The author of HQplayer Jussi Laako released design of this dac on Apr 18.
It's a pure DSD 1-bit 33-level dac with full support from HQplayer.
Signalyst
Anyone interested?
Interested in getting boards? http://www.diyaudio.com/forums/digital-line-level/254935-signalyst-dsc1-6.html#post4175377
Important grounding and gain modifications: http://www.diyaudio.com/forums/digital-line-level/254935-signalyst-dsc1-2.html#post4104608
It's a pure DSD 1-bit 33-level dac with full support from HQplayer.
Signalyst
Anyone interested?
Interested in getting boards? http://www.diyaudio.com/forums/digital-line-level/254935-signalyst-dsc1-6.html#post4175377
Important grounding and gain modifications: http://www.diyaudio.com/forums/digital-line-level/254935-signalyst-dsc1-2.html#post4104608
Last edited:
Network resistors
I'm wondering about Tolerances of Network Resistors and other parts (filter) are suitable for success.
I'm wondering about Tolerances of Network Resistors and other parts (filter) are suitable for success.
I have been talking to Jussi today about this as well as looking into costs to have some DAC & PS boards manufactured here locally (Denver). The reason is that Jussi has NO INTEREST in providing hardware.Anyone interested?
So, my initial thought is have 5 - 10 of each board manufactured depending on how the cost estimates come out. I know that several member of Computer Audiophile forum have expressed interest in the DSC1. Besides, if I am making 1 board it is easy enough and more cost effective to make several.
I am NOT a DIY person so be interested in teaming up with someone who can assemble the boards if we wanted to provided assemblied boards. My contribution, besides getting boards manufactured, could also be sourcing parts to populate the board using Jussi's provided BOM.
Just doing a lot of "what if" thinking at the moment but saw this thread & thought I would throw it out there.
I have been talking to Jussi today about this as well as looking into costs to have some DAC & PS boards manufactured here locally (Denver). The reason is that Jussi has NO INTEREST in providing hardware.
So, my initial thought is have 5 - 10 of each board manufactured depending on how the cost estimates come out. I know that several member of Computer Audiophile forum have expressed interest in the DSC1. Besides, if I am making 1 board it is easy enough and more cost effective to make several.
I am NOT a DIY person so be interested in teaming up with someone who can assemble the boards if we wanted to provided assemblied boards. My contribution, besides getting boards manufactured, could also be sourcing parts to populate the board using Jussi's provided BOM.
Just doing a lot of "what if" thinking at the moment but saw this thread & thought I would throw it out there.
I just populated mine and still waiting for my amanero.
here's my BOM https://docs.google.com/spreadsheets/d/1Lm1L7b_GjFbQFs8_TOuCNU8q3gvgvTgiFdx_7qeBoPM/edit?usp=sharing
Feel free to comment and change things.
cost wise
pcb: depends on batch size
parts: about 130 sourcing from mouser and 20 for transformers from antek. 97+shipping for amanero
case: 0-unlimited
Last edited:
I'm wondering about Tolerances of Network Resistors and other parts (filter) are suitable for success.
For the 15k ones used by design the best I can find is 2%.
It wouldnt be too hard to alter the board to use SMD resistors which can go 0.02% for less $100.
Thanks mcluxun!!! Very much appreciated.
Where did you get your boards? If you had them fabricated, which shop?
One more question, are the outputs balanced or single ended?
Where did you get your boards? If you had them fabricated, which shop?
One more question, are the outputs balanced or single ended?
Last edited:
Thanks mcluxun!!! Very much appreciated.
Where did you get your boards? If you had them fabricated, which shop?
A friend of mine made those for me.
Usually oshpark is my to go place for small batches.
Are the outputs balanced or single ended? I didn't see anything on Jussi's website.
single ended
According to Jussi, a new version (DSC2) is in the works with balanced output which would better for me. 😀single ended
According to Jussi, a new version (DSC2) is in the works with balanced output which would better for me. 😀
I would love to see a DSC2 design.
But before that I need to finish my DSC1.
As far as SE/Bal concerns there are lots of solutions out there to transform from one to another.
To me, a pair of transformers can get this job done nice and easy.
But before that I need to finish my DSC1.
I trust that you will keep updating this thread as your DSC1 progresses as I am very interested in your efforts!! I know that several people over on the Computer Audiophile forum would be interested as well given that HQ Player has gained in popularity over the past few months.
List of fixes I have made (I'll update the documents when I have time):
1) Either change I/V feedback resistor R1/R2 to 1 kOhm (this is what I did), or drop resistance of the network (8k network would be optimal).
2) IC11/IC12 non-inverting input pin 3 should be the point where DGND and AGND are connected, I added jump wires to the pin. While doing this I cut the AGND-DGND R18 (bead) connection on PSU board.
3) IC11/IC12 output pin 6 should be connected to AGND with 100k resistor, I connected from C520/C530 input side.
1) Either change I/V feedback resistor R1/R2 to 1 kOhm (this is what I did), or drop resistance of the network (8k network would be optimal).
2) IC11/IC12 non-inverting input pin 3 should be the point where DGND and AGND are connected, I added jump wires to the pin. While doing this I cut the AGND-DGND R18 (bead) connection on PSU board.
3) IC11/IC12 output pin 6 should be connected to AGND with 100k resistor, I connected from C520/C530 input side.
Got mine running with xmos.
DSC01277 by https://www.flickr.com/people/76879216@N08/, on Flickr
DSC01272 by https://www.flickr.com/people/76879216@N08/, on Flickr
where i2s to dsd matches as:
xmos DSC1
DATA DSDR (pin3)
SCLK BCLK (pin4)
LRCK DSDL (pin5)
3.3v DSDEN (pin7)
DSC01273 by https://www.flickr.com/people/76879216@N08/, on Flickr
My xmos is running with old firmware which goes up to 3072000Hz and DoP only on HQplayer.
under uint8 mode HQ doesnt allow me to select sample rate.
(Jussi what's your setup with amanero on HQ? and ever try to run it with XMOS?)
In that case the noise is way too loud to enjoy music.
It's a problem related to the rate because I can clearly hear differences between 2822400 (dsd64) and 3072000.
Only thing i can tell right now is the music is flowing very well, definitely different from anything Ive heard.
And strangely this thing can sing with out power ??????


where i2s to dsd matches as:
xmos DSC1
DATA DSDR (pin3)
SCLK BCLK (pin4)
LRCK DSDL (pin5)
3.3v DSDEN (pin7)

My xmos is running with old firmware which goes up to 3072000Hz and DoP only on HQplayer.
under uint8 mode HQ doesnt allow me to select sample rate.
(Jussi what's your setup with amanero on HQ? and ever try to run it with XMOS?)
In that case the noise is way too loud to enjoy music.
It's a problem related to the rate because I can clearly hear differences between 2822400 (dsd64) and 3072000.
Only thing i can tell right now is the music is flowing very well, definitely different from anything Ive heard.
And strangely this thing can sing with out power ??????
Last edited:
under uint8 mode HQ doesnt allow me to select sample rate.
uint8 mode is just for custom special hardware that supports 8-bit "PCM" at 1/8th DSD rate. So it shouldn't be used. Either DoP or ASIO DSD (subject to driver support). For ASIO DSD, set SDM Pack to "none" and the driver support will be auto-detected.
(Jussi what's your setup with amanero on HQ? and ever try to run it with XMOS?)
I'm using ASIO driver in native DSD mode (SDMPack=none) and DSD5v3 or DSD7 modulator (DSD5v3 modulator will appear in HQP 3.5.0). Filter is poly-sinc-short-mp-2s. Also DoP at 6.1 MHz works.
I didn't try XMOS yet, I'll try to get some board for the purpose and try it out.
In that case the noise is way too loud to enjoy music.
It's a problem related to the rate because I can clearly hear differences between 2822400 (dsd64) and 3072000.
I tried also DSD64 and didn't encounter any significant noise problems, although the Sallen-Key filter's fc is too high for it and it will also have some roll-off already at 20 kHz because fc of the conversion stage scales linearly with sampling rate. Taking the last 8-bit shift register out and changing cut-off of the Sallen-Key would improve DSD64, but of course with a trade-off in performance.
And strangely this thing can sing with out power ??????
Without PSU? IIRC, there shouldn't be output without power to the output mute relay. And of course none of the active parts would operate correctly.
The mute actually performs really nice with Amanero, so I don't get any extra noises.
Here's picture how I implemented the fixes. Notice short wires for the I/V pin 3, I just removed a bit of solder stop mask from the DGND ground plane and used short L-wires from the pin to the nearest DGND plane point.
An externally hosted image should be here but it was not working when we last tested it.
xmos DSC1
DATA DSDR (pin3)
SCLK BCLK (pin4)
LRCK DSDL (pin5)
3.3v DSDEN (pin7)
I note one criticial thing missing here and in your pictures. You need to feed 3.3V supply to the Amanero interface pin 10 (top row, left-most pin). You can take this from the DSC1PSU 3V3 which is unused at the moment with Amanero board (since Amanero's feed is used). This power is needed to run the IC1 and IC2.
For the SJ1 jumper, I have the top part connected (middle and upper pads).
Jussi, if AGND is pin3 of ad844, so where is DGND point? My chose is the GND closed Amanero, which is after the decouling of psu.
Jussi, if AGND is pin3 of ad844, so where is DGND point? My chose is the GND closed Amanero, which is after the decouling of psu.
If I understand it correctly it's where those two GNDs merge.
uint8 mode is just for custom special hardware that supports 8-bit "PCM" at 1/8th DSD rate. So it shouldn't be used. Either DoP or ASIO DSD (subject to driver support). For ASIO DSD, set SDM Pack to "none" and the driver support will be auto-detected.
I'm using ASIO driver in native DSD mode (SDMPack=none) and DSD5v3 or DSD7 modulator (DSD5v3 modulator will appear in HQP 3.5.0). Filter is poly-sinc-short-mp-2s. Also DoP at 6.1 MHz works.
I didn't try XMOS yet, I'll try to get some board for the purpose and try it out.
I tried also DSD64 and didn't encounter any significant noise problems, although the Sallen-Key filter's fc is too high for it and it will also have some roll-off already at 20 kHz because fc of the conversion stage scales linearly with sampling rate. Taking the last 8-bit shift register out and changing cut-off of the Sallen-Key would improve DSD64, but of course with a trade-off in performance.
Without PSU? IIRC, there shouldn't be output without power to the output mute relay. And of course none of the active parts would operate correctly.
The mute actually performs really nice with Amanero, so I don't get any extra noises.
Here's picture how I implemented the fixes. Notice short wires for the I/V pin 3, I just removed a bit of solder stop mask from the DGND ground plane and used short L-wires from the pin to the nearest DGND plane point.
An externally hosted image should be here but it was not working when we last tested it.
Just confirmed that my xmos is running older version of firmware and can not go beyond dsd64, I had to try higher rate with amanero.
As far as the pin connection goes. Do I need 3.3v on other pins: 7, 10 and 11.
And for SCJ1 I should short 1 and 2?
Last edited:
- Home
- Source & Line
- Digital Line Level
- Signalyst DSC1