The problem of a standard R2R scheme is the zero crossing. Value 1 would be LSB set only, value 0 all bits zero and value -1 all bits one. This big change is likely to cause a transition effect, regardless of the level of music. In contrast, the signed magnitude architecture will have only the LSB set for +1 and -1. So the zero transition will have minor bit switches.
Still, with standard R2R with precise enough resistors, I think you will do better than 8 or 10 bits. The remaining bits up to 24 bit will at least reduce the quantisation effects / noise. Though it will not improve linearity.
Fedde
Still, with standard R2R with precise enough resistors, I think you will do better than 8 or 10 bits. The remaining bits up to 24 bit will at least reduce the quantisation effects / noise. Though it will not improve linearity.
Fedde
Since i have a very limited knowledge of SMD devices other than caps and resistors - I am wondering:
Are there voltage regulators on the 1941 board or would the user supplied power supply be the only regulation applied to the voltages?
I have had a 1941 for almost six months and have yet to get it up and running, I am ashamed to say.
Just need to connect the SALAS BIBs but I am always concerned about cascading regulators.
If anyone knows please specify for all three rails, please.
THANKS!
Are there voltage regulators on the 1941 board or would the user supplied power supply be the only regulation applied to the voltages?
I have had a 1941 for almost six months and have yet to get it up and running, I am ashamed to say.
Just need to connect the SALAS BIBs but I am always concerned about cascading regulators.
If anyone knows please specify for all three rails, please.
THANKS!
A dam19x1 is so easy to get running, so you should be a little ashamed
Just connect power and digital source and take output directly. You can use the firmware defaults, no need to mess with anything.
The two +- analog power just feed the vref regulators.
The digital supply feeds all digital parts with multiple regulators for the lower voltages, and the 570 oscillator have its own LP5907.
To get started you can use just a couple of simple +5V and -5V supplies, you get a good result even with that.
There is a default volume for each setting, line out or headphones. You can set the power up defaults, see dac1541 manual.
I now remember that I swapped channels for headphones out. The dam19x1 firmware is based on the dac1541 firmware, basically the same, and I managed to wire the headphone outputs incorrectly and fixed that in firmware....
Plan that the next dam19x1 firmware release will be corrected. Or you can also just swap the wires to the headphone connector.
I prefer the sound of gear without active voltage regulation.
The only thing "worse" than a regulator is having them in cascade.
I am going to try "brute force" lclc supplies first. And then if disappointed can return to the SALAS project.
Yes, I am ashamed to have put this off so long.
An even more embarrassing confession is that in the interim I have been using the FOCUSRITE Scarlett box I use for making measurements with REW as my DAC.
Doing adjustments on the loudspeakers has become an obsession that has taken me away from soldering projects. I highly recommend the XILICA DSP boxes but they can lead to obsession ...
I hope I am in for the biggest surprise of my life but that two hundred dollars box sounds better than it has any right to but then I am not a gear swapper so I have no basis.
Hoping night and day difference is what I hear.
Thanks for the advice and admonition!
Take care,
The only thing "worse" than a regulator is having them in cascade.
I am going to try "brute force" lclc supplies first. And then if disappointed can return to the SALAS project.
Yes, I am ashamed to have put this off so long.
An even more embarrassing confession is that in the interim I have been using the FOCUSRITE Scarlett box I use for making measurements with REW as my DAC.
Doing adjustments on the loudspeakers has become an obsession that has taken me away from soldering projects. I highly recommend the XILICA DSP boxes but they can lead to obsession ...
I hope I am in for the biggest surprise of my life but that two hundred dollars box sounds better than it has any right to but then I am not a gear swapper so I have no basis.
Hoping night and day difference is what I hear.
Thanks for the advice and admonition!
Take care,
A dam19x1 is so easy to get running, so you should be a little ashamed
Just connect power and digital source and take output directly. You can use the firmware defaults, no need to mess with anything.
The two +- analog power just feed the vref regulators.
The digital supply feeds all digital parts with multiple regulators for the lower voltages, and the 570 oscillator have its own LP5907.
To get started you can use just a couple of simple +5V and -5V supplies, you get a good result even with that.
something I was wondering... do the boards allow separate analogue supplies for each channel?
It could be done - but not easily.
Everything on the 1941 is so small you would have to be really good with SMD components that are really small.
I did it on my first generation board and have no intention of doing it with this one. There were other folks involved back then who showed the way. nige2000 stands out as the pioneer.
AS with everything - it is all compromise. Of course, having separate supplies would be good. The only problem is getting to that will most likely bring about other problems that will, at least, equal, the "badness" of shared supplies.
I sincerely believe that SOEKRIS has made many improvements with this board as compared to the first generation that make going to this trouble just not worth it and could, after all is said and done and there would be LOTS to do and the end result will fulfill BARANEK'S LAW - the one about nothing sounding as good as the loudspeaker you made for yourself.
After going through that misery your mind will refuse to hear any faults.
Everything is possible. The question is: is it worth it?
Everything on the 1941 is so small you would have to be really good with SMD components that are really small.
I did it on my first generation board and have no intention of doing it with this one. There were other folks involved back then who showed the way. nige2000 stands out as the pioneer.
AS with everything - it is all compromise. Of course, having separate supplies would be good. The only problem is getting to that will most likely bring about other problems that will, at least, equal, the "badness" of shared supplies.
I sincerely believe that SOEKRIS has made many improvements with this board as compared to the first generation that make going to this trouble just not worth it and could, after all is said and done and there would be LOTS to do and the end result will fulfill BARANEK'S LAW - the one about nothing sounding as good as the loudspeaker you made for yourself.
After going through that misery your mind will refuse to hear any faults.
Everything is possible. The question is: is it worth it?
I am sure SOEKRIS will gave that approach complete approval!
If one is dead set to have dual mono there is no question, if you value your time and mental health, this would be the least expensive way to do it.
So obvious, it did not even occur to me!
Is there a way to set gain for each board with one control?
Could one board have slightly different gain than the other making them impossible to match gains between both? Not that that would be impossible to live with ..
If one is dead set to have dual mono there is no question, if you value your time and mental health, this would be the least expensive way to do it.
So obvious, it did not even occur to me!
Is there a way to set gain for each board with one control?
Could one board have slightly different gain than the other making them impossible to match gains between both? Not that that would be impossible to live with ..
One dB can be large for balance but one has to just have a drink or two and it no longer matters!
Audio and alcohol must always be enjoyed in tandem.
It would be instructive to give this a try and hear if it makes any important difference.
The rest of my system is dual mono so the temptation is there. Maybe next year ...
Audio and alcohol must always be enjoyed in tandem.
It would be instructive to give this a try and hear if it makes any important difference.
The rest of my system is dual mono so the temptation is there. Maybe next year ...
Thanks for this gem, have always ignored them but they're great for the odd laugh.
" The output is simple: there is no audible difference. There just isn't to my ears. "
So what do you find funny/outrageous about this opinion?
Simply changing the filter changes the way the dam sounds, I can certainly tell the difference in each from the pack of 4 that I use, it couldn't possibly be the same sound as other DACs. Someone needs to measure music and compare the outputs for accuracy, not merely analyzing test tones.
"Gain" is only determined by Vref. I don't see any issues sharing a single Vref among several boards.
Vref set by 1% TL431 which mean <0.1dB signal level difference between boards. Of course typically better....
It's much faster and easier to just test using simple test tones, than to do comprehensive listening test....
Thanks for this gem, have always ignored them but they're great for the odd laugh.
" The output is simple: there is no audible difference. There just isn't to my ears. "
I still don't understand what's funny/outrageous about an opinion?
Not saying that you are wrong and "ASR" is right (or the other way around). But is there any grounds for taking one statement over the other? BTW, about "ASR", I know a little about his work and career, about you (and the analog guy with the tongue sticking out), absolutely nothing. Guess who, lacking any direct evidence, inspires more trust.
It is fascinating and more than a little ironic that so many people who choose to devote their time to audio measurements are completely lacking of even basic listening abilities.
But of course this does not make their pointlessly obsessive measurements any less pointless. The important thing is they inspire trust.
Why the sudden interest in this thread?
Cornix cornici oculos non effodiet
But of course this does not make their pointlessly obsessive measurements any less pointless. The important thing is they inspire trust.
Why the sudden interest in this thread?
Cornix cornici oculos non effodiet
It is fascinating and more than a little ironic that so many people who choose to devote their time to audio measurements are completely lacking of even basic listening abilities.
But of course this does not make their pointlessly obsessive measurements any less pointless. The important thing is they inspire trust.
Why the sudden interest in this thread?
Cornix cornici oculos non effodiet
Says who?
Otherwise, good question. As you were, carry on and don’t forget to keep that tongue hanging out. It helps a lot making a point.