Topping DX5 II

[JIW]

It wasn't loud, that's my point though, on the same amp my Aryas are actually loud on low gain at -28dB as a reference comparison. I was playing music, any music but the main test track I first played is Ratchets by HEDEGAARD. The source chain was Blusound Powernode going COAX into the DX5 II > XLR out to the Susvara.

Either the DX5 II's power isn't what it's specced up to be, or the Susvara is faulty. Now I know Hifiman has its QC issues, but on a £6000 pair of headphones??

According to Amir's measurements, the Susvara has a sensiitivity of 94-20*log10(1.29) = 91.79 dB SPL at 1 V RMS and the Arya has a sensitivity of 94-20*log10(.214) = 107.39 dB SPL at 1 V RMS, i.e. the Arya is 15.60 dB more sensitive. Your statements suggests they are 28 dB more sensitive assuming the DX5 II works properly. Are you sure the gain settings you state are in fact correct? Is there a noticable change in volume when changing gain?

 

[audiofun]

Are you sure it switches to B? I've never read this about this model. The high gain mode of the Rhinelander must be activated by setting two jumpers inside.
Well, if that is really the case, the test isn't really valid... On the other hand, I'll test just to know how hot it will go on the surface. In any case it can go loud: With the volume pot at 12:00, I can now hear my BD T1 Mk 2 from upstairs, and it is a high impedance headphone (IIRC closer to 1 kOhm than to the stated 600 Ohms).

you don't use your hand to switch to B. It will switch automatically.

 

[respice finem]

you don't use your hand to switch to B. It will switch automatically.

I didn't say that. The high gain mode is switchable only by jumpers inside the unit (and is set to high now).
I can probably find a diagram if you're interested. No relay click when turning up etc.

 

[Roland68]

Are you sure it switches to B? I've never read this about this model. The high gain mode of the Rhinelander must be activated by setting two jumpers inside.
Well, if that is really the case, the test isn't really valid... On the other hand, I'll test just to know how hot it will go. In any case it can go loud: With the volume pot at 12:00, I can hear my BD T1 Mk 2 from upstairs, and it is a high impedance headphone (IIRC closer to 1 kOhm than to the stated 600 Ohms).

"Switching" is the wrong word.
In Class A, the power amplifier is biased by the quiescent current. When the bias voltage range is exceeded, it transitions seamlessly to Class B. I can't imagine that it has such low power that it doesn't need a heatsink in Class A. However, this is common practice with Class A/AB, as with the Singxer SA-1, for example, which is advertised as Class A but delivers higher power as Class B. And that has a large heatsink that gets very hot.

 

[audiofun]

I didn't say that. The high gain mode is switchable only by jumpers inside the unit (and is set to high now).
I can probably find a diagram if you're interested. No relay click when turning up etc.

again, when Class AB switch to B mode, it don't use a relay. the transistors will do the work automatically
Almost 95%+ the headphone amplifiers reviewed on this forum are using AB, btw. When output voltage is within range, it operates in class A mode. Otherwise class B mode.

 

[respice finem]

OK, thanks and sorry for OT.

 

[robbiekhan]

According to Amir's measurements, the Susvara has a sensiitivity of 94-20*log10(1.29) = 91.79 dB SPL at 1 V RMS and the Arya has a sensitivity of 94-20*log10(.214) = 107.39 dB SPL at 1 V RMS, i.e. the Arya is 15.60 dB more sensitive. Your statements suggests they are 28 dB more sensitive assuming the DX5 II works properly. Are you sure the gain settings you state are in fact correct? Is there a noticable change in volume when changing gain?

100% sure that the gain setting was correct, I was in low gain initially and was amazed at how little loudness there was at my expected volume level on the dial, then remembered as I use low gain on the Arya as high is not necessary, so flicked to high gain and everything I have posted here is of the Susvara in high gain on the amp.

 

[JIW]

100% sure that the gain setting was correct, I was in low gain initially and was amazed at how little loudness there was at my expected volume level on the dial, then remembered as I use low gain on the Arya as high is not necessary, so flicked to high gain and everything I have posted here is of the Susvara in high gain on the amp.

How is the volume of the Susvara with the Luxsin X9? What gain and volume setting do you use to get to a similar volume as the DX5 II at -10 dB in high gain?

Assuming the DX5 II works properly, for the X9 in low gain, it should be about 20*log10(5) - 20*log10(8.7) = -4.81 dB, in medium gain about 20*log10(5) - 20*log10(11) = -6.85 dB and in high gain about 20*log10(5) - 20*log10(16.7) = -10.48 dB.

 

[robbiekhan]

The Susvara wasn't mine, i took the DX5 II to the AV store to use it with the Susvara they had in to audition. I wanted to hear what the DX5 II was capable of on true flagship top of the top headphones (I'm ignoring Unveiled as they don't interest me at all and exist alongside as a new signature, not a replacement!).

But in terms of X9 vs DX5 loudness, between Arya and HE1000se which are my current headphones, both sound about as loud as each other when the following values are applied:

DX5 II: High gain (it has no medium gain option), -30dB volume, balanced output.
X9: Medium gain (automatically detected and set upon connection of headphones), -28dB volume, balanced output.

To me the X9 drives both Arya and HE1000se louder and cleaner sounding than the DX5 II does. Pushing the volume beyond -28dB my ears hurt, not from the loudness but from how harsh the upper midrange becomes (could that be considered distortion?), the X9 has no such trait and will just get louder and louder until my ears can't take any more.

 

[JIW]

The Susvara wasn't mine, i took the DX5 II to the AV store to use it with the Susvara they had in to audition. I wanted to hear what the DX5 II was capable of on true flagship top of the top headphones (I'm ignoring Unveiled as they don't interest me at all and exist alongside as a new signature, not a replacement!).

But in terms of X9 vs DX5 loudness, between Arya and HE1000se which are my current headphones, both sound about as loud as each other when the following values are applied:

DX5 II: High gain (it has no medium gain option), -30dB volume, balanced output.
X9: Medium gain (automatically detected and set upon connection of headphones), -28dB volume, balanced output.

To me the X9 drives both Arya and HE1000se louder and cleaner sounding than the DX5 II does. Pushing the volume beyond -28dB my ears hurt, not from the loudness but from how harsh the upper midrange becomes (could that be considered distortion?), the X9 has no such trait and will just get louder and louder until my ears can't take any more.

The loudness checks out. The levels are within around 20*log10(5) + 10 - 30 - (20*log10(11) - 28) = 1.15 dB according to the specs.

 

[wilhelm2015]

And the spectrogram for the fun of it:

View attachment 461567

Doesn't show down to DC, only down to 3Hz but you can tell the absolute mess.

hi, this measurement is very... impressive. how did you get this measurement? from what i understand this means that whatever signal you gave was audibally distorted to your reference, correct? what was your reference?
i thought toppings usually have impeccably clean good multitone response, how did this output get distorted so badly?

Also, can you simply use a EQ, maybe even the builtin one, to filter out the DC component?

 

[ziggurcat]

It wasn't loud, that's my point though, on the same amp my Aryas are actually loud on low gain at -28dB as a reference comparison. I was playing music, any music but the main test track I first played is Ratchets by HEDEGAARD. The source chain was Blusound Powernode going COAX into the DX5 II > XLR out to the Susvara.

Either the DX5 II's power isn't what it's specced up to be, or the Susvara is faulty. Now I know Hifiman has its QC issues, but on a £6000 pair of headphones??

I think if you've said that the AADAC is even quieter with the Susvara headphones, then I would argue that it's not necessarily the DX5 II (or even the AADAC). The thing about any electronics is it doesn't matter what price these things are, they are all susceptible to component failure/manufacturers defects/etc... Even a pair of headphones as expensive as those Susvaras. It could have even been the Bluesound Node. Who knows.

I was asking about the content because I know that not all recordings are created equally. I use Nigel Good's "Space Plus One" and "Flight" often to test new gear because those recordings are mastered (and I'm probably not using the correct terms, here) at a higher SPL. They're also really clean-sounding, and test a lot of different areas (imaging, bass performance, treble, etc...). My imported-from-CD copy of Rush's Hemispheres is mastered loads quieter. With settings on the D50 III in preamp mode, L50 HPA set to Low gain, volume knob on the L50 maxed, I can comfortably set the volume to -25dB on the D50 III for the Nigel Good tracks -- loud enough without any discomfort. For the Rush album, I have to raise the volume up to -15dB. Headphones are closed-back FiiO FT1s (98dB sensitivity, 32ohm). That's sort of the reasoning behind the content question -- I wanted to rule out the content as a contributing factor

 

[Sokel]

hi, this measurement is very... impressive. how did you get this measurement? from what i understand this means that whatever signal you gave was audibally distorted to your reference, correct? what was your reference?
i thought toppings usually have impeccably clean good multitone response, how did this output get distorted so badly?

Also, can you simply use a EQ, maybe even the builtin one, to filter out the DC component?

It's just the file, not a comparison.
Spectrogram does not show distortion, only levels vs frequency vs time.

 

[Jarrett]

Excluding the MX5 no one talks about, why isn't Topping focusing on making DACs that can power speakers? Their DAC/amp combos can only power headphones.

 

[Roland68]

Excluding the MX5 no one talks about, why isn't Topping focusing on making DACs that can power speakers? Their DAC/amp combos can only power headphones.

What exactly don't you like about the MX3s and MX5?

 

[Jarrett]

What exactly don't you like about the MX3s and MX5?

Not enough power and uses older chips inside I think

 

[WondrousHippo]

BTW, just to confirm for the Mac users: Topping Tune works just fine with VMWare Fusion. Just follow all the instructions to install Windows 11 Arm, install Topping Tune, ignore the "not compatible" alert, and you can tweak EQ settings with it without any trouble. Make sure you give VMware access to the USB device, it should give you a prompt when you plug in via USB, or you can click one of the device icons on top of the VMware window.

The VM runs a bit slow with default settings on my Mac Mini M4 (tbf, I didn't add a product key, maybe that's part of it?) and I can't figure out how to get it to not be just 1024x768 (the app is made for wider screens), but it works and was relatively painless, just time consuming to register the account with VMware and then set up Windows.

Update: Actually it looks like my EQ changes didn't save from yesterday so YMMV

 

[Roland68]

Not enough power and uses older chips inside I think

Not really outdated; it uses relatively recent Merus 12070s for Infineon, and for me personally, they sound better than the newer Merus MA5332.

The DX5 II is a sensible, compact device combining a DAC and HPA and doesn't involve many compromises, among other things, because the required power supplies are the same for both devices.

But with a DAC/amplifier combination, things are different, as you have to place the high-performance power supply and amplifier, with the corresponding interference, directly next to sensitive small-signal and digital circuits. This, in a much too small housing, significantly reduces the measured values.
If you're looking for something like that, you'll find it at SMSL, etc., e.g., the AO300, RAW-HA1 (very similar), etc. The AO300 was dropped from my setup due to its lack of transparency and strong load dependence. There are also reasons why the analog output of the DAC is rarely provided on such devices.

For this reason, I personally would prefer combinations like the E30 II (lite), D50 III, etc., with the PA5 II, PA7, A5(se), A7(SE), etc.
A notable exception are the SABAJ A30a and SMSL VMV A2 (very similar in design), although they are digital from start to finish and also have a certain size.

That doesn't mean that Topping isn't working on such a compact device combining a DAC and speaker amplifier, or that it won't release one in the future.
But the market for such a device is many times smaller than for the DX5 II, and therefore a much greater risk for Topping. The market doesn't need another second- or third-class device with many limitations in this area, because there are already dozens of them.

A reasonable price is also important so that a certain volume can be sold; otherwise the device wouldn't be profitable.
Devices like a combination of the E30 II Lite and PA5 II, or the D50 III and PA7, or even the DX5 II with PA5 II or PA7 would certainly be interesting.
Especially if Topping provides additional preamp and subwoofer outputs with DSP-based high- and low-pass filters, in addition to the PEQ.

 

[Jarrett]

Not really outdated; it uses relatively recent Merus 12070s for Infineon, and for me personally, they sound better than the newer Merus MA5332.

The DX5 II is a sensible, compact device combining a DAC and HPA and doesn't involve many compromises, among other things, because the required power supplies are the same for both devices.

But with a DAC/amplifier combination, things are different, as you have to place the high-performance power supply and amplifier, with the corresponding interference, directly next to sensitive small-signal and digital circuits. This, in a much too small housing, significantly reduces the measured values.
If you're looking for something like that, you'll find it at SMSL, etc., e.g., the AO300, RAW-HA1 (very similar), etc. The AO300 was dropped from my setup due to its lack of transparency and strong load dependence. There are also reasons why the analog output of the DAC is rarely provided on such devices.

For this reason, I personally would prefer combinations like the E30 II (lite), D50 III, etc., with the PA5 II, PA7, A5(se), A7(SE), etc.
A notable exception are the SABAJ A30a and SMSL VMV A2 (very similar in design), although they are digital from start to finish and also have a certain size.

That doesn't mean that Topping isn't working on such a compact device combining a DAC and speaker amplifier, or that it won't release one in the future.
But the market for such a device is many times smaller than for the DX5 II, and therefore a much greater risk for Topping. The market doesn't need another second- or third-class device with many limitations in this area, because there are already dozens of them.

A reasonable price is also important so that a certain volume can be sold; otherwise the device wouldn't be profitable.
Devices like a combination of the E30 II Lite and PA5 II, or the D50 III and PA7, or even the DX5 II with PA5 II or PA7 would certainly be interesting.
Especially if Topping provides additional preamp and subwoofer outputs with DSP-based high- and low-pass filters, in addition to the PEQ.

Thanks, yeah I definitely see them taking the power amp direction over this, I guess wisely. I've never owned a Class D so I'm curious to try something like the PA5 II or Plus model for my 4 ohm speakers. Maybe the latter just to see what giving them an extra 40 W would do.

 

[ziggurcat]

BTW, just to confirm for the Mac users: Topping Tune works just fine with VMWare Fusion. Just follow all the instructions to install Windows 11 Arm, install Topping Tune, ignore the "not compatible" alert, and you can tweak EQ settings with it without any trouble. Make sure you give VMware access to the USB device, it should give you a prompt when you plug in via USB, or you can click one of the device icons on top of the VMware window.

The VM runs a bit slow with default settings on my Mac Mini M4 (tbf, I didn't add a product key, maybe that's part of it?) and I can't figure out how to get it to not be just 1024x768 (the app is made for wider screens), but it works and was relatively painless, just time consuming to register the account with VMware and then set up Windows.

The one thing to be aware of with the ARM-based Windows 11 is that if you try to install the most recent drivers for PC, it will not. You will get a "current system is not supported" error. I have not figured out a way to run non-ARM-based drivers.

I know what you mean about the resolution, but there is a way to fix that (I fumbled through the settings, and got it to work properly). I believe when you launch VMWare, the top menu bar go to Virtual Machine > Settings > Display and pick either stretch screen to fit or use fusion settings. You can then resize the window to fill as much of your screen as you can.

You do not need to get a Windows license to use Topping Tune, thankfully. I think it's just required for larger system changes, and I won't use it for anything more than running a few PC programs that aren't on Mac. It doesn't seem to have any issues with speed on my M1 Mac Mini.

edit: @WondrousHippo Topping is going to release an update that (hopefully) fixes the PEQ settings bug (as of July 3, it was 5-10 days, so probably in the next week or so, now)