Topping DX3 Pro

My final takes after 17 months of using this product: https://www.superbestaudiofriends.org/index.php?threads/topping-dx3-pro.7498/page-4#post-309031



http://en.tpdz.net/products_detail/productId=39.html

So, this is my new toy to play with from now on (received this morning). This will mainly serve my main measuring dac-amp and bluetooth playback (for iOS/Android gaming or podcast, live news streaming, and so forth).

Regardless of well-documented huge hates toward this company (and that A-blahblah guy), this particular product attracted me in some aspects.


1. This is akm topping, not cirrus or ess-sabre topping (now the company name makes sense, doesn't it? lol).

Like somebody said before (couldn't remember exactly), it's not very likely to fail with akm chip. Their manual and application guide is so (too?) specific.

And dx3 reported to use akm4493eq and opa1612, my bet is topping closely followed reference lpf circuit parameters. Capable engineers such as Schiit or Cranesong may choose different ways though. Anyway I speculate the implementation would be what AKM engineers believe their chip to sound like. Not that bad.


2. Low output impedance

I kinda like topping over audio-gd, because at least they don't lie in spec.

Many friends might remember S19 r2r dac. Look at what agd claimed AND sbaf actually measured:

• http://www.audio-gd.com/Master/Singularity19/Singularity 19EN_Specs.htm
• https://www.superbestaudiofriends.org/index.php?threads/audio-gd-s19-technical-measurements.4305/

(This was enough for me to completely discredit any specs agd claims.)

Topping was relatively honest. For DX7S (their top of the line dacamp), they reported Zout as 10 ohm for unbalanced outputs and 20 ohm for balanced ones (shockingly high). http://en.tpdz.net/products_detail/productId=32.html

This time around, they claim dx3 has <1 ohm Zout. As a conservative guess, I think dx3 zout is around 1 ohm. Low enough [1].


3. Bluetooth input

I secretly enjoy lots of smartphone games, and bluetooth-connectable dac would be very cool (to me).


Extra. Black

Like the design. It's sleek and cute. For a tech toy for computer uses nothing wrong in design to my liking.


So, how does it sound? Well, so far (8-9 hrs of listening) I kinda like it, at least as a dac. Better than any previous topping dacs I experienced before (d10, d30, d50, etc). Also outperforms cheaper agds (r2r blahblah line). And even sounds a bit better than jot 4490 (pre-lyr3 era) if I recall correctly (as a dac not a combination) and fulla 2 (no idea about which revision; purchased in Jan 2018). In particular, for speaker uses, it performs more decently than initially expected.

But in short, for this budget, buy modi 3 [2] if you're an audio purist.

More sonic impressions to come.


Footnotes
[1] Of course this doesn't mean their head amp would be wonderful -- and indeed I confirmed they're not. At best, my expectation was "not mega-shitty". Probably sufficient for a measuring amp.
[2] It's a BIG shame that Modi 3 has lost GS 2019 award imho. Convert-2 (winner) is absolutely awesome. No doubt. But it's >$2k. It deserves to its asking price. But modi 3 does way, way, way more than what it asks to pay.

 

I think topping is fine for convenience units.

My cousin need a small cheap class-dead amp to power cheap outdoor speakers for the patio, they rent a small room on Airbnb.

Got them a real cheap topping all-in-one, looks like a cheaper version of this, no display but has (scratchy misaligned) treble/bass knobs.

It's fine, it's tiny, it makes sound, it plays from Bluetooth. I'm satisfied and so are they, and their guests.

 

I use a topping class D amp whenever I feel guilty about running a Class A amp to watch TV shows <_<

 

Early impressions (dac part)

I've been using or used two noteworthy 449x implementations: ADI-2 DAC and Modi 3 (both ak4490). Will use these two as reference points to explain what DX3 is capable of and how well. SD sharp filter used for both ADI2 and DX3.

Speaking of the dac performance of DX3 (bw2 used as amp), the most impressive point was staging.

Evaluating with Lore OG floorstanders, rendered stage is large, articulate, and with a good sense of 3D. Modi 3 did great on this but dx3 is a tad better. Very good distance (listening seat to virtual images). Likely outperforms ADI2 -- but can't say this strongly because speaker placements changed recently and external subwoofer connected no longer.

Using headphones, this magic isn't as evident as with speakers. Modi 3 sounds better with Aeolus suede. With Arya, Modi 3 and DX3 are nearly on par. Guessing it's somewhat related to physical openness (possibly associated with air compliance and acoustic impedance).

All the three were great on imaging. Hard to decide a clear winner.

As reported by many, Modi 3 addressed 4490's bass problem very well. Unnatural softness and mellowness (or bloomness depending on definitions) were mostly taken out with Modi 3 compared to ADI2. DX3 interestingly sits in between. It seems that Schiit put their output stage know-how to get things right while AKM fundamentally upgraded their converter performance in 4493. Meaningful change but not as dramatic as modi 3's accomplishment. Of course anything mentioned falls greatly behind of Convert-2 in bass slam and articulation.

Regarding transience and attacks, I like dx3 better than the others. Slightly edges out modi 3 in delivering right timing and strength of attacks. More in the direction toward convert-2.

As I recall, ADI2 was relatively stronger in microdetails and blackground, compared to modi 3 and by extension to dx3 as well. But had a noticeable veil -- I am not talking about nitpicking slight veil of yggdrasil a2 -- which was somewhat remedied in modi 3. DX3 is similar but feels a bit more shaded than the other two. With speakers to less extent though.

Anyway so far, DX3 remarkably favors Lore speakers and Arya. Scratching the head.
Also, it feels that some fundamental improvements were made in 4493 over 4490.

Intentionally restrained from making any assessments on dx3 builtin amp because bw2 is in way too different league. Will compare with budget or value amps to collect more thoughts.

By the way, attached is my current speakers setup. Picture taken at the listening position.
(Between-speaker distance = 9.5 ft, speaker-listener distance = 7.5~8 ft, zero toe-in, normally enjoy at 75dbC SPL)

 

Day 2

• Builtin amp wasn't magic. No surprise. Think of a slightly lesser version of rme adi-2 hp out if you're familiar with. Slightly muddier and rougher. Background is quite black (nearly on par with rme) for this price level though. For $200, I am 90% sure that modi 3 + magni 3 is a sonically wiser way to spend similar money.
• Good thing was that I could get very clean (non-clipped and no weird artifacts) responses with low-Z planars even at 100 dbSPL -- my original purpose as a measuring dac-amp replacing fulla 2.
• I like dx3 speaker-staging more and more as I get used to its rendering. Seriously considering to replace modi 3 with dx3, because I am a headphone-first guy, and speakers are generally for less-serious uses. I already have the world-best serving dac (convert-2) for serious listening, so modi 3 is having hard time in justifying itself.
• Convenience-wise dx3 spanks modi 3. It can bluetooth -- and as expected it raised my well-being to another level -- and it can remote.
• No Eitr magic. Either dx3 already has decent usb or it isn't resolving enough to identify usb craps. May be combination?
• Strangely, dx3 annoyingly responses to small changes of coax spdif input. Even very softly touching coax cable resulted in "clicks" (likely related to mute relays). This wasn't the case for modi 3.

 

After spending a week with dx3, I finally decided not to return it (although doing so and re-ordering via massdrop may save $10 or so haha).

Here are some notes after a few more comparisons:

• My earlier assessment on builtin hp out was a bit unfair because I figured out "lower" volume (by roughly 1 db) with dx3 in comparative listening. Comparisons described below were done with matched volume by minidsp ears.
• Magni 3 is overall a better amp (compared dx3+magni with dx3 only), particularly when paired with senn dynamics (modded 800 and 600), in terms of better timbre and more appropriate amount of good warmth. But I found dx3 hp out has a better synergy with my current-hungry arya. Arya with magni 3 sounds a bit too edged on high mids. By comparison, dx3 hp out and arya combinedly produce more rounded and less sibilant sound. Not at the level of BW but tonal direction is similar to my ears.
• Dx3 is actually a quite good AIO imho. Judging with Arya and Ether 2, superior to similarly priced competitors including o2+odac, lcx+sdac, and d50. Only two-boxed modi3+magni3 stack wins over dx3 pro. Schiit stack with eitr has a tighter and more nuanced bass while dx3 as an aio (wo eitr) has a tad less lazy transience. Of course everything mentioned is suboptimal compared to convert-2 and bw combo, as they should for price-delta.
• Think this as a junior RME. I believe it is a good way to put. As AIO rme adi2 and tp dx3 share a lot (both strengths and weaknesses) in common. DX3 may lose some details and blackness but gain better time behavior and less digital nastiness.
• Once get familiar, orange oled front screen is really beautiful. Truly so when looking at it at night.

 

so in your humble opinion rme adi 2 dac has nasty digital artifacts?

 

Yes. I've seen nearly all the opamp-based head amps (particularly built in amps) had the same problem more or less. I also believe this problem can be remedied or aggravated by the followings (no science only speculation):

• opamp used
• how local/global feedback and/or servo are designed

 

Thanks. I gotta try this. I Also want to be able to stream Bluetooth, mostly so that my and my wife's friends can play their awful hip hop when we are Hanging out and drinking. They all think that my music is old man music and is a buzz kill. turns out I don't have enough drake and Kendrick Lamar in my play lists.

I've been seeing some reports of hardware failures but it's too damn cheap to pass., so I Signed up for a silver one on massdrop.

 

I was interested in this too mainly due to the included remote control, but there are endless problem reports, with many units flat-out dying. Ironic how this enthusiastic recommendation blew in Amir's face over at his forum.

 

Is there any specific place to look at problem or failure reports? (I don't regularly read headfi, asr or reddit)

I am feeling I was just lucky to avoid such failures lol.

 

Massdrop product chat and the thread over at AmirScience Review.

 

That's very disappointing. I was going to buy one of these for S's&G's but I will definitely not be doing that knowing there seems to be high risk of failure.

 

After looking at quite a few reports, now I am thinking

1. Amazon return policy is wonderful.. at least it would protect me from any failures up to 30 days
2. Can't recommend this comfortably when returning is restricted like massdrop (afaik they only offer returns for doa items)
3. If risk-averse, just go for modi-magni combo and forget bluetooth/remotes

Btw, I didn't upgrade firmware. Kinda think failures seem more likely with upgraded firmware.

 

I like how you did due diligence here while Amir refuses to do so, there's not even a warning in small print in his original post.

 

I had this thing. It's a pretty good AIO imho agreeing with Vtory. I suppose you can see it as a mini RME, it has a good enough implementation of the 4493 chip, good enough USB+driver and good enough headphone amp. On budget you can do a lot worse..

The main drawbacks to me are
-stock power supply holding it back, and might be related to the random defect DX3s
-definitely buy somewhere with a good warranty/return policy

I personally sold mine for what I bought it for during single's day so no money lost. I sold it not because it didn't perform well, it actually performed better than I expected. I sold it because like most D-S Dacs, I just don't like chip sound, period. I only listen to R2R and/or NOS DACs nowadays..

 

Official update on the issue:
https://www.audiosciencereview.com/...c-and-headphone-amp.4967/page-142#post-150957

 

It's a bit weird that TP did announce it on ASR instead of their official web page. Of course it's not only TP. I've already seen some audio/headphone companies do the similar things on head-fi. Personally, I believe all such info MUST be on the official website first for various reasons (informational, ethical, and so forth)

Anyway, let me summarize TP's announcements to prevent unnecessary exposure to asr. lol

---
Problem: Fail to recover from the stand-by mode.

Possible reasons: something might go wrong in manually fixing surface-mount errors for op-amps (opa2140) so that short circuit activated power-protection.

TP action: improving soldering process AND adding 100hr tests

Consumer action
1. Contact [email protected] if met the problem. TP will pay for shipping.
2. If no problem after using for a while, likely the unit might be non-defective one.