• Disclaimer: This Vendor's Forum is a paid-for commercial area. Unlike the rest of diyAudio, the Vendor has complete control of what may or may not be posted in this forum. If you wish to discuss technical matters outside the bounds of what is permitted by the Vendor, please use the non-commercial areas of diyAudio to do so.

RD30 – Quad multibit d/a converter

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
We all know how good a PCM1704 can be. Currently priced at 75$ per mono chip, and it seems to be the only 24bit R-2-R “dinosaur” available. I wanted to give to this chip that well-deserved digital front end, with a reclocker that offers superb jitter rejection(asynchronous FIFO style), a powerful digital filter, I2S, S/PDIF, AES and USB inputs (all accepting HD audio), convenient user interface with alphanumeric display and remote. I wanted the digital filter to be switched off when needed – for pure NOS operation. I wanted true balanced operation and this requires four PCM1704 on board. Could not stop it – I wanted the current outputs summed 2 by 2 for single ended output if needed – with the benefits of double current swing and double sample rate through analog linear interpolation.
The current outputs from PCM1704 are directly available, so any compatible I/V stage can be used, tube or solid-state.
The result is the RD30 dac, which took about 2 years of tests and bare r&d in all these critical areas to be completed. It is a very compact design, due to the fact that all the features are built into the powerful Xilinx Spartan6 chip XC6SLX16. Measuring only 124 x 125 mm and including all ultra-low noise power management and all input conenctors on board, only AC power is needed to fire up the RD30.
 

Attachments

  • RD-30B.jpg
    RD-30B.jpg
    173.6 KB · Views: 1,733
  • RD-30A.jpg
    RD-30A.jpg
    175.2 KB · Views: 1,689
Last edited:
Custom building blocks – FPGA :

- custom digital receiver
- asynchronous FIFO with digital PLL / DCXO, loop corner frequency 0.5 Hz
- Bytemorph 2.0 digital filter,60 bit internal, 384/768k , 1500 taps, -145 dB stopband att.*
- audio formatter and volume control , 256 steps
- user interface manager
*can be bypassed for native(NOS) mode

Digital inputs :

- AES/EBU : discrete lvds translator - 192/24
- S/PDIF : discrete lvds translator - 192/24
- USB : asynchronous, FX2LP interface – 192/24*
- I2S over HDMI (native LVDS or CMOS) – 192/24
*drivers available for Win,Mac, Linux

D/A conversion stage/supply/outputs :

- PCM 1704 x 4 multibit
- true balanced operation or single ended
- direct current output : 1.2mA x 4, 2.4mA x 2 respectively
- linear power supplies only, low noise regulators
- schottky rectifiers

User interface options :

- control buttons + led indicators
- control buttons + alphanumeric display, 4 x 20 characters
- remote control (rc5)
 
Optional front panel board :
- type A : 4x20 alphanumeric display, IR sensor, control switches
- type B : leds for sample rate signaling, IR sensor, control switches
Remote control also works with T30 transport.
 

Attachments

  • RD30_char_display.JPG
    RD30_char_display.JPG
    39.7 KB · Views: 1,573
  • RD30_char_display.pdf
    82.3 KB · Views: 172
  • RD30_LED_display.pdf
    50 KB · Views: 151
  • RD30_remote.jpg
    RD30_remote.jpg
    48.3 KB · Views: 431
Last edited:
There are many product levels :
- digital front end only (no PCM1704 on board)
- digital front end + 4 x PCM1704
Each version can be ordered with or without front panel board, either led or alphanumeric version.
A custom toroidal transformers set is available as well.
Feel free to contact me about pricing & availability.
 
Looks promising.

Folks who are expert at digital stuff have said about the made in China DACs that they are not so good due to layout and bypassing errors... I wonder about such things... have you got published data and graphs, like off an Audio Precision that shows technical performance by any chance?

Regards,

_-_-bear
 
Looks promising.

Folks who are expert at digital stuff have said about the made in China DACs that they are not so good due to layout and bypassing errors... I wonder about such things... have you got published data and graphs, like off an Audio Precision that shows technical performance by any chance?

Regards,

_-_-bear

Hi Bear,
our company is located in Europe - Romania and all products are designed and manufactured locally. There is no relation with China whatsoever.
The pcb is 4-layer with impedance control over certain signal traces. Routing BGA chips is not a trivial task, and we spent a lot of time optimizing the layout. I will post performance graphs soon.
 
Some technical questions on the RD-30:

- Are you generating the clock in the FPGA (soft PLL), or are you using crystal(s)?
- Is the Bytemorph DF oversampling to 768kHz or 384 kHz?
- Are you doing oversampling in integer multiples, or ASRC? I see only 384/768 kHz mentioned - for n x 44.1 kHz input I'd like to see 352.8 or 705.6 kHz.
 
Some technical questions on the RD-30:

- Are you generating the clock in the FPGA (soft PLL), or are you using crystal(s)?
- Is the Bytemorph DF oversampling to 768kHz or 384 kHz?
- Are you doing oversampling in integer multiples, or ASRC? I see only 384/768 kHz mentioned - for n x 44.1 kHz input I'd like to see 352.8 or 705.6 kHz.

There are 2 main clocks involved, and both are external XO's. One is a high-frequency clock involved in sampling the input signal; second - and key of the reclocking system, combined with fpga internal ram buffer - is the Si570 DCO. This allows the creation of a PLL entirely in the digital domain (all-digital-pll). The performance of the ADPLL is essentially given by the performance of the Si570, which offers a remarkably low phase noise for a DCO.
The Bytemorph DF is actually a collection of 3 independent filters; they are indeed integers (no ASRC). There is a 8x filter for 44.1k and 48k; a 4x filter for 88.2 or 96k input; and a 2x filter for 176 and 192k. The filters are switched on-the-fly when sample rate changes. I will come back with detailed DF performance in the next posts.
 
Last edited:
As promised, digital filter performance graphs in detail.
DF1 data:
Filter type : Parks McClellan linear phase equiripple
Taps : 1510
Input sample rate : 44.1k. 48k
Output sample rate : 352.8k, 384k
Stopband attenuation : -144 dB
Passband ripple : +/-0.00008 dB
 

Attachments

  • fs_overall.jpg
    fs_overall.jpg
    65.5 KB · Views: 740
  • fs_step.jpg
    fs_step.jpg
    49.5 KB · Views: 731
  • fs_transition.jpg
    fs_transition.jpg
    71.1 KB · Views: 716
DF2 data:
Filter type : Parks McClellan linear phase equiripple
Taps : 1510
Input sample rate : 88.2k , 96k
Output sample rate : 352.8k, 384k
Stopband attenuation : -145 dB
Passband ripple : +/-0.00007 dB
 

Attachments

  • 4x_overall.jpg
    4x_overall.jpg
    64.3 KB · Views: 698
  • 4x_step.jpg
    4x_step.jpg
    49.8 KB · Views: 688
  • 4x_transition.jpg
    4x_transition.jpg
    63.8 KB · Views: 191
DF3 performance:
Filter type : Parks McClellan linear phase equiripple
Taps : 1510
Input sample rate : 176.4k , 192k
Output sample rate : 352.8k, 384k
Stopband attenuation : -145 dB
Passband ripple : +/-0.00008 dB
 

Attachments

  • 2x_overall.jpg
    2x_overall.jpg
    67.8 KB · Views: 177
  • 2x_step.jpg
    2x_step.jpg
    49.2 KB · Views: 173
  • 2x_transition.jpg
    2x_transition.jpg
    73.1 KB · Views: 171
Hi Lupulroz
I'll be happy to receive it as soon as possible; I have a lot of HD Tracks at 24/192 kHz to listen. For the moment I can`t use my old DAC with CS4398 for 24/192kHz sample rates I play them from my Dune HD Max media player.

Maybe you will help me to take from this player via HDMI I2S signal for this DAC.

Look forward for more information and opinion about this DAC if someone already have it.
 
Hi Lupulroz
I'll be happy to receive it as soon as possible; I have a lot of HD Tracks at 24/192 kHz to listen. For the moment I can`t use my old DAC with CS4398 for 24/192kHz sample rates I play them from my Dune HD Max media player.

Maybe you will help me to take from this player via HDMI I2S signal for this DAC.

Look forward for more information and opinion about this DAC if someone already have it.

Hi Gabriel,
a HDMI receiver core is possible to implement in a future release of software. However this require lots of testing and there is also the problem of HDCP (digital content protection) which usually applies to HDMI.
 
Impressive set of features, well done!
I notice that the graphs in posts 13, 14 have switched places.
Any room left for different types of digital filters...minimum phase apodizing and the likes? Accesible via remote of course :)

Thank you Sidiy! You are right, I switched graphs, my bad :)
About different digital filters : that's the idea! There are enough hardware resources ,in short time a variety of filters will be available, via remote. We are working now on system to upgrade the firmware via usb input, to make the board easy upgradable in the field.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.