I installed 1.19 firmware with new 4k filters.
i have feeling that new soft filter (i think is F7) has different sound signature as old NOS. Scope shows similar pattern of sine waves as for example ak4490,ak4495, ak4497 SUPER SOFTqazi nos filter not pure NOS square sine pattern as old one But when i measured output gain from 10-20khz response goes up from 3,9V to 4,8Vpp. Why? i think that is the reason that sound is not so relaxed as i wish and stage is quite deep but not so wide left to right and top to bottom as expected.
It is any post high frequency attenuation filter included in the script? I think it should goes minus 1,5db to 20khz not plus 1,5db to 20khz.
Did anybody measured frequency response of new filters?
i have feeling that new soft filter (i think is F7) has different sound signature as old NOS. Scope shows similar pattern of sine waves as for example ak4490,ak4495, ak4497 SUPER SOFTqazi nos filter not pure NOS square sine pattern as old one But when i measured output gain from 10-20khz response goes up from 3,9V to 4,8Vpp. Why? i think that is the reason that sound is not so relaxed as i wish and stage is quite deep but not so wide left to right and top to bottom as expected.
It is any post high frequency attenuation filter included in the script? I think it should goes minus 1,5db to 20khz not plus 1,5db to 20khz.
Did anybody measured frequency response of new filters?
Hi,
I'm in search of a "natural sound" DAC and decided to buy the DAM1021. Been reading through several pages and I didn't manage to solve all my Qs, so here we go:
- Can J2 power an Arduino?
- Pops: does still the positive rail drop faster than the negative rail on power off?
Thank you so much
I'm in search of a "natural sound" DAC and decided to buy the DAM1021. Been reading through several pages and I didn't manage to solve all my Qs, so here we go:
- Can J2 power an Arduino?
- Pops: does still the positive rail drop faster than the negative rail on power off?
Thank you so much
OK that makes sense. Thank you.
What's the use case scenario for J2 power rails then? There are 1.2V 3.3V and even +-5V. I can imagine maybe powering a Toslink board? What else?
Peter,
Interesting assessment. I'm VERY interested to learn more about doing a separate clock regulator... funny noone found that before.
As you can see in the pix in this post:
Soekris' DAC implementations
... I can't easily see the clock / regulator area due to the regulator mounted directly over it. Can you share a picture of that area with the resistor marked? That way I can get a couple additional regs on the way and prepare to do that mod once they arrive.
Also interested in hearing/seeing more about your implemention of the 'superreg' transistor mod. I did the capacitance-only Vref mods along with Paul's (may his website RIP) recommended resistor and capacitor changes. I've been intrigued by the 'superreg' mod and think it is something I need to try.
Since doing the mods outlined in the post and a few following it linked above, I've also gotten 4 recent-implementation recent ESS DAC chip-based DACs running... 2 Twisted Pear Buffalo-IIIPro implementations, Ian's prototype ES9028Q2M-based RPi board using Onetics transformer outputs, and the still-being-tweaked Allo Katana. At their best, they beat my modified Soekris in detailing and start to get near it in naturalness, thought the Soekris still has that 'R2R' sound that is closer to good vinyl than most digital reproduction. Given that continued strength of the modified Soekris, I'm looking at ways to increase the level of detailing... your two mods sound like promising directions to pursue.
THANKS!
Greg in Mississippi
Last edited:
Hi Greg
What i called a "superreg" mod is known in this thread as the BC transistors mod. It involves the addition of a pass transistor controlled by the opamp output and provides a lower and more uniform with frequency output impedance from the regulators. Sorry, don't have a link to the specific posts but it should be easy enough to find.
To power up the clock from its own regulator you have to remove the 1R resistor seen in the pic. The bottom pad of this resistor is directly connected to the PS pin of the clock and this is where an independent regulator can be connected. I got the raw dc and ground connection from the pads of the onboard 3.3v regulator which continues to power up the microcontroller. Waiting for an LT3042 to arrive for this particular spot.
Yesterday i tried something else. Connected a spare FIFO from Ian between the JLSounds usb/i2s board and the DAM. Simple 317 regulator, crystek 45/49. The reason was idle curiosity. Perhaps if the input jitter is very low, the servo PLL will not have to do much work. Not sure if this makes sense, but at least it was easy enough to try. Used micro coax cables between the FIFO and the DAM, which were also quite short, perhaps 5-6cm each.
Not sure what exactly was i expecting to hear, only the sound turned unexpectedly bad. To such an extent that i had to check twice if my speakers were playing with a correct phase. All resolutions played fine with no technical glitches, only the sound was compressed, thin and bright and the perception of acoustic space was completely ruined. I use the same FIFO setup in my 9018 and it sounds completely different there. Took the FIFO out and the DAM went back to its glorious sound. Quite baffled.
What i called a "superreg" mod is known in this thread as the BC transistors mod. It involves the addition of a pass transistor controlled by the opamp output and provides a lower and more uniform with frequency output impedance from the regulators. Sorry, don't have a link to the specific posts but it should be easy enough to find.
To power up the clock from its own regulator you have to remove the 1R resistor seen in the pic. The bottom pad of this resistor is directly connected to the PS pin of the clock and this is where an independent regulator can be connected. I got the raw dc and ground connection from the pads of the onboard 3.3v regulator which continues to power up the microcontroller. Waiting for an LT3042 to arrive for this particular spot.
Yesterday i tried something else. Connected a spare FIFO from Ian between the JLSounds usb/i2s board and the DAM. Simple 317 regulator, crystek 45/49. The reason was idle curiosity. Perhaps if the input jitter is very low, the servo PLL will not have to do much work. Not sure if this makes sense, but at least it was easy enough to try. Used micro coax cables between the FIFO and the DAM, which were also quite short, perhaps 5-6cm each.
Not sure what exactly was i expecting to hear, only the sound turned unexpectedly bad. To such an extent that i had to check twice if my speakers were playing with a correct phase. All resolutions played fine with no technical glitches, only the sound was compressed, thin and bright and the perception of acoustic space was completely ruined. I use the same FIFO setup in my 9018 and it sounds completely different there. Took the FIFO out and the DAM went back to its glorious sound. Quite baffled.
Attachments
I don't know what was envisioned, but I would not use the dac to power other devices