Topping D900 DAC Review

[WillBrink]

Shouldn't this be called Topping D900 discreet "1-bit" balanced Pre amp with smart DAC with PEQ? The pre amp part seems the key aspect of that thing no? A good measuring DAC are common thing these days, and short $, but as a reference quality pre amp and DAC, that's something to be interested in. Am I missing something in that? 10 band PEQ but no active room correction? Offer that with Dirac or another ARC and that would be a hell of product. Otherwise, have to connect MiniDSP to it? Other?

 

[338h10]

Sorry, you misunderstood me. Perhaps that's because I'm not a native English speaker and chose the wrong term.
I mean, a repair guarantee for a period of at least 10, ideally 20 years. Ideally with fixed rates or upper limits.

I don't want my device, which cost almost €2000, to become a paperweight after 3 or 5 years, or for repair costs to be so high that it's not worth it.
For example, Violectric/Lake People repairs even 30-year-old devices at absolutely fair prices.

A major problem with Topping, and other similar manufacturers, is that they give absolutely no thought to potential repairs during the development and planning of their devices. Otherwise, they wouldn't use extremely cheap and problematic very small SMD components in so many places.
This is unacceptable for me in a €500 device, even less so in a €1000 one, and in this price range, it's an absolute no-go.
Topping consistently demonstrates excellent development, but they completely fail in this discipline. As they say in sports, a perfect freestyle routine, but a failure in the compulsory exercises.

If you know chip design, you know that generally, the smaller the better, in terms of efficiency and reliability. Just think of longer signal path and larger power usage with the larger chips! Expensive DIPs vs modern SMDs that are efficiently produced at a much lower price? Scientifically (thank you, ASR), modern SMDs all day everyday. I had a very expensive Wadia DAC (>$10K 20 years ago) -- the DAC became obsolete in a few years such that no repair costs justified bringing it back alive.

It is also wildly unreasonable for any manufacturer to "guarantee" any "fixed" labor rate for any period beyond 2-3 years, given the inflation happenning everywhere. Of all the places, howeever, it is still virtually guaranteed that labor costs in/from China would still be less inflated, which is why I buy from there. They charge an honest price for what they produce, affordable or expensive.

 

[Herbert Shooler]

I have this DAC and its terrific,thanks Amir we are indeed spoilt these days,one wonders if these DACS are the final limit in performance,we have maxed out.

 

[DrCWO]

Nice review, waited for that since some time. Thanks' @amirm

As @amirm wrote "produce some of the best noise+distortion numbers I have measured:" I just compared the measurement of the low price Topping D50 III with the D900.
Nearly the same but the linearity of the D50 III is better at low levels. Also jitter performance is a bit better with the D50 III and take a look at the
ultrasonic output caused by noise shaping.

I know that none of this makes any difference when listening.

I'm just surprised by the marketing hype Topping is putting on this device. Sure, the D900 looks good and is correspondingly expensive. But in my opinion, it's in no way better than state of the art a chip DAC.

 

[DrCWO]

I'm not an expert but if 4 volts are the standard I'd bet 10 volts is not safe with power amps in general. Please experts, chime in

If you use a power amp for example NCx500 with only 11dB amplification this is absolutely fine.

 

[DrCWO]

The setup is Roon linked Zones for the source.

In couples zones Roon produces timing errors in all Zones except the master. I don't wander about sonic impact.

 

[DrCWO]

Otherwise, have to connect MiniDSP to it? Other?

Why not use Roon FIR for room correction?

 

[Roland68]

If you know chip design, you know that generally, the smaller the better, in terms of efficiency and reliability. Just think of longer signal path and larger power usage with the larger chips! Expensive DIPs vs modern SMDs that are efficiently produced at a much lower price? Scientifically (thank you, ASR), modern SMDs all day everyday. I had a very expensive Wadia DAC (>$10K 20 years ago) -- the DAC became obsolete in a few years such that no repair costs justified bringing it back alive.

It is also wildly unreasonable for any manufacturer to "guarantee" any "fixed" labor rate for any period beyond 2-3 years, given the inflation happenning everywhere. Of all the places, howeever, it is still virtually guaranteed that labor costs in/from China would still be less inflated, which is why I buy from there. They charge an honest price for what they produce, affordable or expensive.

If you were familiar with modern chip design, you'd know that not everything is better in SMD, that not everything can be implemented as well and without limitations in SMD, and that compromises often have to be made.

Furthermore, your post has absolutely nothing to do with the problem I was writing about.
The point is that the smallest and smallest packages of the cheapest SMD resistors and the cheapest SMD ceramic and film capacitors, cut from large sheets, are the cheapest and worst type of component available.

They not only fail more frequently, but many times more often. Even worse is the fact that these cheap SMD components don't just fail, as happens with high-quality SMD and through-hole components, but also usually cause a short circuit.
This short circuit often damages other circuits, and in the case of a short to ground, troubleshooting becomes incredibly time-consuming and uneconomical, even with a circuit diagram.

It's not about fixed hourly rates, but about realistic repair times, which wouldn't change. This is standard practice in the industry.
We ourselves have been supplying our own products to industry worldwide for over 60 years, and a repair estimate of 1.5 hours that we gave 30 years ago still applies today.

Repairs are also always related to sustainability.

 

[Cogito]

Great performance. Just $1,000 overpiced IMHO. So many great and inexpensive DACs out there.

 

[Roland68]

Great performance. Just $1,000 overpiced IMHO. So many great and inexpensive DACs out there.

And you're a good example of what I described in post #48, and how well the market's self-regulation works.

If you only care about good measurements, functionality and sound, and not everything else, then you can find excellent DACs for €79-110, even with balanced output and internal power supply, from €150 you can find something with balanced output, internal power supply and galvanic USB isolation, and for €330 you can find a DAC with analog volume control and output buffer.

The overall effort involved in the D900 is simply more expensive, more exclusive, and spread across much smaller production runs.

 

[SoySalsaPicante]

There's also the question of whether this version of their 1-bit converter is worth the difference over their standard one in the DX9 Discrete (which has exactly half the number of resistors), because that one is less money and includes an excellent headphone amp (plus an analog volume control, just like this D900).

 

[338h10]

If you were familiar with modern chip design, you'd know that not everything is better in SMD, that not everything can be implemented as well and without limitations in SMD, and that compromises often have to be made.

Furthermore, your post has absolutely nothing to do with the problem I was writing about.
The point is that the smallest and smallest packages of the cheapest SMD resistors and the cheapest SMD ceramic and film capacitors, cut from large sheets, are the cheapest and worst type of component available.

They not only fail more frequently, but many times more often. Even worse is the fact that these cheap SMD components don't just fail, as happens with high-quality SMD and through-hole components, but also usually cause a short circuit.
This short circuit often damages other circuits, and in the case of a short to ground, troubleshooting becomes incredibly time-consuming and uneconomical, even with a circuit diagram.

It's not about fixed hourly rates, but about realistic repair times, which wouldn't change. This is standard practice in the industry.
We ourselves have been supplying our own products to industry worldwide for over 60 years, and a repair estimate of 1.5 hours that we gave 30 years ago still applies today.

Repairs are also always related to sustainability.

You are playing and twisting words -- it's not "everything" as you said, but "generally" as I clearly indicated.

SMDs are better at heat management, board layout, etc., which are proven in modern laptops (a bigger challenges for chips than DACs/ADCs) -- today's laptops are far more reliable than the ones 20 years ago, removing the use of SSD from consideration (even though SSDs are full of SMDs themselves).

Would anyone here like a through-hole processor? Can a through-hole processor remotely accommodate the layout and other requirements needed for a 4nm processor? Are we to give up 4nm processors because no manufactuer can guarantee a fixed repair time for them (because absolutely no one will make such a guarantee)? I hope the answers are not yes, yes, and yes!

 

[WillBrink]

Why not use Roon FIR for room correction?

I don't know much about it, especially compared to others. Has anyone tested it?

 

[WillBrink]

Great performance. Just $1,000 overpiced IMHO. So many great and inexpensive DACs out there.

Many. How many reference level pre amps + DACs out there for that $?

 

[SoySalsaPicante]

I don't know much about it, especially compared to others. Has anyone tested it?

I haven't specifically tested it, but I imagine it's similar process for anything else with similar FIR filter capability where you use REW for measurement and filter calculation and then enter the filter data into Roon after you've created the filters using REW.

 

[SoySalsaPicante]

Many. How many reference level pre amps + DACs out there for that $?

Three good, "single-box," options right off the bat are Topping DX9, Topping DX9 Discrete, and Topping DX7 Pro+.

Then if you're willing to go two-box, just pick your choice of DAC and pair it with Topping A70 Pro, Topping L70, or Topping Pre90.

And that's strictly speaking only of Topping gear. World's your oyster if you include other brands. Luxsin X9 is great.

 

[WillBrink]

I haven't specifically tested it, but I imagine it's similar process for anything else with similar FIR filter capability where you use REW for measurement and filter calculation and then enter the filter data into Roon after you've created the filters using REW.

Dirac and ARC Genesis don't requite REW. You run the sweeps, they apply their own algorithms, and (hopefully) improve the room effects. You can use REW, and make adjustments manually too, personally let ARC genesis do it's thing automatically. I did follow a vid by Gene D on making some minor manual tweaks with ARC to enhance the base.

 

[Thighsofsteel]

I would get the D70 and pocket the £1200, to spend on beer and debauchery

 

[SoySalsaPicante]

Dirac and ARC Genesis don't requite REW. You run the sweeps, they apply their own algorithms, and (hopefully) improve the room effects. You can use REW, and make adjustments manually too, personally let ARC genesis do it's thing automatically. I did follow a vid by Gene D on making some minor manual tweaks with ARC to enhance the base.

Correct--but I don't know of any DACs that have DIRAC, except for miniDSP Flex with analog outputs, and those don't have analog volume controls. You'd have to combine a miniDSP Flex with digital output with one of these DACs that have analog volume control like the ones I listed in order to have such a combination.

 

[SoySalsaPicante]

I would get the D70 and pocket the £1200, to spend on beer and debauchery

Might as well save yourself even more with an E50 II at that rate.