Review and Measurements of Massdrop THX AAA 789 Amp

[Music1969]

Hi all, I just wanted to give an update regarding the crosstalk at the lowest gain. It occurred to me, since I had to put the volume control at 4 o'clock in the test that had the crosstalk issue, perhaps it's the potentiometer that's causing it. So I did another test, this time with gain setting "2", with 20 dBu input and the volume set to 1-2 o'clock to reach 6 dBV output, and I get a similar, albeit lower, increase in crosstalk!

So it looks like the increase in crosstalk is caused by the potentiometer (and varies with the position of the potentiometer), and not the gain setting.

View attachment 18352

Hi, is this with single ended RCA inputs?

If so, is the same issue there with balanced XLR inputs?

Cheers

 

[pkane]

Hi all, I just wanted to give an update regarding the crosstalk at the lowest gain. It occurred to me, since I had to put the volume control at 4 o'clock in the test that had the crosstalk issue, perhaps it's the potentiometer that's causing it. So I did another test, this time with gain setting "2", with 20 dBu input and the volume set to 1-2 o'clock to reach 6 dBV output, and I get a similar, albeit lower, increase in crosstalk!

So it looks like the increase in crosstalk is caused by the potentiometer (and varies with the position of the potentiometer), and not the gain setting.

View attachment 18352

Interesting. So is the crosstalk increasing with increased volume position of the potentiometer? Can you measure crosstalk at a few more positions to get a better sense of how it changes? I normally have it set near the maximum level position, just a notch or two below, and then control volume via software.

 

[TheBatman_Yo]

Hi all, I just wanted to give an update regarding the crosstalk at the lowest gain. It occurred to me, since I had to put the volume control at 4 o'clock in the test that had the crosstalk issue, perhaps it's the potentiometer that's causing it. So I did another test, this time with gain setting "2", with 20 dBu input and the volume set to 1-2 o'clock to reach 6 dBV output, and I get a similar, albeit lower, increase in crosstalk!

So it looks like the increase in crosstalk is caused by the potentiometer (and varies with the position of the potentiometer), and not the gain setting.

View attachment 18352

As a guy that doesn't really understand what all this means, should I be worried? Is this crosstalk issue audible?

 

[wadec22]

Interesting. So is the crosstalk increasing with increased volume position of the potentiometer? Can you measure crosstalk at a few more positions to get a better sense of how it changes? I normally have it set near the maximum level position, just a notch or two below, and then control volume via software.

you should always max software volume to the dac and control volume from the amplifier for max fidelity. i'm not going to pretend to know why, but I know the JDS manuals and others always recommend this.

 

[wadec22]

As a guy that doesn't really understand what all this means, should I be worried? Is this crosstalk issue audible?

he said earlier, certainly not audible via single ended output, not sure about balanced. balanced output would be interesting, as I would assume with the higher headroom for power, you would be setting the potentiometer lower.

 

[L-Train]

Hi, is this with single ended RCA inputs?

This is with balanced inputs.

Interesting. So is the crosstalk increasing with increased volume position of the potentiometer? Can you measure crosstalk at a few more positions to get a better sense of how it changes?

Crosstalk seems to increase with increased volume position, but I only have 2 data points so I can't be certain. I'm with you though, it would be interesting to see where/how crosstalk improves since at full volume crosstalk is practically non-existent. On the other end, my soundcard only has 2 input level sensitivities (20 dBu and 6 dBV) so measuring crosstalk at lower levels would probably be dominated by noise and may not be entirely accurate.

As a guy that doesn't really understand what all this means, should I be worried? Is this crosstalk issue audible?

You shouldn't be worried. I don't believe it's an audible issue. I haven't heard any issues with crosstalk while listening to my 789.

As a side note, I looked at Tyll's headphone amp measurements here: https://www.innerfidelity.com/headphone-amp-measurements and saw that many amps also have a similar increase in crosstalk vs. frequency.

 

[helloworld]

This is with balanced inputs.


Crosstalk seems to increase with increased volume position, but I only have 2 data points so I can't be certain. I'm with you though, it would be interesting to see where/how crosstalk improves since at full volume crosstalk is practically non-existent. On the other end, my soundcard only has 2 input level sensitivities (20 dBu and 6 dBV) so measuring crosstalk at lower levels would probably be dominated by noise and may not be entirely accurate.


You shouldn't be worried. I don't believe it's an audible issue. I haven't heard any issues with crosstalk while listening to my 789.

As a side note, I looked at Tyll's headphone amp measurements here: https://www.innerfidelity.com/headphone-amp-measurements and saw that many amps also have a similar increase in crosstalk vs. frequency.

looks like jds o2 has a similar problem, maybe jds atom has this problem too? @amirm

 

[Veri]

The amp is all sold out for now, btw.

 

[Music1969]

This is with balanced inputs.

Thanks for confirming. Single ended may be slightly worse?

Wonder what @AndrewMason may think?

 

[seslan]

The amp is all sold out for now, btw.

Is there any information on when the next drop will happen? I was not able to find anything on massdrop's discussion thread.

 

[Kyle / MrHeeHo]

Is there any information on when the next drop will happen? I was not able to find anything on massdrop's discussion thread.

request the drop and they will send you an email ahead of time

 

[pkane]

you should always max software volume to the dac and control volume from the amplifier for max fidelity. i'm not going to pretend to know why, but I know the JDS manuals and others always recommend this.

That’s mostly an audiophile myth. A small software volume adjustment does not reduce fidelity in any audible way.

 

[wadec22]

That’s mostly an audiophile myth. A small software volume adjustment does not reduce fidelity in any audible way.

interesting. the JDS crew is not one I take for being susceptible to audio myths. they always recommend setting software to full volume and dac to 24 bit. right or wrong, I'll continue to follow their recommendations. they seem the opposite of all audio snake oil.

 

[pkane]

interesting. the JDS crew is not one I take for being susceptible to audio myths. they always recommend setting software to full volume and dac to 24 bit. right or wrong, I'll continue to follow their recommendations. they seem the opposite of all audio snake oil.

Maybe they know something about their design we don't. A 24 bit signal normally has enough of room to attenuate in digital domain without destroying information, at least 2-3 bits (12-18dB) on a good DAC.

Digital volume control may not always be bit-perfect, but it will be very close. Besides, bit-perfect is not something I'm looking for as I've been using DSP for the past 6-7 years

My software player volume control performs volume adjustments in double-length floating point, so about 52 bits of actual resolution. This represents about a 168dB volume range before the lower 24 bits are in any danger of being harmed. That's enough headroom for my listening tastes

 

[wadec22]

Maybe they know something about their design we don't. A 24 bit signal normally has enough of room to attenuate in digital domain without destroying information, at least 2-3 bits (12-18dB) on a good DAC.

Digital volume control may not always be bit-perfect, but it will be very close. Besides, bit-perfect is not something I'm looking for as I've been using DSP for the past 6-7 years

My software player volume control performs volume adjustments in double-length floating point, so about 52 bits of actual resolution. This represents about a 168dB volume range before the lower 24 bits are in any danger of being harmed. That's enough headroom for my listening tastes

I have no doubt that you know what you are talking about, I've just always followed their recommendations as they are one of the few companies I feel like I can actually trust. From the manuals of the El Amp & OL DAC:

"For maximum signal fidelity, set the DAC to 24 bit audio and adjust digital volume slightly below 100% on your computer. Only control volume from the amplifier."

 

[wyrainer]

I have no doubt that you know what you are talking about, I've just always followed their recommendations as they are one of the few companies I feel like I can actually trust. From the manuals of the El Amp & OL DAC:

"For maximum signal fidelity, set the DAC to 24 bit audio and adjust digital volume slightly below 100% on your computer. Only control volume from the amplifier."

First, Sorry for my poor English, I hope I can explain it clearly.

I don't know if I am right, you can judge it yourself

Let me make some assumption:
1. The distortion and noise will come when Analog signal goes from DAC to AMP, let's make a assumption that the distortion and noise is 1db
2. the dac outputs 50db at max volume
3. I hope that my headphone gives me 100db

then:
1. set pc volume 10db, because I wish get 100db, the amp will amplify 10 times, meanwhile the distortion and noise will be amplified 10 times , my headphone will get 10db distortion and noise
2. set pc vulume 50db, the amp will amplify 2 times, so my headphone will get 2db distortion and noise

so it is better to set pc and dac volume max

 

[Roen]

@AndrewMason

Does the op-amp balanced configuration (otherwise known as BAL -> SE -> BAL) introduce common mode ground noise along this chain?

 

[pkane]

First, Sorry for my poor English, I hope I can explain it clearly.

I don't know if I am right, you can judge it yourself

Let me make some assumption:
1. The distortion and noise will come when Analog signal goes from DAC to AMP, let's make a assumption that the distortion and noise is 1db
2. the dac outputs 50db at max volume
3. I hope that my headphone gives me 100db

then:
1. set pc volume 10db, because I wish get 100db, the amp will amplify 10 times, meanwhile the distortion and noise will be amplified 10 times , my headphone will get 10db distortion and noise
2. set pc vulume 50db, the amp will amplify 2 times, so my headphone will get 2db distortion and noise

so it is better to set pc and dac volume max

That makes sense. A lower output at the DAC will require a greater gain at the amplifier to get to the same output level. So, noise level will be increased as the result, while the signal will remain the same.

 

[wyrainer]

That makes sense. A lower output at the DAC will require a greater gain at the amplifier to get to the same output level. So, noise level will be increased as the result, while the signal will remain the same.

thank you，that is what I want to say...

 

[ReAlien]

If your DAC has balanced output, then my preference would be the 789 as that deals with ground loops.

Is this really some audible problem? I mentioned making a balanced cable from my portable DAC (Opus 1s — in fact, it has such type of connection mentioned in the manual) to amp connection at some other forum and got laughed at by many members, they say it's a marketing thing and I should not bother...