xDuoo TA-20 Tube Balanced Headphone Amplifier Review

[hmscott]

That's a nice sized pairing between those two units, the SMSL M500 looks like it fits right there alongside the TA-20, sizes pair nicely.

Also, the locally built Hungary cable looks solid - if not a bit expensive.

Have you put it all together? How does it perform? Have you picked up any other tubes to try in the TA-20?

IDK what the output of the M500 is on XLR, but as above I needed to reduce the drive from the D90 MQA XLR output by -9.0dB to be able to "turn up the volume" on the TA-20 without distortion. I have the D90 MQA set at -12.0dB right now, seems fine either way, I may go back to -9.0dB.

I did get a few more tidbits of feedback from Xduoo to my questions, specifically how they would rate each of the tubes they earlier listed as having been tested on the TA-20 by Xduoo:

As translated by Hifigo from Xduoo:

PSVANE 12AU7 这个管在人声方面会更好， ( better on vocal)
GOLD LION 12AU7 这个管在器乐方面表现会更好，( better on instruments)
Mullard 12AU7 这个管低频会更好一些, ( better on bass)

Google translate:

PSVANE 12AU7 This tube will be better in terms of vocal, (better on vocal)
GOLD LION 12AU7 This tube will perform better in instrumental music, (better on instruments)
Mullard 12AU7 this tube will have better low frequency, (better on bass)

I realized I should test the Matched Gain pairs of tubes I ordered in addition to the TA-20 to see if they were functional. Good thing I did test them as one pair - the PSVANE 12AU7-S Art Series pair had a bad tube - very low gain. Easy to notice, I swapped tube socket positions to see if the low gain moved with the tube, and it did. So I've started the Amazon return.

The PSVANE 12AU7-T MKII tubes are both strong and do indeed have a vocal forward presentation, so much so I'd recommend them to anyone that has trouble hearing the vocals in videos, TV, etc. I can turn down the volume to 30 and the voices stand out all the way down.

The normal volume setting I'm using on Youtube is 52 for voice centered media. The movie streaming services don't have as much drive and need to have the volume turned up to 65. For movies, I turn up the sound a bit higher than voice media to enjoy the dynamics of the presentation. Again, the top volume setting on the TA-20 is 95.

I haven't done much music testing yet, the tubes are fresh out of the box and tubes have already been rotated through several pairings, and it has hardly been 18 hours since I unboxed and set up the TA-20. More later.

Update: Something else I noticed, TA-20 China labeled "generic" tubes are printed with the same shade of Red and a similar font to the PSVANE tubes listed here:
https://www.aliexpress.com/item/1005001745244020.html

I've asked Xduoo via Hifigo.

 

[Noodles]

@hmscott So i got some of my tubes delivered today and i slotted in the "matching" pair of Psvane 12AU7-S.
Decent at first listen.
However microphonic (not by a lot) and one of the channels is noisier... so i have no idea how they matched them... ?
Still have some JJ tubes to test, have a few of them here and i will wait some 20 hours to see if the PSVANE improves.

 

[muslhead]

i owned this and found it did not have enough gas in the tank to drive my hifiman he1000v2's. By the time i pushed it to high listening levels, distortion was horrible.

 

[Noodles]

@muslhead I can understand that completely, but that's where you need to step up to medium gain tubes. Most 12AU7's are low gain.
12AT7 is where it's at for you.

 

[hmscott]

@hmscott So i got some of my tubes delivered today and i slotted in the "matching" pair of Psvane 12AU7-S.
Decent at first listen.
However microphonic (not by a lot) and one of the channels is noisier... so i have no idea how they matched them... ?
Still have some JJ tubes to test, have a few of them here and i will wait some 20 hours to see if the PSVANE improves.

I'm not sure I'll rebuy the 12AU7-S Art Series until they come in a factory boxed set of Gain Matched tubes. They may be perfecting the manufacturing of the new tube as both of us got a "bad" tube in the pair. The weak gain tube in my pair wasn't salvageable, so a full return is all I could do. I'll likely wait till PSVANE offers the 12AU7-S's in factory boxed set like the all their other lines - and the 12AU7-T MKII

@muslhead I can understand that completely, but that's where you need to step up to medium gain tubes. Most 12AU7's are low gain.
12AT7 is where it's at for you.

I'm not sure that is effectively true. That's why the 12AX7 overdrives the SS output and distorts. Overdriving the SS final amp won't help.

I'm waiting to hear back from Xduoo on their testing results for 12AT7 / 12AX7 tubes in the TA-20, they hadn't tested them yet. I'm hoping it sparks some ideas for future tube headphone amps besides telling us if the 12AT7 will work in the TA-20.

Fortunately, my hardest to drive headphones have plenty of headroom with the TA-20, but I do hope Xduoo follows up with a 3000mW version based on the TA-30 DAC/AMP - with a headphone AMP only version @ 3000mW for headphones like @muslhead is trying to drive.

I also tested the RCA "Baldwin" Gain Matched / Noise tested NOS tubes from audiotubes.com, and they sound great - no noise during play.

One tube does emit a "ting" when flicking it with a forefinger nail, but it doesn't continue the ring, so it's not a failure. I do recall tubes being this way long ago. The ones that sustained the "ring" were failures as being too microphonic.

The other tube is silent against physical assault. The balance between the two is perfect, there is no imbalance - same for the PSVANE 12AU7-T pair and stock generic pair that came with the TA-20, all well Gain Matched.

The RCA "Baldwin" tubes are not vocal forward presentation, which I do miss a bit after the PSVANE 12AU7-T MKII's, but they filled out the sound stage full left to right, unlike the new PSVANE / stock tubes which are slowly opening up their sound stage presentation and started out with a very narrow sound stage.

Update: I received an email back from the vivatubes.com guys about the values on the sticker on the edge of the boxes, and which ones matter for gain matching. The PSVANE numbers for the previous-gen matched set in the box from the factory were exactly equal - for both halves of each tube and tube vs tube.

That may be a problem when matching these tubes in the field, the accuracy is limited to the "tubes on hand", so the 12AU7-S Art Series single tube boxing were matched as best possible with stock on hand. I'll mention that to them and ask that we wait for a boxed set or new inventory that has exactly matching gain between tubes:

18.4mA/17.4mA - Left Tube, Side 1/2 Gain readings
19,0mA/18.5mA - Right Tube, Side 1/2 Gain readings
2110gm/2065gm Left Tube, Side 1/2 transconductance
1640gm/2030gm Right Tube Side 1/2 transconductance

"The numbers with the decimal points are the Gain test
results, and these are what are used to match each set. The other
four-digit numbers are simply the "transconductance" values, indicating
that the electrons are flowing properly within the tube. They do not
need to be close in value to each other."

PSVANE boxed set only listed gain for both sides of each tube which matched exactly at 9.5mA/9.5mA 9.5mA/9.5mA, but didn't list transconductance, the RCA "Baldwin" gain matched tubes didn't actually list gain readings, only transconductance for both sides of both tubes. I'll ask Brent about that.

The well-measured pair(s) of tubes in the TA-20 sound great to my ears, with distinct sound attributes for each make/model.

 

[Noodles]

AU tubes have an amp factor of 18, AT are around 40 and AX are between 90 and 100.
Soo yeah AX are a no go.
But I got a pair of AT I can test later

 

[muslhead]

@muslhead I can understand that completely, but that's where you need to step up to medium gain tubes. Most 12AU7's are low gain.
12AT7 is where it's at for you.

No, the answer for me was solid state providing 6watts. That solved my problem

 

[hmscott]

AU tubes have an amp factor of 18, AT are around 40 and AX are between 90 and 100.
Soo yeah AX are a no go.
But I got a pair of AT I can test later

That would be nice to know if the TA-20 output actually increases - a lower setting on the volume indicated required for the same loudness on the same headphone with the 12AT7 vs 12AU7.

So far all of the pairs of tubes I've run through the TA-20 resulted in the same output - tested at 55 / 65 indicated volume. It looks to me like the pre-amp drive stage is limited in the effect on TA-20 final output.

No, the answer for me was solid state providing 6watts. That solved my problem

How much headroom remains with that 6 watt amp? Would half that power be enough? I'm thinking Xduoo might release a 3000mW headphone amplifier only product to match the output of the TA-30.

Which SS headphone amp is it?

 

[Noodles]

AT7's are now in, no distortion, no noise or microphonics.
JJ ECC81 / 12AT7

Used HD800 as a reference.
On AU7 - i listen comfortably at around 77 on the dial
Now i can't go past 70. 77 is now quite uncomfortable to listen to.

The amplification is not linear however, there's not much going on until like 55~ and then it becomes exponential.

AX7 tubes wont do on this amp, i'm not even going to try, but that was always something i kinda knew.

With the JJ tubes i actually had to check if i had not enabled the bass eq on the RME DAC... it definitely adds lows and also makes more sparkles in the highs. I don't perceive anything different going on with vocals.

 

[muslhead]

That would be nice to know if the TA-20 output actually increases - a lower setting on the volume indicated on the same headphone with the 12AT7. So far all of the tubes I've run through resulted in the same output - tested at 55 / 65 indicated volume. It looks to me like the pre-amp drive stage is limited in the effect on TA-20 final output.

How much headroom remains with that 6 watt amp? Would half that power be enough? Which SS headphone amp is it?

Schiit Modius

 

[Noodles]

Schiit Modius

As long as you run that shiit ballanced, it's a fine amp.

 

[hmscott]

AT7's are now in, no distortion, no noise or microphonics.
JJ ECC81 / 12AT7

Used HD800 as a reference.
On AU7 - i listen comfortably at around 77 on the dial
Now i can't go past 70. 77 is now quite uncomfortable to listen to.

The amplification is not linear however, there's not much going on until like 55~ and then it becomes exponential.

AX7 tubes wont do on this amp, i'm not even going to try, but that was always something i kinda knew.

With the JJ tubes i actually had to check if i had not enabled the bass eq on the RME DAC... it definitely adds lows and also makes more sparkles in the highs. I don't perceive anything different going on with vocals.

Interesting, now I'm going to have to try an AT7

Remember, I heard "sparkles" in the highs which turned out to be distortion from overdriving the TA-20 with the D90 MQA output - backing off to -9dB and then -12dB on in D90 MQA Pre mode output stopped the sparkles.

I've read that the JJ ECC82's are good tubes too; one of the tubes Xduoo tested.

Thank you for testing, please do come back with more feedback as you run the 12AT7 over time.

 

[Noodles]

Remember, I heard "sparkles" in the highest which turned out to be distortion from overdriving the TA-20 with the D90 MQA output - backing off 9dB-12dB stopped the sparkles.

No distortion here, it's quite clean. What i mean is that specifically metal instruments sound a bit more .... zingy.

I quite like the JJ's also have a set of their gold pin JJ 802S and a pair of 802S tin pins (both AU7's).
If the sound signature is the same across their line, i think i found what i was looking for.

 

[hmscott]

No distortion here, it's quite clean. What i mean is that specifically metal instruments sound a bit more .... zingy.

I quite like the JJ's also have a set of their gold pin JJ 802S and a pair of 802S tin pins (both AU7's).
If the sound signature is the same across their line, i think i found what i was looking for.

Which JJ model are you running? Link(s)?

 

[Noodles]

The gold pin balanced matched JJ 802S either one or both of them trip the protection circuit on the TA-20.
The regular tin pins 802S work just fine ... now i need to see what the hell is going on there ...

 

[Noodles]

These are tripping the protection:
https://www.jj-electronic.com/en/ecc802s-gold

These work fine:
https://www.jj-electronic.com/en/ecc802s

The gain is on par with their AT7 counterpart.

Looks like they weren't lying when they say it's a medium gain tube.

 

[hmscott]

The gold pin balanced matched JJ 802S either one or both of them trip the protection circuit on the TA-20.
The regular tin pins 802S work just fine ... now i need to see what the hell is going on there ...

These are tripping the protection:
https://www.jj-electronic.com/en/ecc802s-gold

These work fine:
https://www.jj-electronic.com/en/ecc802s

The gain is on par with their AT7 counterpart.

Looks like they weren't lying when they say it's a medium gain tube.

Yipe, probably just a bad tube

These are the ones I was looking at, they are the Gold pin label JJ tubes:

JJ ECC82 / 12AU7 Gold Pin Preamp Vacuum Tube | TubeDepot.com

The same online store has the RCA Clear Top for $300/pair... I got the "Organ"/"Baldwin" branded version from audiotubes.com with gain matching and noise testing for $60/pair shipped.

BLACK SABLE RCA 12AU7 Clear Top | TubeDepot.com

I'll steer clear of that place... sheesh.

A lot of variance in price for the NOS RCA Clear Top tubes:
https://www.google.com/search?q=RCA+12AU7+Clear+Top

As far as "medium gain", which tubes are those? I noticed the PSVANE T MKII match pair are 9.5mA vs S Art Series are 2x that, I wonder if it's the testing method? There is no audible increase in volume on the good tube of the Art Series pair, I'm surprised you are noticing volume differences between tubes and with that AT7.

 

[hmscott]

These are tripping the protection:
https://www.jj-electronic.com/en/ecc802s-gold

These work fine:
https://www.jj-electronic.com/en/ecc802s

You might try swapping one gold pin with a tin pin tube to see if you can find the offending gold pin tube, or if both gold pin tubes are tripping protection - then the gold pin tube design / build would be incompatible with the TA-20.

 

[Noodles]

Yipe, probably just a bad tube

Yes, one of the gold pins is busted... now gotta return it ehhhh.

BLACK SABLE just means cryo treated. Supposedly lasts longer than normal but yeah idk. The ones you got a cryo treated these:
https://www.jj-electronic.com/en/ecc82-12au7-gold

Regarding the gain - the PSVANE is on par with the mullards, which are on par with the EH - low gain.
After you mentioned said gain i retested with different tubes in each socket (yes you can do that). AT7's are definitely louder than the AU7's.
I also tested to see if the JJ 802s is on par with the EH and yes they are the same gain. Incidentally the sound change was more subtle than i expected:
The JJ gives metallic instruments the zing, EH is very soft overall. Bass is comparable. Treble is also softer on the EH, but has slightly accentuated "S" compared to the JJ.

My greatest befuddlement came when i plugged in the stock tubes... they sound quite close to the JJ's without the metal zingy and have more reverb on low notes, where the JJ has more reverb on acoustic instruments.

Going back to the PSVANE to see if i get some hours on them the microphonics will disappear.

 

[hmscott]

Schiit Modius

The Modius is a DAC, did you mean the Magnius?

https://www.schiit.com/products/magnius

Easy to mix up those names

What the heck is a Magnius? That doesn’t sound Norse.
It’s like Modius. When we were developing a bigger, better, balanced Modi, we started jokingly calling it “Modius Maximus.” The name stuck. So when we also started working on a bigger, better, balanced Magni Heresy, we started calling it “Magnius.” That name stuck too. So now you’re stuck with Modius and Magnius.

But you’re conflating Norse mythology with Roman terminology!
Right.

But that doesn’t make sense!
Dude, you used the word “conflating.”

Sigh. Okay. Fine. I guess that’s all we should expect from a company that gets all its Norse mythological information from comic books.
And Wikipedia. Don’t forget Wikipedia!

The Magnius measurements and spec's look good, how does the audio drive compare to the TA-20? What is the impedance for your
headphones? And, which DAC do you use?

Spoiler: Magnius power / specs

Balanced Headphone Output:
Maximum Power, 16 ohms: 6.0W RMS per channel
Maximum Power, 32 ohms: 5.0W RMS per channel
Maximum Power, 50 ohms: 3.2W RMS per channel
Maximum Power, 300 ohms: 1000mW RMS per channel
Maximum Power, 600 ohms: 500mW RMS per channel

Single-Ended Headphone Output:
Maximum Power, 16 ohms: 2.2W RMS per channel
Maximum Power, 32 ohms: 2.0W RMS per channel
Maximum Power, 50 ohms: 1.3W RMS per channel
Maximum Power, 300 ohms: 300mW RMS per channel
Maximum Power, 600 ohms: 150mW RMS per channel

Note: Maximum power rated at 1% THD, high gain

THD+N:
Low Gain, Balanced Output: Less than 0.0001% (-119dB) at 4V RMS into 32 ohms
High Gain, Balanced Output: Less than 0.0003% (-110dB) at 4V RMS into 32 ohms
Low Gain, SE Output: Less than 0.003% (-90dB) at 1V RMS into 32 ohms
High Gain, SE Output: Less than 0.004% (-87db) at 2V RMS into 32 ohms

IMD:
Low Gain, Balanced Output: Less than -112dB at 4V RMS into 32 ohms, CCIF
High Gain, Balanced Output: Less than -110dB at 4V RMS into 32 ohms, CCIF
Low Gain, SE Output: Less than -73dB at 1V RMS into 32 ohms, CCIF
High Gain, SE Output: Less than -72dB at 2V RMS into 32 ohms, CCIF

SNR:
Low Gain, Balanced Output: Greater than 125dB, unweighted, referenced to 4V RMS
High Gain, Balanced Output: Greater than 115dB, unweighted, referenced to 4V RMS
Low Gain, SE Output: Greater than 105dB, unweighted, referenced to 1V RMS
High Gain, SE Output: Greater than 110dB, unweighted, referenced to 2V RMS

Crosstalk:
Less than -90dB, 20 Hz-20 kHz, either gain, 300 ohm load
Less than -70dB, 20 Hz-20 kHz, either gain, 32 ohm load

Output Impedance: Less than 0.1 ohms at either gain
Input Impedance: 50k ohms

Gain:
Balanced Output: 1 (0db) or 5 (14db)
Single-ended Output: 0.5 (-6dB) or 2.5 (8db)

Topology: Composite amplifier with OPA1688 input buffers, LME49724 differential amplifier, TPA6120A2 current feedback output stage, DC coupled.

Protection: Standard failsafe DC power input and muting relay

Power Supply: “Wall wart” style 24VA 15VAC transformer, linear regulated +/- 13.5V rails with over 6,000uF filter capacitance

Power Consumption: 6W idle, 21W max

Size: 9 x 6 x 1.5”

Weight: 2 lb

APx555 Report for Magnius