TANCHJIM Luna Portable HP Amp & DAC Review

[CedarX]

View attachment 406817

I don’t really think Luna has such complicated power circuitry.

It’s hard to see from the official image, but it seems to utilize RiCore RT6863D and single USB chipset.

View attachment 406818

…which is quite similar setup with Moonriver 2 Ti as well.

Can’t really find something fancy here.

Also noticed that, for their KA15 dongle, FiiO claims a “…real-time voltage and current monitoring for dynamic power adjustment…”. FiiO doesn’t call it “Class-H”, but it reads like it.
EDIT: forgot to mention that there is also a Class AB/H setting. Not sure what it does exactly and how it is different from this dynamic power adjustment: CS43198 mode?

The KA15 is not unlike Luna or Moonriver 2 Ti: dual CS43198 + SGM8262 (?confirmed on Luna?), but the MCU is different: a relatively unknown SPV5048Pro instead of the ComTrue CT7601, and it appears to be a lot more powerful: supports the integrated display (with animations!), UAC 1.0/2.0 modes, 10-band PEQ, coax output... KA15 is also cheaper—a better buy conditional upon actual measurements (I bought one).

EDIT2: FiiO rep. confirmed on head-fi that the KA15 desktop (switchable) mode enables more “power support”—assumed to be higher rails peak voltage.
The Class AB/H switch only affects the CS43198. Since it’s a DAC-only, it does not impact much the power consumption.

Luna Class AB/H modes may do the exact same thing: switching the CS43198 between Class AB and Class H operations, but without affecting the output opamps and rail supply…. If it’s the case, that would have a very small impact and fall into the Marketing-type stuff.

 

[Svensson]

If you only use unbalanced output, then the (still) mighty E1DA 9038D has the same SNR and outputs 2.6 Vrms vs 2 Vrms of the Tanchjim Luna. However, It may not make a any meaningful difference for IEMs and the Tamchin form factor of having the dedicated volume control should not be overlooked.

 

[GeekyBastard]

I don’t really think Luna has such complicated power circuitry.

It’s hard to see from the official image, but it seems to utilize RiCore RT6863D and single USB chipset.

…which is quite similar setup with Moonriver 2 Ti as well.

Can’t really find something fancy here.

Apparently this chip got some serious design flaws. ( #1 ) ( #2 ) ( #3 )

 

[CedarX]

Apparently this chip got some serious design flaws. ( #1 ) ( #2 ) ( #3 )

What Ivan (@IVX) outlined in that thread is the clipping behavior of thist opamp. He would never use it for an amp (dongle) output stage, but he does use it in one of his products (forget which one) because it’s cheap and perfectly suitable in that particular application.

OTOH, there are so many OEMs using this opamp in high-power dongles—Tanchjim is only one of several— that I now doubt whether these design flaws are that serious…

 

[mc.god]

What Ivan (@IVX) outlined in that thread is the clipping behavior of thist opamp. He would never use it for an amp (dongle) output stage, but he does use it in one of his products (forget which one) because it’s cheap and perfectly suitable in that particular application.

OTOH, there are so many OEMs using this opamp in high-power dongles—Tanchjim is only one of several— that I now doubt whether these design flaws are that serious…

They are also used in desktop units like Topping DX1 (Ricore) and Fiio K11 (equivalet SGM)

 

[LMDA1]

Anyone recently purchased/received one and made it work on MacOS? It plugs-and-plays and works as expected on an Ipad, but upon connect to an M2 MBP with OS v15.1, it momentarily appears and then disappears. Different cables and ports with same results. Looked around the web but can't find a single successful report of this use case, everyone else is plugging into Windows/Android/iOS.

 

[LGD_]

Has anyone found an ASIO driver for the Luna?
I looked at the Tanchjim (Shenzhen) website and came up dry there.

 

[potatosalad]

Has anyone found an ASIO driver for the Luna?
I looked at the Tanchjim (Shenzhen) website and came up dry there.

Try this:

https://tanchjim.com/en/WHQL/

 

[mc.god]

Has anyone found an ASIO driver for the Luna?
I looked at the Tanchjim (Shenzhen) website and came up dry there.

It's based on Comtrue 7601 usb bridge so the native driver should do
https://www.comtrue-inc.com/doku.php?id=ct7601_files

 

[gatucho]

Hello guys,

Do you think this cable will be fine for connecting the Luna and charging my phone at the same time?

https://www.amazon.com/gp/aw/d/B0CNGY6GQ4/

I know there are other more typical USB C splitters, but this has a male conector, which would avoid an extra cable into the Luna.

What do you think?

What I'm also wondering is that the page says that it supports 18W PD, but my phone uses up to 45W, would this damage the cable or it will limit itself?

 

[tccalvin]

CS43198 and CS43131 DAC sound defects

reference-audio-analyzer.pro

At 0:23
Left AK4377, right CS43131/CS43198 (Class H Mode)

AK4377
View attachment 406661
CS43131/CS43198
View attachment 406662

By the way, the clicks also manifest on my Apple Dongle (which AFAIK uses a different Cirrus Logic chip) and on the integrated DACs of my iPad and Pixel.

My UA Volt 2 has a CS4272 and it doesn't exhibit the issue BUT it has analog volume pots so I assume the output voltage of the DAC is fixed there.

Not sure why this is being portrayed as an issue specific to the CS43198 and CS43131.

Makes me wonder if these tests have ever been independently verified.

Has anyone actually played that test file on an AKM DAC (or any other DAC chip) at different volumes to figure out if the clicks really do not appear?

 

[destinoazell]

Hello I just bought the Luna... and annoyingly there is a noticeable hissing sound when plugging with either 3.5 or 4.4. The hissing become more subtle after the music is played or paused. Did any Luna owners facing this?

 

[Hasan Aydin]

By the way, the clicks also manifest on my Apple Dongle (which AFAIK uses a different Cirrus Logic chip) and on the integrated DACs of my iPad and Pixel.

My UA Volt 2 has a CS4272 and it doesn't exhibit the issue BUT it has analog volume pots so I assume the output voltage of the DAC is fixed there.

Not sure why this is being portrayed as an issue specific to the CS43198 and CS43131.

Makes me wonder if these tests have ever been independently verified.

Has anyone actually played that test file on an AKM DAC (or any other DAC chip) at different volumes to figure out if the clicks really do not appear?

I have discovered that as well… Unfortunately, several CS DACs have this issue.

 

[danadam]

Not sure why this is being portrayed as an issue specific to the CS43198 and CS43131.

Makes me wonder if these tests have ever been independently verified.

Has anyone actually played that test file on an AKM DAC (or any other DAC chip) at different volumes to figure out if the clicks really do not appear?

These were played through ALSA and captured using ADI-2 Pro FS at +13 dBu and 96 kHz sampling rate. Volume levels are: full, -6 dB, -12 dB, -18 dB. The captured files were normalized to -0.1 dBFS.

Tanchjim Space, volume modified in ALSA:

Tanchjim Space, volume modified using its buttons:

Fiio K3 hi-gain:

Dragonfly Red:

Samsung dongle:

JCally JM6:

Fiio JA11:

 

[tccalvin]

These were played through ALSA and captured using ADI-2 Pro FS at +13 dBu and 96 kHz sampling rate. Volume levels are: full, -6 dB, -12 dB, -18 dB. The captured files were normalized to -0.1 dBFS.

Tanchjim Space, volume modified in ALSA:

View attachment 409191

Tanchjim Space, volume modified using its buttons:

View attachment 409192

Fiio K3 hi-gain:

View attachment 409193

Dragonfly Red:

View attachment 409194

Samsung dongle:

View attachment 409195

JCally JM6:

View attachment 409196

Fiio JA11:

View attachment 409197

Wow. Thank you for taking the time to test this.

Is there any conclusion we can derive from this data? Sorry, I'm not familiar with all the devices here. I know some of them don't use CS chips at least, and for those the clicks don't seem to be there.

 

[Hasan Aydin]

Wow. Thank you for taking the time to test this.

Is there any conclusion we can derive from this data? Sorry, I'm not familiar with all the devices here. I know some of them don't use CS chips at least, and for those the clicks don't seem to be there.

Only Tanchjim Space uses CS chips, all the others do not. But I'm not sure about Samsung Dongle and Fiio JA11.

Now we need someone with a Tanchjim Luna who can switch between the two modes and measure it.

 

[tccalvin]

Only Tanchjim Space uses CS chips, all the others do not. But I'm not sure about Samsung Dongle and Fiio JA11.

Now we need someone with a Tanchjim Luna who can switch between the two modes and measure it.

I believe the Fiio KA11 (CS43131) can switch between Class AB and Class H, but I'm not sure about the JA11. Reported specs say it has an "integrated chip". Not sure what that means in practice.

 

[multiformous]

Fiio JA11 (as well as JCally JM12 and VE ODO) uses KT02H20 chip.

 

[CedarX]

I believe the Fiio KA11 (CS43131) can switch between Class AB and Class H, but I'm not sure about the JA11. Reported specs say it has an "integrated chip". Not sure what that means in practice.

I don’t think the FiiO KA11 has this feature enabled in the app.
The KA15 allows you to switch between Class AB and Class H, but it only affects the dual CS43198 (KA15), the SGMxxx output opamps are not affected and always operate in the same (Class AB?) mode.

 

[danadam]

Is there any conclusion we can derive from this data? Sorry, I'm not familiar with all the devices here. I know some of them don't use CS chips at least, and for those the clicks don't seem to be there.

As already mentioned, only Tanchjim Space is using CS43131. Others are using:

• Fiio K3 - ESS ES9038Q2M
• Dragonfly Red - ESS ES9016
• Samsung dongle - ???
• JCally JM6 - Conexant CX31993
• Fiio JA11 - KTMicro KT02H20