Congrats abraxalito on this milestone !
So, only experienced DIYers can assemble this DAC or is it noob friendly (for someone like me) ?
So, only experienced DIYers can assemble this DAC or is it noob friendly (for someone like me) ?
Thanks for your kind words. I'd say you might find it challenging if you're a total noob, but then again neither do you need to be highly experienced.
Most of the parts are 0805 SMT - if you have no experience at all with surface mount then you'll need to equip yourself with a desk magnifier and a small-tipped iron (I use 2.3mm but that's on the large side I reckon) plus some narrow gauge solder (0.5mm typically). There aren't any fine pitch ICs, just normal small outline packages with 1.27mm pin spacings. The only leaded parts are the electrolytics as I have an aversion to SMT 'lytics.
Most of the parts are 0805 SMT - if you have no experience at all with surface mount then you'll need to equip yourself with a desk magnifier and a small-tipped iron (I use 2.3mm but that's on the large side I reckon) plus some narrow gauge solder (0.5mm typically). There aren't any fine pitch ICs, just normal small outline packages with 1.27mm pin spacings. The only leaded parts are the electrolytics as I have an aversion to SMT 'lytics.
Interesting !
I'm planning to use the other DAC (Chinese WM8740 DAC) I have for TV (SPDIF) and need a higher quality for PC and iPad/iPhone (USB input). Also it need to be compact.
I'm planning to use the other DAC (Chinese WM8740 DAC) I have for TV (SPDIF) and need a higher quality for PC and iPad/iPhone (USB input). Also it need to be compact.
Hi richard,
I've been reading all of your nos dac threads, had been really insightful! I was thinking of building one of your lingdac's but i don't have much time on hand atm. A kit for the phidac would be amazing, can i still join to buy 2? If i understand correctly the Phidac has the 3 stages in 1 (buffer, dac and low noise supply for single 8Vdc in)?
I have two more questions;
I want to use 24bit 96khz signal comming from a minidsp 2x4hd master. As far as i can see this combo should work fine but can I mod the output filter (NOS droop came by a lot of times in your threads) to the 96khz?
The input impedance of my amplifiers are 6k and 3k3, this shouldn't be a problem right? If I rememered correctly the output impedance would be low enough with a buffer stage and the 10x impedance rule.
Chris
I've been reading all of your nos dac threads, had been really insightful! I was thinking of building one of your lingdac's but i don't have much time on hand atm. A kit for the phidac would be amazing, can i still join to buy 2? If i understand correctly the Phidac has the 3 stages in 1 (buffer, dac and low noise supply for single 8Vdc in)?
I have two more questions;
I want to use 24bit 96khz signal comming from a minidsp 2x4hd master. As far as i can see this combo should work fine but can I mod the output filter (NOS droop came by a lot of times in your threads) to the 96khz?
The input impedance of my amplifiers are 6k and 3k3, this shouldn't be a problem right? If I rememered correctly the output impedance would be low enough with a buffer stage and the 10x impedance rule.
Chris
Last edited:
Hi Chris
I'm not selling fewer than 10 at a time myself but since there is some demand I've asked my girlfriend to help out and she's happy to deal with the smaller quantities. I'll PM you later with her email.
Yes, that's broadly right - except the PSU is 9VDC min, 12V max. Oh and the stages are dac, passive filter, I/V, active filter on PhiDAC. LingDAC has no active filter but a more complex passive one.
Glad you asked that question - I rather foresaw such a request and have included a spare capacitor site on the board to help accommodate different active filter characteristics. I've not done the redesign for 96k yet but it shouldn't take me more than an hour. As for 24bits the DAC's only going to recognize 16, is it possible to dither for that in the miniDSP?
Certainly not a problem for PhiDAC to drive such loads. The Zout is 15ohm (created by a series resistor).
Could I reduce the series resistor to lower Zout or short it ? I plan to make one DAC to drive my K812 ( 36 ohms ) so low Zout is very welcome in that case!
I'd say definitely do NOT short it out, its there for a purpose which is to decouple capacitive loads from the AD8017. They're only rated to drive something like 15pF without any series resistance.
What I've done to drive headphones is replace the 15ohm with a string of 3 ferrite beads, forming about 3uH of output inductance. This works fine - you'll get sub-1ohm Zout in the audio band then.
What I've done to drive headphones is replace the 15ohm with a string of 3 ferrite beads, forming about 3uH of output inductance. This works fine - you'll get sub-1ohm Zout in the audio band then.
Thanks for the answers!
I have a smps power supple (Meanwell EPS15-12, 80mV p-p ripple, act as standby power as well) for the miniDSP. Would it be benefitial to add a LDO like a LT3045 to step down to around 10Vdc in front of your board?
Great! My knowledge is very limited and asked this in a thread on the miniDSP forum; MiniDSP : Adding dac with i2s out pins 2x4HD and SRC options (1/1)
No reactions yet.
I suspect the output is always 24bit (no dither options). But without dither, the tda1387 should be able to take the first 16bits (ignoring the rest, truncating???) and still work it's wonders right? I don't know how much impact it would have on the sound:/
As far as I understand the last bits are most of the time empty or noise, since there is almost no use for all that dynamic range. (correct me if i'am wrong😕)
I'll be watching my inbox:3
I have a smps power supple (Meanwell EPS15-12, 80mV p-p ripple, act as standby power as well) for the miniDSP. Would it be benefitial to add a LDO like a LT3045 to step down to around 10Vdc in front of your board?
Great! My knowledge is very limited and asked this in a thread on the miniDSP forum; MiniDSP : Adding dac with i2s out pins 2x4HD and SRC options (1/1)
No reactions yet.
I suspect the output is always 24bit (no dither options). But without dither, the tda1387 should be able to take the first 16bits (ignoring the rest, truncating???) and still work it's wonders right? I don't know how much impact it would have on the sound:/
As far as I understand the last bits are most of the time empty or noise, since there is almost no use for all that dynamic range. (correct me if i'am wrong😕)
Last edited:
In the following thread they are discussing this since i'am using the dsp for my digital volume control;
Review and Measurements and miniDSP 2x4 HD DSP and DAC | Page 2 | Audio Science Review (ASR) Forum
Due to the volume control and eq most of the bits are empty when the volume is not at 100%?
Not sure if this is right...
Ling Dac Listening Test Review
Hi Folks,
Some possibly interesting info for those considering Ling dacs. Abraxalito (Richard) sent me a Ling dac to listen to and compare with the high end dacs I am working with. This particular Ling dac came outfitted with a couple of small output isolation transformers at the audio outputs which can be seen in the picture below.
How does it sound for a $5 dac? Pretty amazing at the price. First, it doesn't sound like anything like the modern sigma delta dacs. In fact, there is one respect in which it can beat them, which is audibility of low level reverb tails. They are very easy to hear and sound nice being there. By way of comparison, Benchmark DAC-3 is much worse in that respect, and Allo Katana (correctly powered) is somewhere in between. But, this dac wins for that.
At this point I should probably mention that the Ling dac is receiving I2S from a CM3361A board, and is the Windows default sound device set to 44.1kHz default to match the files I am playing. Clock quality would probably be better from an Italian Amanero board as compared to the CM3361A, and that might change the sound some.
In what respects are the high end sigma delta dacs better? Well, it depends a lot on the particular implementation. The cheap Chinese ES9038Q2M dacs are pretty distorted to my ears, at the low end of the cost spectrum. So are those that use wall warts and inadequately implemented SMPS. However, the high end Sabre dacs are far superior in every way except for reverb tail audibility. I am thinking of things like very smooth yet very detailed sound, stereo imaging, ability to accurately reproduce punchy percussion that sounds real, etc. Pretty much everything, except the reverb tails.
So, the question then seems to become how can one enjoy the best sound from Ling dac? The first thing I would say is don't run the digital volume control feeding Ling dac at full volume then control the playback volume with another volume control in the analog part of the system. At full digital volume Ling dac can get kind crunchy from intersample over clipping, at least it sounds like that type of thing. With the digital volume set a bit lower, maybe -4dB, that's probably enough to avoid problems.
The other thing I would say is that Ling sounds best to me out of the close near field of the speakers where some room sound starts to fill in, or off to the side a bit out of the way of direct HF. In that respect, headphones might not be the best choice for maximum enjoyment, but I didn't try it.
How expensive of commercial dacs can Ling dac compete with in terms of sound quality? I think it depends on what one likes and dislikes about different sounds. Real good dacs cost up around $2,000 and up, IMHO. All dacs below that level have more problems than just missing reverb tails. For example, I once tried a well known $350 dac with a good reputation and listened only a few seconds before noticing the sound was smeared and phasy, as compared to a really good dac. It went right back in the box and back to the store. Just happened to be something not right for me.
Here is what I would say, for a $5 kit, or for whatever someone might sell a pre-built one, Ling dac is definitely worth trying. I might well prefer it to a $50 or even $150 ES9038Q2M dac with a voltage output stage or wall wart powered with SMPS noise. Some might prefer Ling to even more expensive units. Again, it depends what one likes and dislikes given that there will be some shortcomings in all dacs.
EDIT: Just did a follow-up experiment with findings that are very good news. I added about 100uf of parallel film caps at the power terminals of Ling dac. Big difference! Much more smooth highs. Now we are talking about a dac that can compete with dacs up into hundreds of dollars, IMHO. The caps add about $50 to the cost of the dac, of course. And the transformers on this Ling dac would add a little more cost too, but I think people could be pretty happy with the add-on options. For me, the upgrades would be must-haves.
Hi Folks,
Some possibly interesting info for those considering Ling dacs. Abraxalito (Richard) sent me a Ling dac to listen to and compare with the high end dacs I am working with. This particular Ling dac came outfitted with a couple of small output isolation transformers at the audio outputs which can be seen in the picture below.
How does it sound for a $5 dac? Pretty amazing at the price. First, it doesn't sound like anything like the modern sigma delta dacs. In fact, there is one respect in which it can beat them, which is audibility of low level reverb tails. They are very easy to hear and sound nice being there. By way of comparison, Benchmark DAC-3 is much worse in that respect, and Allo Katana (correctly powered) is somewhere in between. But, this dac wins for that.
At this point I should probably mention that the Ling dac is receiving I2S from a CM3361A board, and is the Windows default sound device set to 44.1kHz default to match the files I am playing. Clock quality would probably be better from an Italian Amanero board as compared to the CM3361A, and that might change the sound some.
In what respects are the high end sigma delta dacs better? Well, it depends a lot on the particular implementation. The cheap Chinese ES9038Q2M dacs are pretty distorted to my ears, at the low end of the cost spectrum. So are those that use wall warts and inadequately implemented SMPS. However, the high end Sabre dacs are far superior in every way except for reverb tail audibility. I am thinking of things like very smooth yet very detailed sound, stereo imaging, ability to accurately reproduce punchy percussion that sounds real, etc. Pretty much everything, except the reverb tails.
So, the question then seems to become how can one enjoy the best sound from Ling dac? The first thing I would say is don't run the digital volume control feeding Ling dac at full volume then control the playback volume with another volume control in the analog part of the system. At full digital volume Ling dac can get kind crunchy from intersample over clipping, at least it sounds like that type of thing. With the digital volume set a bit lower, maybe -4dB, that's probably enough to avoid problems.
The other thing I would say is that Ling sounds best to me out of the close near field of the speakers where some room sound starts to fill in, or off to the side a bit out of the way of direct HF. In that respect, headphones might not be the best choice for maximum enjoyment, but I didn't try it.
How expensive of commercial dacs can Ling dac compete with in terms of sound quality? I think it depends on what one likes and dislikes about different sounds. Real good dacs cost up around $2,000 and up, IMHO. All dacs below that level have more problems than just missing reverb tails. For example, I once tried a well known $350 dac with a good reputation and listened only a few seconds before noticing the sound was smeared and phasy, as compared to a really good dac. It went right back in the box and back to the store. Just happened to be something not right for me.
Here is what I would say, for a $5 kit, or for whatever someone might sell a pre-built one, Ling dac is definitely worth trying. I might well prefer it to a $50 or even $150 ES9038Q2M dac with a voltage output stage or wall wart powered with SMPS noise. Some might prefer Ling to even more expensive units. Again, it depends what one likes and dislikes given that there will be some shortcomings in all dacs.
EDIT: Just did a follow-up experiment with findings that are very good news. I added about 100uf of parallel film caps at the power terminals of Ling dac. Big difference! Much more smooth highs. Now we are talking about a dac that can compete with dacs up into hundreds of dollars, IMHO. The caps add about $50 to the cost of the dac, of course. And the transformers on this Ling dac would add a little more cost too, but I think people could be pretty happy with the add-on options. For me, the upgrades would be must-haves.
Attachments
Last edited:
Thanks Mark for the listening impressions !
But, can you tell us more about mid and high frequency reproduction ?
Is it sharp and brittle or slightly dark and smooth sounding ?
How would you describe the frequency response ?
Thanks again ��
But, can you tell us more about mid and high frequency reproduction ?
Is it sharp and brittle or slightly dark and smooth sounding ?
How would you describe the frequency response ?
Thanks again ��
@eljoantonyn
I just added an edit at the bottom of the review describing film caps for improve power quality. With the film caps HF is smooth and distortion is low enough probably not to bother most people. I liked it, although it is still not in the 'really' good class.
Without the film caps, it is a bit brittle at HF (a bit more than I would prefer, personally), but smooth in the mid frequencies. Frequency response was not measured by me, but most dacs are ruler flat. If they sound bright or dark, or weak in the bass, it probably still looks flat in a frequency response measurement. What causes such sound quality variations is beyond the scope of the discussion for now, since it could take us rather off topic.
I just added an edit at the bottom of the review describing film caps for improve power quality. With the film caps HF is smooth and distortion is low enough probably not to bother most people. I liked it, although it is still not in the 'really' good class.
Without the film caps, it is a bit brittle at HF (a bit more than I would prefer, personally), but smooth in the mid frequencies. Frequency response was not measured by me, but most dacs are ruler flat. If they sound bright or dark, or weak in the bass, it probably still looks flat in a frequency response measurement. What causes such sound quality variations is beyond the scope of the discussion for now, since it could take us rather off topic.
I would be interrested in which audio transformers, caps and psu you used. Do you have pictures?:3
The transformers were supplied by Abraxalito, maybe he can say what they are.
The film caps were Wima mks4 soldered together in parallel using the following values: 10+10+22+33+33 (all values in uf). I have found this particular combination to work quite well in other cases too. However, adding as little at 15uf of small value film caps in parallel has been reported to make some easily audible immediate change for the better.
The film caps were Wima mks4 soldered together in parallel using the following values: 10+10+22+33+33 (all values in uf). I have found this particular combination to work quite well in other cases too. However, adding as little at 15uf of small value film caps in parallel has been reported to make some easily audible immediate change for the better.
Big thank you to Mark
Thanks very much indeed for such a long and detailed review of PhiDAC (by the way its PhiDAC, not lingDAC which I sent) - I appreciate the time spent and attention to detail in your writing about it. I will respond to individual points you raised in a later post, particularly I'm interested in the power supply issue.
On the transformers, I mentioned how they were built here : lingDAC - cost effective RBCD multibit DAC design
Thanks very much indeed for such a long and detailed review of PhiDAC (by the way its PhiDAC, not lingDAC which I sent) - I appreciate the time spent and attention to detail in your writing about it. I will respond to individual points you raised in a later post, particularly I'm interested in the power supply issue.
On the transformers, I mentioned how they were built here : lingDAC - cost effective RBCD multibit DAC design
Happy to give it a try. Have a short follow up, which is that later the today I went back for a 2nd listening session, this time accompanied by my son. I still notice a lot of distortion compared to the dacs I am working on, but my son pointed out something that I didn't quite make out because I was listening to the distortion and not listening with a more open mind. He said it sounded a bit phasey, like a wha wha pedal. When he pointed it out I could hear there was something that sounded a lot like it could be a bit of a hard peak in the frequency domain with some associated phase shift. Don't know if the transformers might have been interacting with the cable going to the preamp or what. Maybe I will have a chance to bypass them or try adding some resistive loading, just as kind of an experiment to see if any effect.
The transformers do have quite significant leakage inductance - they're designed that way to minimize the coupling of input to output. I made measurements of the 'leakage' and compensate for it with an output cap and zobel network (series RC). That said its not optimum for all load impedances - I tuned it for 47k. Go much lower than that and there will be some droop at higher frequencies but that's going to be less than 1dB I reckon going down to 22k.