Topping DX5II Balanced DAC and Headphone Amp Review

[ngs428]

What is this? Is QC really that bad?

https://www.reddit.com/r/headphones/s/YNUt1OWPdy

Just saw that post there too. This one explanation of the issue seemed to be the most clear…

“It's the design of the 4.4mm jack combined with poor output control on the amp. The ground pin on the 4.4mm plug is at the base, so the ground connection on the amp's jack is the outermost one. Thus as you insert the plug, each pin on the plug will make contact with the ground pin in the jack. If another pin on the plug is touching one of the signal output pins, then you're effectively grounding that output. Since the 4.4mm allows for bridged amping (called balanced in the headphone world but that's inaccurate), you're effectively grounding a bridged output which you must never do. That effectively short-circuits the amp and can release the magic smoke amongst other things.

Amps should have some sort of relay that only allows output when the plug is fully inserted to prevent this. Some amps like the ones in the Sony Walkmans simply don't connect the ground pin (it's not needed for bridged amping). And in all cases, as a precaution, inserting and removing a 4.4mm plug should always be a single swift motion so that the plug pins are never touching the wrong jack pins for long.”

So how do we avoid this? Use the XLR balanced and/or only plug in headphones when the switch on the back is off. Mine has been working fine for several months. I mainly use the XLR though.

 

[audiofun]

My understanding from the previous post is that it's an as yet released feature.

Do you have source for why you only believe it will be one intensity?

you can refer to the dx9 discrete manual.

 

[Timbo2]

you can refer to the dx9 discrete manual.

Convolution、Sample、Off (Default)Note: This is valid for all input and output interfaces (44.1kHz-48kHz/24bit-32bit)

I wonder what in properly translated English the difference between "convolution" and "sample" is.

 

[Mutex]

“It's the design of the 4.4mm jack combined with poor output control on the amp.

So the issue is only on the 4.4mm jack? Is there a mod one can perform? Like lifting the ground of the 4.4mm jack maybe?

I only use the 6.35mm single-ended and have plugged and unplugged it several times since I got it. Had no issue yet. XLR should be fine as well then?

 

[audiofun]

Convolution、Sample、Off (Default)Note: This is valid for all input and output interfaces (44.1kHz-48kHz/24bit-32bit)

I wonder what in properly translated English the difference between "convolution" and "sample" is.

convolution is using fft. sample is using EQ (bs2b algorithm). this matches pretty well with their product flyer.
since "sample" is just EQ, limiting it to 48khz is not necessary. also with bs2b parameters are adjustable. so adding flexible intensity is also trivial.
that's why I said the implementation looks like a toy. they can do much better at zero cost.

 

[Guetta]

Just saw that post there too. This one explanation of the issue seemed to be the most clear…

“It's the design of the 4.4mm jack combined with poor output control on the amp. The ground pin on the 4.4mm plug is at the base, so the ground connection on the amp's jack is the outermost one. Thus as you insert the plug, each pin on the plug will make contact with the ground pin in the jack. If another pin on the plug is touching one of the signal output pins, then you're effectively grounding that output. Since the 4.4mm allows for bridged amping (called balanced in the headphone world but that's inaccurate), you're effectively grounding a bridged output which you must never do. That effectively short-circuits the amp and can release the magic smoke amongst other things.

Amps should have some sort of relay that only allows output when the plug is fully inserted to prevent this. Some amps like the ones in the Sony Walkmans simply don't connect the ground pin (it's not needed for bridged amping). And in all cases, as a precaution, inserting and removing a 4.4mm plug should always be a single swift motion so that the plug pins are never touching the wrong jack pins for long.”

So how do we avoid this? Use the XLR balanced and/or only plug in headphones when the switch on the back is off. Mine has been working fine for several months. I mainly use the XLR though.

So the main issue is if 4.4mm is used? If someone uses 6.35 mm (1/4") stereo jack plug, the AMP would work just fine? Are all reliability issues primarily about this case you mentioned?

 

[ngs428]

So the main issue is if 4.4mm is used? If someone uses 6.35 mm (1/4") stereo jack plug, the AMP would work just fine? Are all reliability issues primarily about this case you mentioned?

I really don’t know any a more than what was at that Reddit post or posted here. I personally have not had any issues. Knock on wood.. but yes, the 4.4mm was reported on Reddit as the issue per what I posted.

 

[ngs428]

So the issue is only on the 4.4mm jack? Is there a mod one can perform? Like lifting the ground of the 4.4mm jack maybe?

I only use the 6.35mm single-ended and have plugged and unplugged it several times since I got it. Had no issue yet. XLR should be fine as well then?

I really don’t know any a more than what was at that Reddit post or posted here. I personally have not had any issues. Knock on wood.. but yes, the 4.4mm was reported on Reddit as the issue per what I posted.

 

[ziggurcat]

So the issue is only on the 4.4mm jack? Is there a mod one can perform? Like lifting the ground of the 4.4mm jack maybe?

I only use the 6.35mm single-ended and have plugged and unplugged it several times since I got it. Had no issue yet. XLR should be fine as well then?

I have unplugged/plugged in the 4.4mm countless times since I've gotten mine with no issue.

 

[ngs428]

I have unplugged/plugged in the 4.4mm countless times since I've gotten mine with no issue.

Same here, so not sure where the issue is. I’m not ruling out plants from other manufacturers trying to knock down the DX5II.

 

[ziggurcat]

Same here, so not sure where the issue is. I’m not ruling out plants from other manufacturers trying to knock down the DX5II.

I wouldn't go that far, but I feel like it's the kind of thing that would have been discovered a long time ago if it was an inherent flaw in the device.

I have also unplugged/plugged the 6.35mm jack in countless times even while the 4.4mm cable is connected (and the DX5 II is on, too) with no problems. Reports of the 4.4mm being loose is a bit suspect to me, though -- I found that you have to push it in a little bit further than what you think is the proper seating. I discovered that after I connected the 4.4mm cable for the first time, and no sound came out of one side. I's always bee a very snug fit with no wiggle at all for me.

The only issue I'm half-facing right now is that sometimes it will not come out of standby mode if it's been sitting for a while. I haven't quite figured out the specific steps to reproduce (or narrowed down the amount time is required while sitting in standby mode), but it tends to happen with the USB input only (i.e. USB is the only active input checked in the Input Option menu). I have not encountered it when using the OPT input only so far, and it (I think) has happened once or twice when both USB and OPT were active in the Input Option menu.

It's also not a 100% issue, and might be once out of every 10-20 times. It might not even happen for other people, so it could very well be a unit-specific issue (for example, the screen could be showing signs of failing). I do know that turning it off/on again from the switch on the back corrects the problem, which leads me to think that it's a software issue, though.

The other thing is it was one of the problems I reported to them with one of the test FWs they sent me while they were fixing a major issue I reported to them, so it could be that issue resurfaced, but it's not as consistently reproduceable.

 

[Mutex]

The only issue I'm half-facing right now is that sometimes it will not come out of standby mode if it's been sitting for a while. I haven't quite figured out the specific steps to reproduce (or narrowed down the amount time is required while sitting in standby mode), but it tends to happen with the USB input only (i.e. USB is the only active input checked in the Input Option menu). I have not encountered it when using the OPT input only so far, and it (I think) has happened once or twice when both USB and OPT were active in the Input Option menu.

I had/have this issue. Unit not waking up when I start the PC while it's connected via USB. It started happening after I plugged it into another USB HUB I have. I need to check connecting it to the other HUB.

 

[ziggurcat]

I had/have this issue. Unit not waking up when I start the PC while it's connected via USB. It started happening after I plugged it into another USB HUB I have. I need to check connecting it to the other HUB.

Mine's connected direct to my Mac, not through a hub (USB-A to USB-B cable). If you happen to observe a time frame or anything else, let me know, and I can cite that info when I put in a formal report with them. I've sort of mentioned that this is happening, but I haven't sent them a proper issue report yet.

 

[melowman]

Can someone confirm that with firmware 1.93, the reconstruction filters that are supposed to be phase linear, are not phase linear at all? None of them.

I did measurements with the ADCiso and the linear phase filters (namely F-2, F-3, F-4 and F-5) have almost the same phase response as the minimum phase filters.

 

[JIW]

Can someone confirm that with firmware 1.93, the reconstruction filters that are supposed to be phase linear, are not phase linear at all? None of them.

I did measurements with the ADCiso and the linear phase filters (namely F-2, F-3, F-4 and F-5) have almost the same phase response as the minimum phase filters.

Did you measure impulse responses?

 

[ngs428]

I wouldn't go that far, but I feel like it's the kind of thing that would have been discovered a long time ago if it was an inherent flaw in the device.

I have also unplugged/plugged the 6.35mm jack in countless times even while the 4.4mm cable is connected (and the DX5 II is on, too) with no problems. Reports of the 4.4mm being loose is a bit suspect to me, though -- I found that you have to push it in a little bit further than what you think is the proper seating. I discovered that after I connected the 4.4mm cable for the first time, and no sound came out of one side. I's always bee a very snug fit with no wiggle at all for me.

The only issue I'm half-facing right now is that sometimes it will not come out of standby mode if it's been sitting for a while. I haven't quite figured out the specific steps to reproduce (or narrowed down the amount time is required while sitting in standby mode), but it tends to happen with the USB input only (i.e. USB is the only active input checked in the Input Option menu). I have not encountered it when using the OPT input only so far, and it (I think) has happened once or twice when both USB and OPT were active in the Input Option menu.

It's also not a 100% issue, and might be once out of every 10-20 times. It might not even happen for other people, so it could very well be a unit-specific issue (for example, the screen could be showing signs of failing). I do know that turning it off/on again from the switch on the back corrects the problem, which leads me to think that it's a software issue, though.

The other thing is it was one of the problems I reported to them with one of the test FWs they sent me while they were fixing a major issue I reported to them, so it could be that issue resurfaced, but it's not as consistently reproduceable.

My current process is to switch it off at the back when not in use and switch it back on when I start to listen. I stopped using stand by mode once I realized the screen is always on. Figured that may help the screen life.

 

[BeeKay]

My current process is to switch it off at the back when not in use and switch it back on when I start to listen.

Or you go the lazy way (like me) and use home automation controlled outlets to do this job.

Nice side effect: this drives the high end power cord guys totally nuts …

 

[melowman]

Did you measure impulse responses?

I should have but no I didn’t. Will do asap. Meanwhile, I’ve written to the support.

 

[Dk33145]

Just saw that post there too. This one explanation of the issue seemed to be the most clear…

“It's the design of the 4.4mm jack combined with poor output control on the amp. The ground pin on the 4.4mm plug is at the base, so the ground connection on the amp's jack is the outermost one. Thus as you insert the plug, each pin on the plug will make contact with the ground pin in the jack. If another pin on the plug is touching one of the signal output pins, then you're effectively grounding that output. Since the 4.4mm allows for bridged amping (called balanced in the headphone world but that's inaccurate), you're effectively grounding a bridged output which you must never do. That effectively short-circuits the amp and can release the magic smoke amongst other things.

Amps should have some sort of relay that only allows output when the plug is fully inserted to prevent this. Some amps like the ones in the Sony Walkmans simply don't connect the ground pin (it's not needed for bridged amping). And in all cases, as a precaution, inserting and removing a 4.4mm plug should always be a single swift motion so that the plug pins are never touching the wrong jack pins for long.”

So how do we avoid this? Use the XLR balanced and/or only plug in headphones when the switch on the back is off. Mine has been working fine for several months. I mainly use the XLR though.

Not quite right, there isn't a connected ground with 4.4mm and this is a separate issue than the other reported issue with a 6.35 to 4.4 short. This has to do with a very momentary short when the negative of the jack contacts the positive of the outlet and vice versa as the jack is being unplugged or plugged in with the amp running. If you have some load then this momentary short can theoretically fry a component. My suspicion is that the user was driving a high impedence set of cans and the amp was turned up high, so even that momentary short damaged some internal components. Bottomline is don't try to hot swap 4.4mm cables when the amp is running (at least turn off the volume). Easy way to check your amp for damage is to just run a few constant mono sine waves. If it sounds good and balanced and no fuzz then good to go.

Edit: Another source of comfort, I don't know about earlier builds, but on my unit there is an audible click when the amp first turns on, so there is circuit protection by a relay switch and presumably it will switch off before any abnormal load damages anything.

 

[Bleib]

OK, crossfeed is now available

DX5 II version V2.07 firmware update - Topping audio

6. What's updated​

*Fixed: Several other known issues
*New: DSP crossfeed function
*Fixed: Several known issues