Hello everybody,
I've recently bought a Topping E50 to be used in a second system, while keeping my old Musical Fidelity VDACii.
The E50 is very different, transparent, clean, wider... there's more of everything, except some musicality on mids that the VDACii has.
My question is:
I don't have the possibility to buy not even try DACs above 1000 euros, but I would like to understand: what makes them so expensive?
What is the plus thay can give compared to cheaper ones? Distortion numbers are impressively low even for those cheaper DACs.
Second part of the question is:
I would like to build a USB-only DAC, buying boards for the digital part, and then implement a tube preamp/buffer for it.
Everything in the same case. What would be your suggestions? What is the state of the art for the DIY? FPGAs?
Thanks in advance,
Kind Regards
Roberto
I've recently bought a Topping E50 to be used in a second system, while keeping my old Musical Fidelity VDACii.
The E50 is very different, transparent, clean, wider... there's more of everything, except some musicality on mids that the VDACii has.
My question is:
I don't have the possibility to buy not even try DACs above 1000 euros, but I would like to understand: what makes them so expensive?
What is the plus thay can give compared to cheaper ones? Distortion numbers are impressively low even for those cheaper DACs.
Second part of the question is:
I would like to build a USB-only DAC, buying boards for the digital part, and then implement a tube preamp/buffer for it.
Everything in the same case. What would be your suggestions? What is the state of the art for the DIY? FPGAs?
Thanks in advance,
Kind Regards
Roberto
@JMFahey very easy to explain.
I now use a Daphile system with DAC (I have only two of them) and class D amp in all three systems I usually use.
PP tube amps are used once in a while. SE amp is under construction, but some other works have higher priority.
The cheapest and fastest way to experiment on adding harmonics in my usual listening systems is a tube preamp.
I now use a Daphile system with DAC (I have only two of them) and class D amp in all three systems I usually use.
PP tube amps are used once in a while. SE amp is under construction, but some other works have higher priority.
The cheapest and fastest way to experiment on adding harmonics in my usual listening systems is a tube preamp.
@jan.didden
could be fashion audio as well, of course, but I'm used to build my own systems (both for instrument and hi-fi), so I would like to try this route as well.
could be fashion audio as well, of course, but I'm used to build my own systems (both for instrument and hi-fi), so I would like to try this route as well.
Sure, more power to you! But in part of your post you seemed to like low distortion and transparency.
That doesn't jive with a tube stage.
Jan
That doesn't jive with a tube stage.
Jan
It depends on the kind of music and the speaker I use (I have a pair of Klipsch RF82, a pair of FHXL with Alpair 11MS speakers and a pair of Silbury with Alpair 10.3): in some cases I liked most high DF low distortion amps, in others lower DF tube PP amps. The two self built speaker cabinets should indeed work with a DF around 2-2.5 instead of the low Zout of the Class D amps.
Then do add a tube stage which actually does something visible (on scope screen), audible, measurable, not just expectation bias fueled "do nothing" ones.
I design and make Guitar amps, and different to generic Audiophiles, Musicians, although also showing some expectation bias, a very Human trait, definitely have well tuned ears , overblown stuff disappears from the Market if not meeting expectations.
This is a gain stage I add to "clean" SS amps, think for Acoustic Guitars, organ, piano, etc.
Even Harmonics by the truckload he he.
Use only ECC189 (or PCC189 with a 7.5V filament supply, DC strongly recommended) and 120-150V +V supply.
Or design your own, all you need is:
For a roughly unity gain, VERY flavoured stage, drive it straight from your DAC and then attenuate output by 10-20X, say 100k-10k attenuator.
If too much of a good thing, remove cathode capacitor and adjust attenuator to get near unity gain.
You will be surprised.
Very often it is the high price that makes them expensive. And obviously the expensive casework.
What makes a good dac is a very different question, but the end result is similar: nothing good comes cheap.
The tubes in a dac is something I don't get though. Tubes excel at voltage amplification and the majority of current crop dacs need neither voltage nor current amplification. An active tube based preamp may still be desirable, but why integrate with the dac? It makes no sense and only creates problems.
You could expand your question beyond DACs, and to about every component in the playback chain, except, maybe, speakers. Take amplification, for example. There‘s no objectively rational reason for anyone to own anything more expensive than a mass-market A/V receiver. The objective specifications of such units routinely exceed the thresholds of human hearing acuity. Simply purchase the least expensive major brand unit which meets your power requirements. The high-end audio market, however, is defined by customers seeking one of two different primary benefits, one intellectual, and the other emotional. High-end audio equipment is a luxury goods market, and operates as one.
Intellectual benefit buyers are objective specifications focused. They seek the knowledge that a given component is, objectively, as close to perfect as possible. Even though those specifications are far beneath the threshold of human hearing acuity. Logically, and understandably, more perfect specification performance equals greater purity of signal transfer. Such buyers judge a component‘s value (price) relative to its specification performance versus alternatives. Delivering ever increasing objective perfection is an increasingly costly engineering and development activity. Which is, at least partly, reflected in price tags which can be orders of magnitude higher than that of A/V receivers, for example.
Emotional benefit buyers are subjective feelings focused. They seek components which give them a certain emotional experience, largely irrespective of that component’s objective specifications. Adding to that emotional experience are non-audio factors, such as the way that component appears to the eye, or status. Rising to audio jewelry status for the most expensive components. Such buyers judge value (price) relative to the human experience which that component provides them relative to alternatives. alternative can include non-audio options, such as exotic sports cars. Which, some audio systems, are priced like
Intellectual benefit buyers are also affected by the way a component appears, except that appearance seems judged in the service of the delivering the objective priorities. Such as, how much the component appears reliable, by way of its construction. My experience is that high-end intellectual buyers typically do not purchase based solely on its objective specifications, Else they would all have A/V receivers as the core of their home system. Just as many emotional benefit buyers like the notion of excellent specifications, so long as those specifications don’t seem to constrain their emotional listening experience. There usually aren’t fixed lines separating intellectual benefit buyers, from emotional benefit buyers. As humans, we are feeling beings that think. There is no clean separation of those two fundamental aspects of our decision making.
Intellectual benefit buyers are objective specifications focused. They seek the knowledge that a given component is, objectively, as close to perfect as possible. Even though those specifications are far beneath the threshold of human hearing acuity. Logically, and understandably, more perfect specification performance equals greater purity of signal transfer. Such buyers judge a component‘s value (price) relative to its specification performance versus alternatives. Delivering ever increasing objective perfection is an increasingly costly engineering and development activity. Which is, at least partly, reflected in price tags which can be orders of magnitude higher than that of A/V receivers, for example.
Emotional benefit buyers are subjective feelings focused. They seek components which give them a certain emotional experience, largely irrespective of that component’s objective specifications. Adding to that emotional experience are non-audio factors, such as the way that component appears to the eye, or status. Rising to audio jewelry status for the most expensive components. Such buyers judge value (price) relative to the human experience which that component provides them relative to alternatives. alternative can include non-audio options, such as exotic sports cars. Which, some audio systems, are priced like
Intellectual benefit buyers are also affected by the way a component appears, except that appearance seems judged in the service of the delivering the objective priorities. Such as, how much the component appears reliable, by way of its construction. My experience is that high-end intellectual buyers typically do not purchase based solely on its objective specifications, Else they would all have A/V receivers as the core of their home system. Just as many emotional benefit buyers like the notion of excellent specifications, so long as those specifications don’t seem to constrain their emotional listening experience. There usually aren’t fixed lines separating intellectual benefit buyers, from emotional benefit buyers. As humans, we are feeling beings that think. There is no clean separation of those two fundamental aspects of our decision making.
Last edited:
Unfortunately, that can be a rather misleading claim. Thresholds of hearing are not hard limits. Rather they are estimates of a limit for the average ear. It means that 50% of the population should be able to hear below the threshold and the other 50% of the population should not be able to hear below the threshold. In other words, thresholds are an estimate of the center of a bell curve, not an extreme point out on a tail. (Also, distributions were likely assumed to be gaussian, which we now know may be wrong in the case of some real world data. The assumption then may have resulted in significant errors in some published threshold studies: so-called outliers may be more common that previously belieived.)
Moreover, there are no published thresholds of audibility for some important things about hearing. For example, there is no published threshold of audibility for perception of soundstage, which is the perception of space with width and depth, and the perception of localization of sound source in that space. Yet a perceptual soundstage is exactly what stereo was intended to be able to reproduce. Otherwise only mono sound would suffice.
Changing the subject a bit, one thing to understand about the people here is that they are typically audiophiles or former audiophiles. Some former audiophiles take on a new set of beliefs to the effect that everything people can hear can be easily measured (which, for example, requires ignoring or rationalizing away the lack of specific measurements for soundstage, reproduction of the auditory cues humans use for spatial localization).
Another problem with so-called objective measurements is that they have rather poor predictive value. In science the true test of a theory is how good it is at predicting the future. However, if someone looks at the measurements of piece of audio gear, it is not possible to predict exactly how that device will sound. What measurements are best at is when the measurements are very poor, then sound is likely to be bad. Thus, its a good idea to listen to a device before buying it just because its measurement's look good. Manufacturers know that some people listen with their eyes, which is to say some people look at measurements rather than or before listening with their ears (thus creating an expectation bias as to the sound that will be heard). As a result, there may be some tendency on the part of some manufacturers to try to game popular measurements. That is, they try to make a low-cost, mediocre device the problems of which do not show up well in typical measurements. If it can beat another device in measurement rankings then hopefully a lot of them can be sold.
Changing the subject once again, it is true that people may shop not only for good sound, but also for good looks. People often like to have a beautiful looking car, or nice looking dining room set, a nice looking computer with some fun lighting options. Whatever it is for a given person, sometimes they prefer something that looks nice as well as serves some other primary function. If that's what they want and they can afford it, then there is not necessarily a problem with it. OTOH if someone can't afford something, then they shouldn't buy it. Just common sense,
Last edited:
I use tubes for the actual digital-to-analogue conversion in my DAC, mainly because I just felt like building something with tubes.
I have 2 TDA1541 dacs: AYA II and my own design with tube I/V buffer without global feedback. Guess which is my preference and actually has lower distortion 😉
Interesting.
Could we see a chassis picture or schematic?
How many tubes do you use?
I would expect 1 per bit per channel or so.
I use solid-state circuitry (an FPGA board and an SRC4392) to convert it to a noise-shaped single-bit format. Regarding the tubes:
85A2 voltage reference
ECC81 crystal oscillator
two times EF80 clock buffers
Six times E88CC, three per channel, for the core of the DAC
Seven neon lamps as part of the user interface
For more information, see:
https://www.diyaudio.com/community/threads/valve-dac-from-linear-audio-volume-13.308860/
https://linearaudio.net/sites/linearaudio.net/files/03 Didden LA V13 mvdg.pdf
https://linearaudio.net/sites/linearaudio.net/files/ThevalveDAC_PCB_FPGA_files for web posting.zip
https://linearaudio.net/sites/linearaudio.net/files/additionaldatavalveDACversion2p1.zip
https://www.diyaudio.com/community/threads/valve-dac-from-linear-audio-volume-13.308860/post-6569767
85A2 voltage reference
ECC81 crystal oscillator
two times EF80 clock buffers
Six times E88CC, three per channel, for the core of the DAC
Seven neon lamps as part of the user interface
For more information, see:
https://www.diyaudio.com/community/threads/valve-dac-from-linear-audio-volume-13.308860/
https://linearaudio.net/sites/linearaudio.net/files/03 Didden LA V13 mvdg.pdf
https://linearaudio.net/sites/linearaudio.net/files/ThevalveDAC_PCB_FPGA_files for web posting.zip
https://linearaudio.net/sites/linearaudio.net/files/additionaldatavalveDACversion2p1.zip
https://www.diyaudio.com/community/threads/valve-dac-from-linear-audio-volume-13.308860/post-6569767