-
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!
Topping DX5 Review (DAC & HP Amp)
- Thread starter amirm
- Start date
Same, I emailed Topping, within less than 24h they had sent the 1.08 file, everything is back to normal after a quick update. Thanks Topping!
I wonder if this is a good upgrade from FiiO K5 Pro (suing RCA) to this one with XLR.
Received my DX5 Lite yesterday. The main difference between this and the C200 is that even when turned off with the remote it still runs a bit warm. SMSL seems to turn of completely but on the other hand it doesn't have any auto-start or auto-turnoff.
Display is really nice, quite nice information. Now waiting for new XLR cables so I can try it out on the main system.
Display is really nice, quite nice information. Now waiting for new XLR cables so I can try it out on the main system.
Twice dac went into protection mode and turns off or into DFU mode when you turn on the dsd files. This is my first dac to react like this. After he goes into DFU mode and after a reboot, he returns. But this is not normal.
Files DSD are familiar and checked many times.
Files DSD are familiar and checked many times.
Should I go with the L70 or DX5 for using with Hifiman Ananda, Focal Clear, and Hifiman Arya? I have the E30 DAC for my two JBL 305 MKII and LSR310 Subwoofer. Was wondering if getting the DX5 would be worth combining all of them into one unit instead of stacking an L70 on top of the E30?
I've had the Topping DX5 for a few month now and like it a lot. I like it so much that I had to do some nit-picking.
First is the wide-band noise. The DX5 is pretty quiet in the audio band, but the noise start to pick up as the measurement bandwidth goes up. I see 2.x uV within 20KHz bandwidth, increasing to about 18uV with 90KHz bandwidth. It registers about 1.5mV with 500kHz measurement bandwidth of an AP2722, and more than 2mV on an Agilent 34401A bench meter. Compared with the similarly priced SMSL SU-9 which is squeegee-clean above the audio band, this is less than ideal. Whether it is the switching power supply noise coupling or something else, Topping may want to look into it.
The second problem is random Left/Right channel delay in the USB mode. It's more likely to happen with higher sample rates. I've seen as much as 35 degree phase shift at 176.4k sample rate, which is about 4.96us. It's easy to reproduce. Just make a few .wav files at different sample rates containing the same 20KHz tone on both channels, and play them through USB. Observe the outputs on an oscilloscope, you'll see the left and right channels sometimes having a different phase. The amount of delay is random, and the delay would remain the same until the sample rate changes again. Since the delay in most cases is less than the sample interval of the source file, it's unlikely caused by anything from the source file to the USB interface, but more likely caused by the ES9068 or the way DX5 handles it during sample rate changes. S/PDIF via coax doesn't have this problem.
The third 'issue' is perhaps a well-known one, that certain manufacturers optimize their product based on how it is typically measured, and not necessarily the everyday usage scenario. We've seen the stellar THD+N performance of the unit at 0dBFS input. The following picture has slightly higher harmonics than the one Amir did because my AP is a lesser model than his APX555. But it's good enough to prove my point.
Once we slightly reduce the input amplitude by 3dB, the harmonics get a lot worse.
Apparently the volume is done digitally (no surprise). Attenuating the input by 3dB has the same effect as setting -3dB volume with 0dBFS input:
If ESS offers some kind of THD optimization feature in their DAC chips that can get better measured numbers, it would be stupid for Topping not to use it. But we know that's good for the measurement only, because nobody listens to 0dBFS sine wave all day.
Despite those minor issues, the DX5 is still my daily driver. From time to time I swap in the SU-9, within a couple of songs the DX5 would take back the driver seat. There must be something that Topping engineers did right. I just wish they can make the product more polished, maybe too much to ask for products at this price range? Also, I'm not suggesting that measurement is useless in picking a product that 'sounds better'; it's just that the limited measurement we've seen so far may not correlate with subjective impressions very well. There is still a lot to explore in that front. Believing in science means that we believe there is a lot more that we don't know than we already know.
For example, I can see that the SU-9 has an analog reconstruction filter with -3dB at about 83kHz. We listen through the filter even with all the hi-res sources. The DX5 on the other hand has a much more flatter response in the analog domain, which somewhat contributed to the higher broadband noise. Is that the major factor that makes them sound differently?
Last but not least, let's take a look at what's with the "Transistor sound":
and "Valve sound":
The sound coloring offered by SMSL is similar, with different combinations of 2nd and 3rd harmonics. If you don't hear a difference at all, congratulations, you could save a lot of $$ by not having to buy a DAC based on the SINAD leader board!
First is the wide-band noise. The DX5 is pretty quiet in the audio band, but the noise start to pick up as the measurement bandwidth goes up. I see 2.x uV within 20KHz bandwidth, increasing to about 18uV with 90KHz bandwidth. It registers about 1.5mV with 500kHz measurement bandwidth of an AP2722, and more than 2mV on an Agilent 34401A bench meter. Compared with the similarly priced SMSL SU-9 which is squeegee-clean above the audio band, this is less than ideal. Whether it is the switching power supply noise coupling or something else, Topping may want to look into it.
The second problem is random Left/Right channel delay in the USB mode. It's more likely to happen with higher sample rates. I've seen as much as 35 degree phase shift at 176.4k sample rate, which is about 4.96us. It's easy to reproduce. Just make a few .wav files at different sample rates containing the same 20KHz tone on both channels, and play them through USB. Observe the outputs on an oscilloscope, you'll see the left and right channels sometimes having a different phase. The amount of delay is random, and the delay would remain the same until the sample rate changes again. Since the delay in most cases is less than the sample interval of the source file, it's unlikely caused by anything from the source file to the USB interface, but more likely caused by the ES9068 or the way DX5 handles it during sample rate changes. S/PDIF via coax doesn't have this problem.
The third 'issue' is perhaps a well-known one, that certain manufacturers optimize their product based on how it is typically measured, and not necessarily the everyday usage scenario. We've seen the stellar THD+N performance of the unit at 0dBFS input. The following picture has slightly higher harmonics than the one Amir did because my AP is a lesser model than his APX555. But it's good enough to prove my point.
Once we slightly reduce the input amplitude by 3dB, the harmonics get a lot worse.
Apparently the volume is done digitally (no surprise). Attenuating the input by 3dB has the same effect as setting -3dB volume with 0dBFS input:
If ESS offers some kind of THD optimization feature in their DAC chips that can get better measured numbers, it would be stupid for Topping not to use it. But we know that's good for the measurement only, because nobody listens to 0dBFS sine wave all day.
Despite those minor issues, the DX5 is still my daily driver. From time to time I swap in the SU-9, within a couple of songs the DX5 would take back the driver seat. There must be something that Topping engineers did right. I just wish they can make the product more polished, maybe too much to ask for products at this price range? Also, I'm not suggesting that measurement is useless in picking a product that 'sounds better'; it's just that the limited measurement we've seen so far may not correlate with subjective impressions very well. There is still a lot to explore in that front. Believing in science means that we believe there is a lot more that we don't know than we already know.
For example, I can see that the SU-9 has an analog reconstruction filter with -3dB at about 83kHz. We listen through the filter even with all the hi-res sources. The DX5 on the other hand has a much more flatter response in the analog domain, which somewhat contributed to the higher broadband noise. Is that the major factor that makes them sound differently?
Last but not least, let's take a look at what's with the "Transistor sound":
and "Valve sound":
The sound coloring offered by SMSL is similar, with different combinations of 2nd and 3rd harmonics. If you don't hear a difference at all, congratulations, you could save a lot of $$ by not having to buy a DAC based on the SINAD leader board!
Ppl, I need your help.
I'm planning to upgrade my Topping DX5 by soldering new op amps to the PCB.
Can I desolder the small OPA1612A (marking OVII TI 15I) and replace them with other op amps that have SOIC8/SOP8 package?
Are the places, that I showed in the photo, the ones that I need to solder the leads/foots to?
Will it work?
Sorry for my english and thanks in advance!
I'm planning to upgrade my Topping DX5 by soldering new op amps to the PCB.
Can I desolder the small OPA1612A (marking OVII TI 15I) and replace them with other op amps that have SOIC8/SOP8 package?
Are the places, that I showed in the photo, the ones that I need to solder the leads/foots to?
Will it work?
Sorry for my english and thanks in advance!
Attachments
How do you know this is a upgrade. What tells you this?
You're right. I don't have knowledge about the circuit design and can't read it. As I see on the board, there are 7 positions that manufacturer uses to place OPA1612 to. I've marked them on the photo (hope that I've marked them right
). Which positions will affect the sound best? Which should I replace to upgrade (?) the sound if I have better (?) op amps?Maybe it will not upgrade anything, even can downgrade
But I want to try.I see that Topping D50s, DX7 Pro, D10s have swappable op amps, so I want to try it with DX5 by soldering.
Attachments
Veri
Master Contributor
- Joined
- Feb 6, 2018
- Messages
- 9,598
- Likes
- 12,040
Sounds like a world of pain.Maybe it will not upgrade anything, even can downgrade
I see that Topping D50s, DX7 Pro, D10s have swappable op amps, so I want to try it with DX5 by soldering.
Does the DX5 usb work without needing a driver? (Recognize in Windows as audio device without needing any kind of driver?).
Veri
Master Contributor
- Joined
- Feb 6, 2018
- Messages
- 9,598
- Likes
- 12,040
Of course, sure! Just not on Windows 7 but shame on anyone still using that dinosaur.
iconoclast
Member
- Joined
- Sep 2, 2022
- Messages
- 52
- Likes
- 48
For some reason I missed that this is just a single-ended headphone amp when I bought it 6 months ago (didn't use it until now); I thought that was just the EX5. Oh well, who am I kidding? It's not like I wasn't going to buy another headphone amp anyway...
Other than that (which is entirely my own fault), I really like it. The knob quality is a step up from the DX3Pro+.
Other than that (which is entirely my own fault), I really like it. The knob quality is a step up from the DX3Pro+.
alekksander
Active Member
Is it Your unit that is disassembled? Can i ask You do show me the volume knob pin? and maybe some measurements of vol knob?You're right. I don't have knowledge about the circuit design and can't read it. As I see on the board, there are 7 positions that manufacturer uses to place OPA1612 to. I've marked them on the photo (hope that I've marked them right
Maybe it will not upgrade anything, even can downgrade
I see that Topping D50s, DX7 Pro, D10s have swappable op amps, so I want to try it with DX5 by soldering.
alekksander
Active Member
@simmconn
very nice post, thanks for sharing it! and the conclusions as well. superb!
personally i'm one step from getting this unit, but this discovery makes me hesitate again…
very nice post, thanks for sharing it! and the conclusions as well. superb!
does it depend of number of tries, or restricted to sample rate? could You show us some more examples of when does it happen and of what come out perfectly?
personally i'm one step from getting this unit, but this discovery makes me hesitate again…
well said, that explains why some folks claim to notice a difference. while it fades, there still are some scenarions and therefore some DACs may sound a bit different to some than to the others.
Last edited:
Similar threads
