complicated than it is, it won't make it sound any better.
Cheers!
Russ
I do believe that being an asyncronous DAC, the starting point/moment to read the digital archive will be tightly defined by your i2s source (the same for both channels).
From then onwards, at 80Mhz or 100Mhz (we are talking about 80/100 Mhz!), you will not notice any difference between channels.
As I see it, jitter interfers in conversion adding noise floor.
Stereophonic coherence is well sustained by that i2s first reference and the digital archive it is not long enough to notice some deviation between the separate clocks.
Please, engineers of this thread, confirm if such guesses are correct.
Well do you think it would be better then a pga 2320?
Audio - Volume Control - PGA2320 - TI.com
im running it into a dcb1 buffer afterwards.
I also need to figure out if i should do the standard iv conversion or transformer or something else for the output.
Audio - Volume Control - PGA2320 - TI.com
im running it into a dcb1 buffer afterwards.
I also need to figure out if i should do the standard iv conversion or transformer or something else for the output.
Basicly, that is correct. However according to the ESS white papers the internal width is 48 bits. If you want to attenuate in such an architecture, you put the 32-bits that can be received in the highest order bits. Only after doing that you go through the division, and then on to analog conversion. With 16 extra bits, it takes a while to loose resolution.
A PGA 2320 adds distortion to the signal. From the numbers it would seem like a worse solution. However, numbers are not important when it comes to audio. Otherwise valves and vinyl would have been ancient history long time ago, and we wouldn't have had MP3 or CDs anymore.
Same goes for the I/V stage. I think the IVY3 is excellent, others will disagree. Some will prefer transformers, some Legato. All that matters is how you think it sounds best.
Hello,
I would like to ask does it make sense to have WM8804 Receiver before Buffalo?
I have Airport N (modded, linear PSU) that might have (unsure) some jitter at optical output, so I have some options:
1. Get good glass optics, short, from AE to Pear Toslink Receiver, then straight to WM8804 chip, then Buffalo via I2S. Does WM really improve in this layout?
2. Before Buffalo arrived, I can only use WM8804 Receiver for my current DAC, using SPDIF between Receiver and DAC. This is the only connection option. The problem is: I don't have good coax cable and not planning to buy one as it is temporary solution. Is all my jitter back, even increased, if I use 1m 75-ohm antenna cable here?
Thanks a lot.
I would like to ask does it make sense to have WM8804 Receiver before Buffalo?
I have Airport N (modded, linear PSU) that might have (unsure) some jitter at optical output, so I have some options:
1. Get good glass optics, short, from AE to Pear Toslink Receiver, then straight to WM8804 chip, then Buffalo via I2S. Does WM really improve in this layout?
2. Before Buffalo arrived, I can only use WM8804 Receiver for my current DAC, using SPDIF between Receiver and DAC. This is the only connection option. The problem is: I don't have good coax cable and not planning to buy one as it is temporary solution. Is all my jitter back, even increased, if I use 1m 75-ohm antenna cable here?
Thanks a lot.
Last edited:
Yes. Basically when you change volume at the PC the result will be limited by the output sample rate.
When changing with volumite the result is better because it is done at a great bit depth.
Still you should get goods results even from the PC, but using the volumite is better.
I can hear a change with the very first step of attenuation in the computer, but additional attenuation seems not to loose any more until about -50 db. This makes sense because with any change all new data is changed (except bit shifting limited to about 6.02 db steps when increasing bit depth) and in my case this new string of numbers generated at 64 bit resolution is dithered to the 24 bit limit of the SPDIF interface. The volumite is more transparent working at both a higher sample rate and bit depth. Simply put I can't imagine a more transparent volume control for a digital source. Add a motorized pot for remote control and enjoy.
That is only true until it is output. Then you are going to hit the 24bit wall.
Transformers on Voltage Output of Buffalo II
Hi, I'm considering dual-mono Buffalo DACs to replace the stock VC-24 digital filter & S-TAC DAC in a modified Sony SCD-1. I'd like to try porting over the mods that I have in this player. This includes a balanced analog section of 1:2 Audio Consulting silver transformers and LME 49600 buffers. These are directly attached to the voltage output of the Sony S-TACT-- which in this application sounds better than using an I/V stage. Will the Buffalo DAC work well in voltage mode using a buffered transformer output? Do I need to add a HF analog filter?
I'm powering the player entirely on 12V SLA batteries and would like to continue with this approach. Can 12Vdc be used to power Placid after removal of rectifiier? Are one or two Placid regs recommended for dual-mono Buffalo?
Thanks!
Hi, I'm considering dual-mono Buffalo DACs to replace the stock VC-24 digital filter & S-TAC DAC in a modified Sony SCD-1. I'd like to try porting over the mods that I have in this player. This includes a balanced analog section of 1:2 Audio Consulting silver transformers and LME 49600 buffers. These are directly attached to the voltage output of the Sony S-TACT-- which in this application sounds better than using an I/V stage. Will the Buffalo DAC work well in voltage mode using a buffered transformer output? Do I need to add a HF analog filter?
I'm powering the player entirely on 12V SLA batteries and would like to continue with this approach. Can 12Vdc be used to power Placid after removal of rectifiier? Are one or two Placid regs recommended for dual-mono Buffalo?
Thanks!
I deleted the posts about trying to reverse a commercial product for two reasons.
1) It is is simply not necessary. You could do the same thing yourself quite easily.
2) We don't discuss consumer products on this thread.
If you find a set of filter coefficients you want to load, you can very easily do so. If someone has some filter they would like me to try they will need to send it to me, or at least tell me more about it than "Marketing Hype Filter X".
Cheers!
Russ
1) It is is simply not necessary. You could do the same thing yourself quite easily.
2) We don't discuss consumer products on this thread.
If you find a set of filter coefficients you want to load, you can very easily do so. If someone has some filter they would like me to try they will need to send it to me, or at least tell me more about it than "Marketing Hype Filter X".
Cheers!
Russ
Last edited:
I would refer you to this thread:
http://www.diyaudio.com/forums/twisted-pear/164294-buffalo-ii-transformers.html
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.