Hi
In the past 2 weeks I have purchased online several different DACs, to try them and see the difference in the sound they give.
One thing I noticed, is that their circuit is always composed of 3 main chips:
1) USB to I2S chip
2) DAC chip, which takes I2S and converts it to Analog
3) An Op-Amp
My question:
If the definition of a DAC is that it outputs Line Out, (which you should then connect to your amplifier)
then why is the Op-Amp needed?
Thank you
In the past 2 weeks I have purchased online several different DACs, to try them and see the difference in the sound they give.
One thing I noticed, is that their circuit is always composed of 3 main chips:
1) USB to I2S chip
2) DAC chip, which takes I2S and converts it to Analog
3) An Op-Amp
My question:
If the definition of a DAC is that it outputs Line Out, (which you should then connect to your amplifier)
then why is the Op-Amp needed?
Thank you
Last edited:
Opamps are employed because that's a convenient way to perform a couple of functions at the output stage of a DAC chip - differential-SE conversion and filtering. Today's opamps have such good THD+N numbers that manufacturers' applications tend to use them exclusively so as not to compromise the attainable measurements from their chips.
There are of course other ways of implementing the diff-SE and filtering functions but none to my knowledge with THD+N numbers as low as those attainable from opamps.
There are of course other ways of implementing the diff-SE and filtering functions but none to my knowledge with THD+N numbers as low as those attainable from opamps.
The don't have to have an op-amp, they can have a discrete gain stage. Many DACs are current out, so you need an I/V stage, often implemented in an op-amp. Others have a voltage output, but you generally buffer that with a line driver of some sort.
There will be DACs for embedded applications that can drive line out directly
There will be DACs for embedded applications that can drive line out directly
Was the FDA-1 DAC created like that because it improves the sound over an Op-Amp?They don't "always" contain opamps. My FetAudio FDA-1 DAC has a discrete transistor IV and no opamps
It is possible to make a $2,000, state of the art, highly regarded and Stereophile recommended dac, which is used by many audiophiles and audio professionals, based largely on a lot of opamps. To wit, Benchmark Media DAC-3. It uses LME49860 opamps.
There is no reason why a carefully designed dac can't sound great, world class in fact, using opamps.
However, the dacs you have been trying are all essentially junk, and junk for a lot of different reasons. Opamps-or-not should be the least of concerns.
I understand that you say this not because of the DAC chips (which some of them are ESS, Burr-Brown, and more),
but because of the implementation, as we spoke about.
They come as a ready made PCB (with soldered parts), but without an enclosure..
And as you could see, they were really cheap, around 15$ on average.
What in your opinion is a (starting) price for a DAC that is not junk?
(and would you narrow the chip also to 1-2 companies?)
First of all, those were not the best dac chips from the manufacturers you cited. They were they bottom of the line low cost parts. They don't sound as good as top of the line parts can. You might want to take a look at the Stereophile recommended equipment list. A very accurate dac at the low price end of the really good sound quality range might cost $2,200.
--------------------------------
Not junk? That's tough. For me the minimum price for a dac I can be happy with that I know of right now is probably about $1,800. That's what it costs to make and sell one of sufficiently high quality.
I could tell you how to make your own pretty good dac for maybe around $200 (a lot of work, however). I don't know where you can buy a finished one that good for less than $1,000. Maybe when Allo Katana v1.2 is released there will be something in the $250 range, but can't say for sure til I hear the final version.
Unfortunately, making dacs that actually reproduce all the data on a CD accurately is still not easy or cheap. That's where we are right now with dacs. It would be nice if it were otherwise, but it just isn't.
Last edited:
Usually current out dacs with transformers attached use low value resistors to do the IV conversion. The transformers bridge the resistors and connect to an external amplifier of whatever type.
If we're talking costs of components and PCBs, transformer for PSU I'd say about $50. But then plenty of DIY time in building, pre-selecting components and testing. That's an estimate based on my lingDAC which is only a 16bit design. Aiming lower than -92dB (roughly the noise floor for RBCD) pushes up costs quite quickly.
From what I understood, it's because the ES9023 is a chip that has most of the surrounding circuitry inside it, built-in.Many ES9023 DACs don't have an external op amp - guess why
Is that the reason?
In the ES9018K2M's case, it's Stereo, 384KHz 32bit.
There are higher ones, like the ES9038Q2M (Stereo, 768KHz 32bit), and other versions of the ES9038 which are 8-Ch,
but as we spoke in that other thread, I do not need 8-Ch,
and from 192KHz to 384KHz or 768KHz, well, I think you need a good enough source file, in order to enjoy the higher sample rates..
Wow
I wish I could connect one like this here, and hear it for a week, to try and see If I notice a difference (and how much of it)..
And more importantly, to see if the difference justirifes moving from paying several tens of dollars, to several thousands of dollars.. on a DAC.
But trying something, that costs 1800$, is a bit problematic..
(for the case where it would not feel a big enough of a change)
OK hopefully it will come out
Haven't heard about it yet..
Nice..
And heavy..
Last edited:
Many ES9023 DACs don't have an external op amp - guess why
Sure, that's why you just bought an ES9038Q2M dac instead?
Not really. I upsample CD audio to 11.2MHz DSD format and find the dac clearly sounds better that way, IME.
Regarding trying an $1,800 dac, obviously you would need a power amp and speakers good enough to reproduce everything the dac is outputting. That costs more, unfortunately, unless perhaps you are willing to use some good use headphones.
Last edited:
The difference between the ES9018K2M's chip price and the ES9038Q2M's chip price is not what causes the expensive DACs to cost hundreed or thousands of dollars.Sure, that's why you just bought an ES9038Q2M dac instead?
Example:
Here you can see a ready made board with the ES9038Q2M, for 43$:
(Breeze Audio)
ES9038Q2M I2S DSD256 DAC HiFi Optical Coaxial Audio Decoder Assembled Board 384K 358962408632 | eBay
It's still far from 200$, 1800$, etc..
So even If you will really benefit from 768KHz 32bit, you can still buy the higher end chip, without going to a high end price..
The question is if the implementation is worth it, so much as to justify ~2000$ instead of 40$..
How do you upsample?
Software I assume?
BTW
Did you see this review?
Review and Measurements of JDSLabs ODAC Rev. B Compared to Fiio E10K | Audio Science Review (ASR) Forum
The winner from the 2 mentioned is none of them,
but the Topping D30, which appears to be not bad at all.
It's a 120$ DAC.
Topping D30 Multi-Function S/PDIF and USB DAC
What do you think about it?
Still junk?
Last edited:
It isn't. Maybe you are confusing a DAC chip with a domestic audio DAC box?spaceman5 said:
It isn't. That is, if you want unfiltered current output then you don't need an opamp. If you want a filtered voltage at low impedance then you need something, and a suitable opamp in a good circuit can do the job.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.