Review and Measurements of Massdrop THX AAA 789 Amp

[Roen]

Differential generally has twice the voltage swing, so effectively has 6dB more gain. Depends on the design, though.

It's theoretically possible to design an amp where either the single-ended or balanced/differential output sounds different from the other. But there's no inherent difference, it's down to design. In general, when you see people claiming balanced sounds better you can simply assume they are suffering from bias and/or failed to volume match.

Double the output impedance as well?

Actually, that’s another question for @AndrewMason

Are there output impedance differences between Bal and SE?

 

[derp1n]

Double the output impedance as well?

Actually, that’s another question for @AndrewMason

Are there output impedance differences between Bal and SE?

It's listed on the specs page...

 

[Roen]

It's listed on the specs page...

Ah....so it is.

Science behind the impedance doubling if the voltage and power are also doubling?

 

[AndrewMason]

Ah....so it is.

Science behind the impedance doubling if the voltage and power are also doubling?

Yes like most amps, the 789 has 2x Zout with BAL out compared to SE out. If HP impedance is left unchanged (say 32 Ohms), then Damping Factor reduces because of the 2x Zout. The 2x Vswing and 4x Pout of BAL don't play a factor in this calculation. Only the ratio of Zout to Zheadphone matters.

Higher Zout becomes a concern if it affects frequency response. Fortunately the 789 has such low Zout levels (50 mOhms SE, 100 mOhms BAL, 1 kHz) that you can confidently use SE or BAL output and maintain high damping factor.

Damping factor > 100 is desirable. In the above example with 32 Ohms headphone, neglecting cable resistance, the Damping Factors would be:
SE out: 640
BAL out: 320

 

[Roen]

Damping Factor = Headphone ZOut / Amp ZOut?

So let's say for a Shure SE846, 16/0.1 = 160 Damping Factor.

Man that IEM is demanding.

 

[AndrewMason]

Yes you got it right on your second line. DF = Zload/Zsource. Agreed, that's an insanely demanding and unforgiving IEM. Have you tried it with 789 on Gain 1?

 

[Music1969]

All good. SE Crosstalk < -100 dB @ 1 kHz is pretty low.

Nice, thanks @AndrewMason .

SE output spec is “Crosstalk: -90 dB, 300 ohms: 0.003162%”

Does that SE OUTPUT spec improve if balanced INPUTS are used with the 789? Or not really?


Similar question for balanced output spec: "Crosstalk: -127 dB, 300 ohms: 0.000045%"

Does this balanced OUTPUT spec apply only to balanced INPUT only - or both balanced and unbalanced inputs?

 

[Roen]

Yes you got it right on your second line. DF = Zload/Zsource. Agreed, that's an insanely demanding and unforgiving IEM. Have you tried it with 789 on Gain 1?

This involves me having to get a SE846, but once I do, I definitely will.

Even though Gain 1 has the "loudest" crosstalk, it's still all below audbility, right?

 

[jackenhack]

Still waiting for mine, it's been in transit for over 2 weeks, DHL sure are taking their time. Funny thing is that it took them one day to get the package from the US to Germany once they had finished processing it in the US but then it's been stuck in Germany for a week

Same problem here. Ordered it on the 14th last month, spent two weeks in Germany and is now in the incapable hands of the Swedish postal system. Gahh!

 

[AndrewMason]

Nice, thanks @AndrewMason .

SE output spec is “Crosstalk: -90 dB, 300 ohms: 0.003162%”

Does that SE OUTPUT spec improve if balanced INPUTS are used with the 789? Or not really?


Similar question for balanced output spec: "Crosstalk: -127 dB, 300 ohms: 0.000045%"

Does this balanced OUTPUT spec apply only to balanced INPUT only - or both balanced and unbalanced inputs?

Hi @Music1969 I believe SE crosstalk is with SE in/out and BAL crosstalk is with BAL in/out. You'd get very similar figures for RCA in, BAL out.

 

[AndrewMason]

This involves me having to get a SE846, but once I do, I definitely will.

Even though Gain 1 has the "loudest" crosstalk, it's still all below audbility, right?

Yes it's way down there. -93 to -100 dB depending on who's measuring it. The crosstalk of a TRS jack alone when loaded is around -75 to -90 dB depending on the headphone impedance. So the "gain 1" crosstalk issue is swamped out if you're using TRS anyways.

 

[Roen]

Yes it's way down there. -93 to -100 dB depending on who's measuring it. The crosstalk of a TRS jack alone when loaded is around -75 to -90 dB depending on the headphone impedance. So the "gain 1" crosstalk issue is swamped out if you're using TRS anyways.

More kind words to comfort my soul.

 

[amirm]

My Question... Zeos and Others have subjectively noticed huge differences in the balanced vs unbalanced performance of this and other DAC/AMPS..... Balanced being noticeably superior. Should this be so?

Balanced out has so much power that it can easily convince someone it is better because it is louder. Equalize the loudness and there should not be noticeable fidelity differences.

 

[Music1969]

Hi @Music1969 I believe SE crosstalk is with SE in/out and BAL crosstalk is with BAL in/out. You'd get very similar figures for RCA in, BAL out.

Thanks @AndrewMason

So the recommendation to use balanced for best performance is really for the headphones output? It's not much to do with the input side, correct?

 

[esm]

This involves me having to get a SE846, but once I do, I definitely will.

I have a set of older SE846 IEMs, I can give them a try tonight to see how they do on the lowest gain setting on both balanced and SE, and see whether the noise floor is audible.

These IEMs are definitely a pain in the ass when it comes to sensitivity.

 

[mi-fu]

Balanced out has so much power that it can easily convince someone it is better because it is louder. Equalize the loudness and there should not be noticeable fidelity differences.

strangely, in the manual of focal clear headphones, it also says using a balanced headphone cable will produce better sound quality. i wonder if it is just marketing BS.

 

[amirm]

strangely, in the manual of focal clear headphones, it also says using a balanced headphone cable will produce better sound quality. i wonder if it is just marketing BS.

Companies that support it certainly tout it this way.

 

[derp1n]

strangely, in the manual of focal clear headphones, it also says using a balanced headphone cable will produce better sound quality. i wonder if it is just marketing BS.


View attachment 18495

No different from many speakers that support biwiring. Market demands it, engineers include it, documentation people ask how to describe it in the manual... the circle of misinformation is complete.

 

[esm]

I have a set of older SE846 IEMs, I can give them a try tonight to see how they do on the lowest gain setting on both balanced and SE, and see whether the noise floor is audible.

So, the stack I tested with:

• DX7s DAC, in DAC mode (max gain), balanced connections into...
• JBL M-Patch 2 passive switch (max gain), balanced connections into...
• THX AAA 789 amp (lowest gain switch setting), feeding into...
• Shure SE846 IEMs circa 2014, using a balanced cable and adapters for both 4-pin XLR and 3.5mm TRS

3.5mm SE: I have to stretch to hear the noise floor at max volume; if I know it's there, I can listen for it (specifically, if I lower the volume and raise it to max quickly), but you'd be hard-pressed to notice it in regular use. I found reasonable listening levels to be about 1/4 to 1/2 of the amp's gain range depending on the source material, so I didn't find it necessary to lower gain at the source at all.

4-pin XLR: it's quiet, but you don't have to focus to notice the noise floor in the top 10% of the volume range. That said, if you try listening to actual source material at max volume, well, you won't have the problem of hearing the noise floor for very long. At reasonable listening volumes, I quickly stopped noticing it, but someone more sensitive (or listening at higher volumes) might be irritated. I found that I had too little gain control directly on the amp, and wound up lowering the gain on the M-Patch 2 to give the amp more range.

On anything but the lowest gain setting, the noise floor is clearly present on both SE and balanced.

tl;dr:

• Only low-gain, and SE is best
• Reduce output gain at the source to give you more control on the amp
• On XLR (with source gain reduction) you'll probably hear the noise floor in quiet sections

 

[Music1969]

JBL M-Patch 2 passive switch (max gain), balanced connections into...

Is it possible that this switch in the chain is affecting things ?

With the switch removed, can you still hear noise on balanced connection or does the noise disappear?