Normally I'd agree with you, but the presentation of the music with the HD800 is as if the singer is right on top of you hehe.
Thanks for the tip - this looks easy to accomplish... Soren also suggested prior a small cap here. Did you try that before changing to the TPS7A4700 adapter (I'm assuming the three pin sort instead of one that is like another linear psu)?
Yes thanks, I have. The unbuffered outputs at J7 are connected with short silver wires to a pair of hefty CMC silver rca’s. The impedance should match as all of my amp inputs have standard impedance.
I do too. I have the Salas Reflektor-D for the isolation coming shortly. Also plan to fiddle around with a battery supply.
Ahh! - you are a Can-manSorry but only via speakers may a DAC be properly judged for its ability to reproduce a music event. This is my experiance.
//
No need to be sorry, I just disagree with you.
So you mean to say that this dac is not suited for cans, or...? The other r2r-ladder dacs I've tried reallly made my cans sing. (yeah, I'm repeating myself, but I have to make a reply).
Søren himself has mentioned on a couple of occasions that he uses his HD650 to test the dac.
Standard is 10K and 100K, which represent quite different load to the DAC. Better clarify.
I am in camp Loudspeakers
TNT, I am not a headphones fan. That said, except stereo imaging and impact, loudspeakers cannot beat headphones in any other area. Most people (including me who has spent money on proper room treatments) listen to the interaction of loudspeakers and room. Many designers and manufacturers evaluate via headphones especially for noise and resolution.
Ahh! - you are a Can-man
//
TNT, I am not a headphones fan. That said, except stereo imaging and impact, loudspeakers cannot beat headphones in any other area. Most people (including me who has spent money on proper room treatments) listen to the interaction of loudspeakers and room. Many designers and manufacturers evaluate via headphones especially for noise and resolution.
Don't ruin your fun!
Don't let talk about Rolls Royce ruin the fun with your Golf.
Consider the fact that now you can now BUY a well designed R-2R DAC and can tinker with it at will. As others have said, if you ruin it, buy another one.
Not only that. Now, for the first time, you can load filters. When was the last time you were able to upgrade the firmware of your DAC?
And: add a transformer, connect some wires and you are ready to HEAR MUSIC. This diy module was designed for listening to music. Other projects are designed for modding
I also have to believe that Soren, being a h/w designer by profession knows a bit of good engineering. The selection of components are those that are appropriate for the performance of the entire system.
Often, the critical area of a DAC is the reference voltage. I have been trying to measure ripple or noise on Vref, but my crappy scope and setup picks up more noise (from wires, environment, etc) than what I can measure. Maybe some of you with sophisticated equipment and set up can measure Vref while the DAC is playing.
Consider that the variation of the resistors ( 0.5% in my case) is the equivalent of 2 mV variation on a 4V Vref. If Vref is regulated at better than 2 mV, any improvements would probably not make any difference.
Don't let talk about Rolls Royce ruin the fun with your Golf.
Consider the fact that now you can now BUY a well designed R-2R DAC and can tinker with it at will. As others have said, if you ruin it, buy another one.
Not only that. Now, for the first time, you can load filters. When was the last time you were able to upgrade the firmware of your DAC?
And: add a transformer, connect some wires and you are ready to HEAR MUSIC. This diy module was designed for listening to music. Other projects are designed for modding
I also have to believe that Soren, being a h/w designer by profession knows a bit of good engineering. The selection of components are those that are appropriate for the performance of the entire system.
Often, the critical area of a DAC is the reference voltage. I have been trying to measure ripple or noise on Vref, but my crappy scope and setup picks up more noise (from wires, environment, etc) than what I can measure. Maybe some of you with sophisticated equipment and set up can measure Vref while the DAC is playing.
Consider that the variation of the resistors ( 0.5% in my case) is the equivalent of 2 mV variation on a 4V Vref. If Vref is regulated at better than 2 mV, any improvements would probably not make any difference.
Wonder how well these combine with SCHITT components?
With the withering away of meaning in the English language you might usher in a new life for the phrase, "that sounds like CRAP" which will have become a compliment!
The price of the Discrete Monica is for a handbuilt fully assembled, functioning and tested dac. Even including some tailormade outputs at my request.
If we look at the components used, the Monica is about the same level as the DAM. Some more caps, but no fancy brands.
So my point still remains, it should be possible to come up with a diy dac that can compete with any high end r2r dac without breaking the bank.
SPDIF input
Hi Soekris!
I just received my R2R board today;-) I'm planning to use also my raspberry pi with attached hifiberry digi- board as coax SPDIF- source. The digi- board's coaxial SPDIF output is realized with a transformer. So, in this case, do I need an additional transformer on the R2R dac's input side? And if not, how is the supply of the 1.2 Voltage to handle?
Thanx!
Hi Soekris!
I just received my R2R board today;-) I'm planning to use also my raspberry pi with attached hifiberry digi- board as coax SPDIF- source. The digi- board's coaxial SPDIF output is realized with a transformer. So, in this case, do I need an additional transformer on the R2R dac's input side? And if not, how is the supply of the 1.2 Voltage to handle?
Thanx!
Søren-
As a few people here are using this strategy I thought this would be an appropriate question for this forum.
I have been running my subwoofer amp from the Buffered outputs (SE- +/GND) and my main amp from the unbuffered outputs simultaneously. When one of the rcas came unplugged I discovered that the sub amp is introducing phase cancellation in its mono summing circuit.
If I reverse one of the buffered output connections on the board (+/GND to -/GND) will it fix the phase issue in the summing circuit?
Also- would this introduce crosstalk into the direct outputs or otherwise risk harming anything on the DAM?
As a few people here are using this strategy I thought this would be an appropriate question for this forum.
I have been running my subwoofer amp from the Buffered outputs (SE- +/GND) and my main amp from the unbuffered outputs simultaneously. When one of the rcas came unplugged I discovered that the sub amp is introducing phase cancellation in its mono summing circuit.
If I reverse one of the buffered output connections on the board (+/GND to -/GND) will it fix the phase issue in the summing circuit?
Also- would this introduce crosstalk into the direct outputs or otherwise risk harming anything on the DAM?
I measured R 2R ladder of the DAM1021 plate.
R, 2R and tolerance graphs measured values.
2R:9967 Oh. (4990+4990-13)
R: 4893.5 Oh. (4890 = (4990 // 245K) + 3.52 2 ladder: 245K = 47 R + 2R + 2R joined by mass is 4890). +3.5 Oh mass resistance).
The 10 first 2R resistors clearly appear of 0.05% in the four ladder.
In the four ladder, R next output decreases resistance , the circuit must be something I do not understand.
R, 2R and tolerance graphs measured values.
2R:9967 Oh. (4990+4990-13)
R: 4893.5 Oh. (4890 = (4990 // 245K) + 3.52 2 ladder: 245K = 47 R + 2R + 2R joined by mass is 4890). +3.5 Oh mass resistance).
The 10 first 2R resistors clearly appear of 0.05% in the four ladder.
In the four ladder, R next output decreases resistance , the circuit must be something I do not understand.
Attachments
I would really like that version n°2 had little holes for us to replace the resistors with the ones of our preference.
Forza!, dear Soren.
M.
How did you measure? "in circuit" and DAC not powered I suppose.
If you have a close look on the PCB, some of your assumptions on the topology seem not to be correct.
You see that "2R" is a 10k (103) with something big (47E) in parallel giving in total something slightly smaller 10k. The backbone of the ladder are the black ones with a resistance of around 5k.
I would really like that version n°2 had little holes for us to replace the resistors with the ones of our preference.
Forza!, dear Soren.
M.
what r would u use ?
would think ps will have more influence than r tolerances
available
I packed by lab away back in 2009, but maybe Sørens DAC can get me back into the game.
Many years ago when working in mobile telecom I discovered how sensitive crystal oscillators are to vibration. A crystal is after all an electro-mechanical device. I have never worked with anything as small as the Si514, but it has a crystal inside, and I bet it does a good job
of converting sound and other vibration into jitter. A clock can measure great in a silent lab, but real life jitter is probably never single digit picoseconds.
When I modified my first CD player some 20+ years ago I dampened the cabinet thoroughly but left the new clock naked.
Listening with and without the cover on made a huge difference. Allowing sound at normal listening levels to have direct
access to the naked board really messed things up.
Next time I modified a player I dampened the cabinet the same way, but I cast the new clock in silicone inside a shielded can
and placed it right next to the filter chip. The DAC chips (TDA1541) were also dampened. This time listening with and without
the cover on there was almost no difference, but then again it was a different player.
DAC chips and some other components are slightly piezo-electric and are therefore microphonic.
Due to the size of the DAM1021 and the use of discrete resistors I think it could benefit a lot from proper acoustic
shielding and vibration dampening.
I think I read somewhere that TotalDAC has encapsulated resistors. This is probably not just for keeping things secret.
I am not suggesting casting the whole board in silicone. There are other ways to do it that also caters for good thermal management.
The heat that the board produces needs to be dissipated away.
I am considering making a conceptual drawing and posting it in the implementation thread.
Some seem to be using very small transformers.
Smaller transformers have higher winding resistances that can lead to severe secondary clipping even with modest capacitor banks.
When the harmonics contained within the clipped waveform reach the capacitor bank, they are turned into current based harmonics
in the ground plane, to which everything else is referenced.
They may no longer be visible on the spectrum analyzer, but they are still there.
I think this DAC deserves a proper unclipped power supply and good shielding/dampening before posting any negative reviews.
Maybe I will just start unpacking my lab.
Simon.
Many years ago when working in mobile telecom I discovered how sensitive crystal oscillators are to vibration. A crystal is after all an electro-mechanical device. I have never worked with anything as small as the Si514, but it has a crystal inside, and I bet it does a good job
of converting sound and other vibration into jitter. A clock can measure great in a silent lab, but real life jitter is probably never single digit picoseconds.
When I modified my first CD player some 20+ years ago I dampened the cabinet thoroughly but left the new clock naked.
Listening with and without the cover on made a huge difference. Allowing sound at normal listening levels to have direct
access to the naked board really messed things up.
Next time I modified a player I dampened the cabinet the same way, but I cast the new clock in silicone inside a shielded can
and placed it right next to the filter chip. The DAC chips (TDA1541) were also dampened. This time listening with and without
the cover on there was almost no difference, but then again it was a different player.
DAC chips and some other components are slightly piezo-electric and are therefore microphonic.
Due to the size of the DAM1021 and the use of discrete resistors I think it could benefit a lot from proper acoustic
shielding and vibration dampening.
I think I read somewhere that TotalDAC has encapsulated resistors. This is probably not just for keeping things secret.
I am not suggesting casting the whole board in silicone. There are other ways to do it that also caters for good thermal management.
The heat that the board produces needs to be dissipated away.
I am considering making a conceptual drawing and posting it in the implementation thread.
Some seem to be using very small transformers.
Smaller transformers have higher winding resistances that can lead to severe secondary clipping even with modest capacitor banks.
When the harmonics contained within the clipped waveform reach the capacitor bank, they are turned into current based harmonics
in the ground plane, to which everything else is referenced.
They may no longer be visible on the spectrum analyzer, but they are still there.
I think this DAC deserves a proper unclipped power supply and good shielding/dampening before posting any negative reviews.
Maybe I will just start unpacking my lab.
Simon.
Last edited:
petition for more FIR2 tabs
With my amplifier/loudspeaker I was quite happy with the DAM and the stock filter (attenuated so that it can not clip) with 96kHz+ material.
But now I use a headphone (Stax SRM1 mk2 + SR Lambda Pro ... a bit vintage). With that the sound was somehow unpleasant. That surprised me, as with the Stax listening was always very relaxed, although the details and stage were better over the speakers.
After some thinking and experiments it seems that the headphone amp (SRM1) has some problems with the high frequencies far from the audio band (suspect the images beyond the nyquist frequency of the intermediate stage, which through nonlinear effects in the (old) headphone amp can lead to audible distortions (my amplifier for the speakers has specified 300kHz bandwidth and the speaker crossover and tweeters seem less affected).
I tried so make a steeper FIR2 filter with rephase. I would have liked to have something that has no influence up to 20kHz and -120dB or more at the nyquist frequency of the intermediate stage. That seems to next to impossible with the available 120 tabs. But I was able to produce something steeper than the stock FIR2 and that made the sound already more relaxed with the Stax.
So
@ Soren would it be possible to increase the number of tabs for the FIR2 filter (Pleeeease). At least for symmetric filters it should be possible to double the number without using much more resources.
With my amplifier/loudspeaker I was quite happy with the DAM and the stock filter (attenuated so that it can not clip) with 96kHz+ material.
But now I use a headphone (Stax SRM1 mk2 + SR Lambda Pro ... a bit vintage). With that the sound was somehow unpleasant. That surprised me, as with the Stax listening was always very relaxed, although the details and stage were better over the speakers.
After some thinking and experiments it seems that the headphone amp (SRM1) has some problems with the high frequencies far from the audio band (suspect the images beyond the nyquist frequency of the intermediate stage, which through nonlinear effects in the (old) headphone amp can lead to audible distortions (my amplifier for the speakers has specified 300kHz bandwidth and the speaker crossover and tweeters seem less affected).
I tried so make a steeper FIR2 filter with rephase. I would have liked to have something that has no influence up to 20kHz and -120dB or more at the nyquist frequency of the intermediate stage. That seems to next to impossible with the available 120 tabs. But I was able to produce something steeper than the stock FIR2 and that made the sound already more relaxed with the Stax.
So
@ Soren would it be possible to increase the number of tabs for the FIR2 filter (Pleeeease). At least for symmetric filters it should be possible to double the number without using much more resources.
100% agree. Many have expressed preference for the NOS filter which, if using the correctly implemented version, is -1dB at 14kHz and -3dB at 20kHz.
Oddly enough I was going suggest last night that Sebastian's comments were much like a Bentley owner complaining that the ride of a Lexus was harsh and unrefined.
I know that my DAM1021 with about 2 hours run time was clearly better in definition and sound stage than a NAD M51, to the point the owner of the M51 wanted to know what was required to build a DAM. The tonal balance and frequency response of the DAM and M51 was very similar.
John Atkinson wrote in May 2014 "I've been using NAD's Masters Series M51 Direct Digital D/A processor ($1999) as one of my references since summer 2013. Jon Iverson had enthused about the M51 in his July 2012 review, concluding that he preferred DACs "that reveal as much as possible about what was captured on the tape or in the digits, and couldn't care less about adding a rose-colored tint to dodgy digital sound."
I firmly believe that the DAM1021 has been designed with the intent of being totally accurate to the source. If you like your music to come with a "rose-coloured tint" you will have add those "filters" yourself - be they trafo's, tubes, class-a output stages, etc, etc.