• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

Hardware Teardown of Topping D30 DAC

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,369
Likes
234,395
Location
Seattle Area
This is a hardware teardown and analysis of Topping D30 USB DAC. No doubt many of you have read my review of the D30 and know that it is my current favorite for a budget DAC ($129). There seems to be two revisions of this DAC out there so this will outline the unit I tested.

Four hex screws releases the back panel allowing the board to easily slide out without much adventure:

Topping D30 USB DAC Top Board Teardown.jpg


The board is clean with no modifications, soldering issues or any problems visible. No, you don't want to "eat of it" but for a value-oriented product, I see nothing wrong with it. Here is the revision of the board by the way:

Topping D30 USB DAC Revision.png


This larger than normal budget DAC board sizes affords very good separation between the digital and analog side. Connectors are nicely spaced out which should reduce crosstalk. On top of that, there is a nice "moat" between the two sides as seen here:

upload_2018-1-10_13-46-52.png


That goes through the entire board as I have roughly shown in the overview picture above. Certainly a nice design practice to keep digital noise from bleeding into analog section (one of the biggest challenges in DACs). The only miss on this front is a front panel switch that selects between digital sources. So signals have to be routed way up there and back. This switch would have been best put in the back. I guess convenience trumped that engineering fidelity.

Back to overall picture, the ubiquitous Xmos async interface chip is used to fetch digital data from USB and convert that to raw PCM data to play by the Cirrus Logic CS4398 DAC to play. Companion CS8416 is used to interface with the S/PDIF coax and optical inputs.

Topping D30 USB DAC XMOS USB Interface Teardown.jpg


Audio-rated Burr Brown/TI Opa2134 low-noise operational amplifier is used to drive the analog outputs.

Topping D30 USB DAC BurBrown OPA2134 Opamp Teardown.jpg


Was pleasantly surprised by use of high-end brand capacitors from likes of Nichicon/Elna/Wima.

In similar vein, use of Omron Japanese relays (but made in China in this instance) is also reassuring.

That theme continued in an impressive manner with tons, and I mean tons, of test point for proper automated testing of the units:

Topping D30 USB DAC Test Points.jpg


Topping D30 USB DAC Test Points 2.jpg


Not that you would want to repair this type of product but if you did, these would come in very handy. Kudos on this! I don't see this much attention to testing in even higher-end products.

This gets us to the backside of the PCB:

Topping D30 USB DAC Bottom Board Teardown.jpg


Nothing too exciting really. The connectors in the back are hand soldered and there is some flux residue there. No solder balls or any other disasters as I for example found on the Schiit Modi 2 Teardown.

Notice how all the test points are brought to the bottom side through gold plated "vias" allowing a bed of nails type of jig to probe them for pass fail testing.

Unlike the rust used parts elsewhere, I was also happy to see some of the beefiest switches on a small product like this:
upload_2018-1-10_14-1-49.png


You could probably pull Sal's car with those heavy duty lugs!!! :D These are not going anywhere.

Only minor nit is the "hot snot" used to stabilize the board. This only works once of course. Now that I have pulled it out, it won't remelt itself. Still, it was good to see it there as it gets rid of the looseness and cheep feeling other products without this can have.

Summary
The Topping D30, just like its measured performance, is quite competently assembled. No revisions or flaws are apparent. Isolation is very good between digital and analog. No sign of used or counterfeit parts. Excellent attention is given to testability which translates into a unit working when you purchase it.

My recommendation to purchase the Topping D30 USB DAC continues! Good job Topping.
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,369
Likes
234,395
Location
Seattle Area

Don Hills

Addicted to Fun and Learning
Joined
Mar 1, 2016
Messages
708
Likes
464
Location
Wellington, New Zealand
The flux residue may be present deliberately. "No clean" flux is common these days. (Search "no clean flux".)
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,369
Likes
234,395
Location
Seattle Area
The flux residue may be present deliberately. "No clean" flux is common these days. (Search "no clean flux".)
Yeh, I assume they are no clean flux. It is a very thin layer.
 

Wombat

Master Contributor
Joined
Nov 5, 2017
Messages
6,722
Likes
6,459
Location
Australia
Just wondering if the PS is supplied by Topping or the vendor?
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,369
Likes
234,395
Location
Seattle Area

neu01

Member
Joined
Nov 2, 2017
Messages
11
Likes
6
IMG_20180111_233220.jpg IMG_20180111_233417.jpg

Thanks again Amir.

Here are pics of my unit, for what it's worth.
 
OP
amirm

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
44,369
Likes
234,395
Location
Seattle Area
Those $0.20 clocks are stupidly far away from the DAC chip.
Just saying'... o_O
I don't know their architecture but I assume the "clock" is generated by the XMOS SoC:

upload_2018-1-12_5-45-15.png


USB comes in from left, and clocked digital out comes from the other end. There is a MUX (chip) that selects between this path and S/PDIF input. The master clock source I see is right next to XMOS as they should be in this architecture. And you want this XMOS away from DAC (chip)'s analog stage so that bit of distance doesn't bother me.

upload_2018-1-12_5-50-22.png


Do you see it differently?
 

Eric Auer

Member
Joined
May 18, 2017
Messages
43
Likes
24

Superdad

Active Member
Audio Company
Joined
Aug 6, 2017
Messages
149
Likes
91
Do you see it differently?

Absolutely. The audio clocks 22.5792MHz/24.576MHz (of multiples thereof) should ALWAYS be as close to the DAC chip as possible. It's fine to also feed the clock back to the XMOS, but the DAC itself runs with (ideally) very nearby XOs. Look at almost any top quality DAC and you will see the proper placement of the clocks. The only clock that should be near the XMOS chip is the 24.0MHz one that it runs from.
 

sssn

Member
Joined
Nov 1, 2017
Messages
92
Likes
74
The jitter performance of the D30 is pretty good, so I wouldn't be too concerned about the clock being too far away.
 

Superdad

Active Member
Audio Company
Joined
Aug 6, 2017
Messages
149
Likes
91
The jitter performance of the D30 is pretty good, so I wouldn't be too concerned about the clock being too far away.

Well it's a cheap an cheerful DAC with a BoM of less than $30, so believe me I was not losing sleep over it. Since Amir commenting on design and parts I figured I'd point out that the layout could easily have been improved. :cool:
 

Jinjuku

Major Contributor
Forum Donor
Joined
Feb 28, 2016
Messages
1,278
Likes
1,180
Absolutely. The audio clocks 22.5792MHz/24.576MHz (of multiples thereof) should ALWAYS be as close to the DAC chip as possible. It's fine to also feed the clock back to the XMOS, but the DAC itself runs with (ideally) very nearby XOs. Look at almost any top quality DAC and you will see the proper placement of the clocks. The only clock that should be near the XMOS chip is the 24.0MHz one that it runs from.

Interesting as I keep seeing a 3rd party multiple clock board for all the different clocks on a motherboard with really long leads and audiophiles just crowing about the improvements.
 

Superdad

Active Member
Audio Company
Joined
Aug 6, 2017
Messages
149
Likes
91
Interesting as I keep seeing a 3rd party multiple clock board for all the different clocks on a motherboard with really long leads and audiophiles just crowing about the improvements.

What part of audio DAC sample clocks (the most important of all, the ones that directly affect jitter performance) needing to be close to the DAC chip's clock pin is not clear? The topic is a DAC.
 

Jinjuku

Major Contributor
Forum Donor
Joined
Feb 28, 2016
Messages
1,278
Likes
1,180
What part of audio DAC sample clocks (the most important of all, the ones that directly affect jitter performance) needing to be close to the DAC chip's clock pin is not clear? The topic is a DAC.

That's clear and it's sage advice of all clock implementations. I keep seeing modified: Motherboards, PCIe USB, NIC's, Routers, Switches with a more precise clock but with long lead wires.
 

DonH56

Master Contributor
Technical Expert
Forum Donor
Joined
Mar 15, 2016
Messages
7,835
Likes
16,497
Location
Monument, CO
I can think of good technical reasons for it to be close or far away... In an ideal system it does not matter how far the clock is from the load if the transmission line is perfect. In the real world, the longer the clock traces, the greater the chance for noise injection into the trace and the greater the loss. However, loss at such low rates is not a concern, and if placing the clock source/driver close to the DAC compromises the layout by e.g. reducing ground and power isolation or by providing a higher coupling factor into the signal, then it is better to keep the clock source further away. In my world it is nice to have the clock reasonably close to the DAC but rarely is it advantageous for it to be very close; too many other issues, and making clean clock traces is usually not too hard. Lots of variables drive component placement on a board (and on a chip).

FWIWFM, YMMV, etc. - Don
 

March Audio

Master Contributor
Audio Company
Joined
Mar 1, 2016
Messages
6,378
Likes
9,317
Location
Albany Western Australia
As always its about implementation and the effects of any actual real world design rather than audiophile dogma. The performance of clock transmission lines can be analysed and optomised. This is nothing unusual rf work.

Whack a femto clock in there, that will solve yor problems ;) any audiophile knows this.
 
Last edited:
Top Bottom