It seems that the minimum detectable interaural time difference is 10 microseconds. In a dual-mono implemention of the dam1021, can someone confirm that the lock time difference will always be only a few microseconds?
It probably won’t make a difference at all for speaker folks but as a headphone user I’m still slightly concerned...
It probably won’t make a difference at all for speaker folks but as a headphone user I’m still slightly concerned...
I do it with the terminal and the screen command as described here:
http://wiki.soekris.info/Connecting_to_the_serial_console
I can at least say that using two independent DAM for L and R channel sounds really good. Super stable and solid soundscape. But I'm still interested in the answer. I use an optical (passive) toslink splitter to ensure that there is no time skew created by the s/pdif distribution to the 2 DACs. I would never trust USB etc to do this however - it must be a synchronous transfer mechanism.
I did ask Sören before (should be in this thread) of any eventual random behaviour of the the handling of audio frames in the buffer etc. but he assured that there was non. Made me go ahead with my plan and it has panned out very good as I now sit and listen to 1. (Sonata for oboe and piano (1938)) - I. Munter
Ogrintchouk, Alexei, Leonid, Hindemith, Paul
A 20th-Century Recital (1970)
2350kbps VBR, flc
using two 5 meter glass toslink cable, each feeding one channel.
//
Well, when even the folks over at SBAF are partially divided whether the Yggi or the DAC1541 is better, then we can definitely agree that Soeren has not botched the design completely ;-)
Disclaimer: I am a very satisfied customer of Soeren’s, and could enjoy the surprise on a vinylist friend of mine that could compare it to a high mass turntable with SOTA arm and pickup. And this without fancy filters!
Roberto
Disclaimer: I am a very satisfied customer of Soeren’s, and could enjoy the surprise on a vinylist friend of mine that could compare it to a high mass turntable with SOTA arm and pickup. And this without fancy filters!
Roberto
I know, I know
You truncate, I tested it.
As my example shows, truncate and round make a difference as the negative of a signal is not always mapped to the same state of the DAC ladder (ladders swapped) as the positive signal. (A serious lack of symmetry
Yes, the difference is at most the magnitude of the last bit. I agree that for listening this is probably not of importance.
But you miss the marketing terms "bit perfect" and "32-bit input".
OK but getting serious again, the missing features make it harder to put the DAM-ladder, with all of it bits, in a predefined state without fiddling around with several filters and tweaked signals.
Interesting choice of music I'm not yet able to appreciate the piece but I'm sure it's beautiful to those who could...
I did read Soren's reply earlier saying that there will be at most several microseconds of delay in either channel, which I could only interpret as being caused by random noise in the chips
My setup basically just shorts the IIS and SPDIF inputs on the two boards with a 20mm tall header. I think someone mentioned in this thread that this is doable. But now I'm not sure if there's synchronous transfer in IIS that could mess things up...
Why did you use two separate toslink inputs boards (i assume, with the mention of 5m toslink cables..) instead of just shorting the SPDIF inputs on the boards, which definitely introduces less variance between the signals picked up by dam1021s? Is there an impedance issue that I overlooked?...
A definitive answer from Soren would be nice...if only for the peace of mind, or the lack thereof... To me, there's value in knowing my system is well-engineered, much more so than having mystical improvements in SQ that's likely the result of wishful thinking.
I'm not yet able to appreciate the piece but I'm sure it's beautiful to those who could...I did read Soren's reply earlier saying that there will be at most several microseconds of delay in either channel, which I could only interpret as being caused by random noise in the chips
My setup basically just shorts the IIS and SPDIF inputs on the two boards with a 20mm tall header. I think someone mentioned in this thread that this is doable. But now I'm not sure if there's synchronous transfer in IIS that could mess things up...
Why did you use two separate toslink inputs boards (i assume, with the mention of 5m toslink cables..) instead of just shorting the SPDIF inputs on the boards, which definitely introduces less variance between the signals picked up by dam1021s? Is there an impedance issue that I overlooked?...
A definitive answer from Soren would be nice...if only for the peace of mind, or the lack thereof... To me, there's value in knowing my system is well-engineered, much more so than having mystical improvements in SQ that's likely the result of wishful thinking.
My DACs are 4,6 meter apart!
5. Kozo's Waltz
Art Blakey & The Jazz Messengers
A Night In Tunisia (2013)
4608kbps CBR, wav
now also sound really good
//
LOL a true "dual-mono" design
While Soren (possibly) is out on holidays and we wait for a reply, do you know if, electrically speaking, I would do myself any good by connecting both + and - signals (8 in total) from each buffered ladder to a 4-pin XLR for headphones? I'm sort of hoping this would cut output impedance by half and have more drive power, compared with only using the 4 non-inverted buffered signal from each ladder... Would this be a disaster?...
Thanks TNT, for all your inputs in this thread!
P.S. I'm still thinking about putting together a real wiki page for the dam1021 line. There are very good tutorials by Dimdim and hifiduino, as well as a wiki stub for "Soekris R2R dac" here on diyaudio. But I was more envisioning something that's updated with discussions on different aspects of the dac, to save everyone of searching/skimming through 700 pages, and allow us to focus on the potentially impactful components/design choices. It would take some efforts so I would appreciate it if someone could chime in on the value of such a wiki page.
)
When you use one board in differential mode you get 3 wires out. The drive impedance out of the ladder is quite high - in the order of several 100 ohms - not so suitable to drive a low ohmic load I would say. But maybe you mean sort of parallel config of two channel - still to high impedance, if it could work - unsure.
Majestic sound from:
1. Ballade No. 1 G-moll, Op. 23
Frédéric Chopin
4 Balladen Barcarolle Fantasie (piano Krystian Zimerman)
1411kbps CBR, wav
:-D
//
"LOL a true "dual-mono" design" - well if you see my room as the housing I suppose you are correct
Majestic sound from:
1. Ballade No. 1 G-moll, Op. 23
Frédéric Chopin
4 Balladen Barcarolle Fantasie (piano Krystian Zimerman)
1411kbps CBR, wav
:-D
//
"LOL a true "dual-mono" design
" - well if you see my room as the housing I suppose you are correct
Last edited:
You have quite a mix of playlist! I sort by composer so it's usually an entire hours of Chopin or more until I hear from someone else
))I'm mostly using dam1021 directly with headphones at the moment so I have to rely on buffered outs. For each board (channel), the output is non-inverted(1) and inverted(2) from left ladder, inverted(3) and non-inverted(4) from right ladder. There's no ground in 4-pin XLR nor does a headphone use one so I'll ignore those...
The choice for me, until Soren finds time to share with us his discrete buffer stage in the form of an add-on board / separate board..., is to use 1 and 3 for 2 pins on the XLR for each channel, or use 1&4 shorted and 3&2 shorted for each channel... The former is standard buffered out with 20R output impedance at 4V RMS, and I'm hoping the latter would be 10R output at 4V RMS, without degrading the sound and preserving the benefits of a fully balanced system of course....
But this could always blow things up and leave me with nothing but magic smoke...
LOL a true "dual-mono" design
While Soren (possibly) is out on holidays and we wait for a reply, do you know if, electrically speaking, I would do myself any good by connecting both + and - signals (8 in total) from each buffered ladder to a 4-pin XLR for headphones? I'm sort of hoping this would cut output impedance by half and have more drive power, compared with only using the 4 non-inverted buffered signal from each ladder... Would this be a disaster?...
Thanks TNT, for all your inputs in this thread!
P.S. I'm still thinking about putting together a real wiki page for the dam1021 line. There are very good tutorials by Dimdim and hifiduino, as well as a wiki stub for "Soekris R2R dac" here on diyaudio. But I was more envisioning something that's updated with discussions on different aspects of the dac, to save everyone of searching/skimming through 700 pages, and allow us to focus on the potentially impactful components/design choices. It would take some efforts so I would appreciate it if someone could chime in on the value of such a wiki page.
That could be very helpful. Being in the midst of setting one up, it is a bit of a hassle to manoeuvre in all the recommendations that goes for elderly versions of the board. So I chime in
That could be very helpful. Being in the midst of setting one up, it is a bit of a hassle to manoeuvre in all the recommendations that goes for elderly versions of the board. So I chime in
Thanks, also check page 677 for spikestabber’s suggestions on mods for rev4. I might just leave the board as is for now.
The same problem with random delays in acquiring lock among the boards exist in your case. An official answer from Soren is probably needed.
You would have the peace of mind if you could work with dam1121, but it requires more work and is more fragile.
Couldn’t find the edit function so I’ll summarize my questions here for those who might know the answers..:
(1) How much random delay would there be between two dam1021’s in a dual mono setup, either with paralleled SPDIF or I2S input? Some explanation on the causes would also be great.
(2) Considering that stacking ladders for each channel will effectively decrease the output impedance in addition to improving SNR, could I do the same, I.e. parallel outputs of the same polarity, with the onboard buffered outputs to cut output impedance by half in balanced dual mono mode? (8 buffered signals from the 4 ladders -> 4 signal, e.g. 4-pin XLR out) My intention is only to increase the drive power, while keeping SNR and signal quality the same.
(3) If the above doesn’t work or is still less than desirable for driving high impedance headphones, is there an available add-on headphone amplifier board that is relatively compact? Any suggestions here on how to integrate the pre-amp/headphone amp in the dac enclosure would be greatly appreciated!
Also, should I get my hopes up for an upcoming add-on head-amp board for dam1021? @Soren I do think that there will be plenty people interested in a discrete head-amp and/or preamp that is state-of-the-art engineering wise with the option of easily integrating with dam1021. You might also be able to remove the buffer stage in the next rev to lower costs and improve compactness. In any case, you seem to have the design already in your other products with great results, and you also mentioned a possible release earlier last year. It seems to me that it would just be GREAT to have an add-on output stage to this already excellent DAC, especially for those of us who aren’t the most interested in tube solutions. Just a (longish) thought....
(1) How much random delay would there be between two dam1021’s in a dual mono setup, either with paralleled SPDIF or I2S input? Some explanation on the causes would also be great.
(2) Considering that stacking ladders for each channel will effectively decrease the output impedance in addition to improving SNR, could I do the same, I.e. parallel outputs of the same polarity, with the onboard buffered outputs to cut output impedance by half in balanced dual mono mode? (8 buffered signals from the 4 ladders -> 4 signal, e.g. 4-pin XLR out) My intention is only to increase the drive power, while keeping SNR and signal quality the same.
(3) If the above doesn’t work or is still less than desirable for driving high impedance headphones, is there an available add-on headphone amplifier board that is relatively compact? Any suggestions here on how to integrate the pre-amp/headphone amp in the dac enclosure would be greatly appreciated!
Also, should I get my hopes up for an upcoming add-on head-amp board for dam1021? @Soren I do think that there will be plenty people interested in a discrete head-amp and/or preamp that is state-of-the-art engineering wise with the option of easily integrating with dam1021. You might also be able to remove the buffer stage in the next rev to lower costs and improve compactness. In any case, you seem to have the design already in your other products with great results, and you also mentioned a possible release earlier last year. It seems to me that it would just be GREAT to have an add-on output stage to this already excellent DAC, especially for those of us who aren’t the most interested in tube solutions. Just a (longish) thought....
Last edited:
Edit to (2): Also, if this scheme works and the output impedance is lowered to 10R, the peak power on a 300R load would be 99.9mW, and the RMS power would be 50mW (4V RMS). Consider that HD650 has 103db/mW sensitivity, 115db SPL requires 15.8mW, another common way to calculate headroom: 50x the power for 95db - 7.92mW. Also, OPA1602 lists 60mA as max current output so we should be safe there.
The RMS power seems to be way over the required headroom for HD650. Can I say then that the buffer on dam1021 is sufficient in terms of power to drive a HD650?
Thanks!
The RMS power seems to be way over the required headroom for HD650. Can I say then that the buffer on dam1021 is sufficient in terms of power to drive a HD650?
Thanks!
The existing buffer seems plenty a lot. Did you try and listen to it? If you place 2 boards in one box (dual-mono), why not synchronise them for the peace of mind? In your build, just make the direct output available so that you can connect an external amp to try and compare.
Are you in possession of boards or planning a purchase?
//
Are you in possession of boards or planning a purchase?
//
Yeah I posted my first build a few pages back. The machines chassis is still a couple days out so I haven’t soldered the second board yet just to make assembly easier. I don’t know of an existing method to sync two dam1021 and would be very interested in one...
SQ wise I can only say it beats my Precision5520 for lack of comparison.... lots of details, great dynamics, loving it. But I do feel like it could be even better in extra high dynamic sections of music/movie... just a feeling really... maybe it’ll get a bit better in dual mono
I’ve decided now to just use the buffers since I will not be able to fit the power supply necessary for an independent output stage that’s also up to my standards. I would only need to swap out my 35VA Talema mini series for a custom one with 3 secondaries - one for dam1021 and 2 for the amps, and add single rail linear regulator in addition to the sigma22 that I have at hand. However I haven’t found one that would fit... let alone the fact that we are no where close to seeing Soren release his discrete head amp add-on...
Direct raw output will be available on back panel through 3-pin XLR and 4x RCA. See my earlier posts for the front and back panel I’ll also post a more comprehensive description once I close the lid on the build
SQ wise I can only say it beats my Precision5520 for lack of comparison
.... lots of details, great dynamics, loving it. But I do feel like it could be even better in extra high dynamic sections of music/movie... just a feeling really... maybe it’ll get a bit better in dual monoI’ve decided now to just use the buffers since I will not be able to fit the power supply necessary for an independent output stage that’s also up to my standards. I would only need to swap out my 35VA Talema mini series for a custom one with 3 secondaries - one for dam1021 and 2 for the amps, and add single rail linear regulator in addition to the sigma22 that I have at hand. However I haven’t found one that would fit... let alone the fact that we are no where close to seeing Soren release his discrete head amp add-on...
Direct raw output will be available on back panel through 3-pin XLR and 4x RCA. See my earlier posts for the front and back panel
I’ll also post a more comprehensive description once I close the lid on the build