Review and Measurements of Massdrop THX AAA 789 Amp

[Andi]

It’s definitely worth considering the SU-8. Other than the mid ESS hump it’s better in most if not all categories of testing over the DX3 Pro.
Subjectively I find it lacking in the bass region, no matter the speaker or headphone, just my option though.

Thanks for the info, good to know since I normally like it a bit bassier.

 

[magor34]

Thanks for the info, good to know since I normally like it a bit bassier.

That's what the ASR forums are for . I was hoping to see the new SMSL M300 put through its paces here, since it should be an upgrade over the SU-8 if the specs are legit. It has one of the high end AKM DAC chips so the ESS hump is not an issue here. But it appears no one has bought it or rather wanna send it to Amir for testing. Which is a shame.

 

[superdeedapper]

You measure this amp as not clipping until 3.5w with the balanced out. According to the published specs of this amp on massdrop, the amp can output 6w (above the 3.5w clipping threshold) into 16 ohms with the balanced out. Does that mean that I will experience noise/clipping if I use 16 ohm headphones with the balanced out? is that how that works, or is it the volume knob that increases the wattage, and I don't get clipping until I turn the volume past a certain point?

 

[elira]

You measure this amp as not clipping until 3.5w with the balanced out. According to the published specs of this amp on massdrop, the amp can output 6w (above the 3.5w clipping threshold) into 16 ohms with the balanced out. Does that mean that I will experience noise/clipping if I use 16 ohm headphones with the balanced out? is that how that works, or is it the volume knob that increases the wattage, and I don't get clipping until I turn the volume past a certain point?

More volume more power. Lower volumes require less power. So it depends what sensitivity the headphones have, but for most headphones 6W will be extremely loud.

 

[Roen]

You measure this amp as not clipping until 3.5w with the balanced out. According to the published specs of this amp on massdrop, the amp can output 6w (above the 3.5w clipping threshold) into 16 ohms with the balanced out. Does that mean that I will experience noise/clipping if I use 16 ohm headphones with the balanced out? is that how that works, or is it the volume knob that increases the wattage, and I don't get clipping until I turn the volume past a certain point?

Check your sensitivity numbers, but I think you'll be deaf before the amp clips.

 

[magor34]

Any guesses as too when the amp will hit 4 volts with an 8 Ohm planar?

 

[JohnYang1997]

Any guesses as too when the amp will hit 4 volts with an 8 Ohm planar?

RMS?

 

[JohnYang1997]

4Vrms or 5.6V peak to 8ohm isn't that hard. It requires 0.7A. One tpa6120 per channel should be capable of pushing it. Not continuous for long time tho, the power dissipation is very high at high current. A massive parallel design will certainly drive 8 ohm load with ease even 4 ohm. 10Vrms into 4ohm is possible.

 

[magor34]

4Vrms or 5.6V peak to 8ohm isn't that hard. It requires 0.7A. One tpa6120 per channel should be capable of pushing it. Not continuous for long time tho, the power dissipation is very high at high current. A massive parallel design will certainly drive 8 ohm load with ease even 4 ohm. 10Vrms into 4ohm is possible.

I'm asking cause it will void the warranty if it hits 4 volts (at 8 Ohms). I'm assuming it's 4Vrms and not peak voltage.

 

[JohnYang1997]

I'm asking cause it will void the warranty if it hits 4 volts (at 8 Ohms). I'm assuming it's 4Vrms and not peak voltage.

What was the question again? Were you asking when an amp can drive such load or in what condition will 789 drive 4v into 8ohm?
I was answering the first one.
If you were asking the second. In short, normal listening is safe, you won't see any problem. Your ear would bleed before the amp stress. It requires 4v input + 0db gain or 2v input + 6db gain to achieve 4v output.

 

[magor34]

Second (I think ), how much gain/volume it'll take to hit 4v. I'll be using XLR input and 4-pin balanced output. So I'll need to use the second gain setting (+6 dB) at noon (?) in order to void warranty? I'm trying to get a grasp of what not to do, before it arrives.

 

[Veri]

I'll be using XLR input and 4-pin balanced output.

Balanced input balanced output gives ridiculous power at low knob setting. SE input balanced output is already loud, balanced in+out really won't stress the amp whatsoever.

 

[dudewithcans]

It’s definitely worth considering the SU-8. Other than the mid ESS hump it’s better in most if not all categories of testing over the DX3 Pro.
Subjectively I find it lacking in the bass region, no matter the speaker or headphone, just my option though.

i felt exactly the same, and do you know what i did? im absolutely certain (i switched multiple times and testet with DT880, even though i didnt do a blind test), that the small amout of basslack of dx3pro comes from the factory default pcm setting.
maybe its just for my headphones, thats possible.

but when i set the pcm setting from default (F3, short delay, sharp rolloff) to F5 (super slow rolloff) i FELT like the bass ist more extended, and overall it just sounded CORRECT to me.

and im certain this was not some kind of psychological "result equals expectation" kind of thing, because I HAD NO EXPECTATIONS for any of the pcm settings in the first place.

donno, maybe you give this a try?

on the other hand i have no technical explanation for this , so this all might just be in my head... but then again, why would you have those settings in the first place if it would make no difference?

 

[Veri]

on the other hand i have no technical explanation for this , so this all might just be in my head... but then again, why would you have those settings in the first place if it would make no difference?

To please the people that think they hear a difference

Or on a more serious note, to appease both the linear phase and minimum phase camp. Most consumers obviously don't really care about this.

 

[Viper Necklampy]

To please the people that think they hear a difference

Or on a more serious note, to appease both the linear phase and minimum phase camp. Most consumers obviously don't really care about this.

I Read in a forum both minimum are wrong technically, but feel more 'extended' in time and that's why he feel it kinda bassy .. Linear phase is the way to be 'right', or maybe even others.. I love Linear phase fast roll off with D50 dac and headphones.. Think i will use that even in my next D1 Dac, which is default filter, i feel with enhanced dynamic even with sharp filter hurt even less the hearing, maybe a little dry or fast, yeah.. Seriously, try it If you can hear the difference even in hi-res upsampled ofc, i can, and don't tell me i can't, I can easily hear it every time i switch

 

[dudewithcans]

well honestly i dont care at all at this point. like i said, i had no expectations whatsoever when switching these settings, so there is no such thing as preoccupation towards or against one of those pcm settings... or the presumtion, that switching between those settings does anything audible at all.

i just switched it, and i think i can hear it sounds different regarding bass extention with the DT880. solid audibly on certain tracks, not by a hair. thats fine with me. thats the reason why i pointed to that settings in the first place.

in the end, im not a "believer" - i like numbers. i like results. i like proof. so, im more biased towards a device that measures good then towards a device thats not measured at all (or at least incomprehensible for me) and someone TELLS ME its good (technically).
but regarding those pcm settings, i for myself really dont need to have scientifically proven results what those settings actually do or dont do, or whether they do something at all.

what would be REALLY FUN though, to have a blind test regarding this - maybe, just for fun, i explain to my girlfriend how to put the dx3pro in program mode and switch between F3 and F5 (well, ok, this part wont be to much fun at all), and i try to find out which setting it is...

cheers.

To please the people that think they hear a difference

works. im pleased.

 

[magor34]

i felt exactly the same, and do you know what i did? im absolutely certain

I was referring to the lack of bass in the SMSL SU-8 not the Topping DX3 Pro.
I’ll post an update when my THX 789 arrives.

 

[magor34]

I have not tried another DAC with my 789 yet but since it's perfectly flat in the FR, any perceived colouring should be down to the source, in my case the SU-8.
When testing the same headphones on various mobile devices with their internal DACs, Lightning-adapters and the Dragonfly Red, they are not even close to the SU-8/THX 789 combo in SNR and DR. BUT there's slightly more warmth on all of them, anywhere between 0.5-2.0 dB, not sure.
The only headphone I own with a warm signature is the Sony MDR-1A, with those there's not much of an issue but with any of my neutral or slightly bright pairs I must say the lack of bass is very noticeable.
I will try to post a follow up side by side DAC test with level matching if there's any interest in that.

 

[magor34]

Balanced input balanced output gives ridiculous power at low knob setting. SE input balanced output is already loud, balanced in+out really won't stress the amp whatsoever.

Yeah, I'll take your word for it. But with it being a planar, it requires a lot more power than a dynamic headphone, no? A reviewer (Max Settings) stated 8-12 Ohm range on the Verum Audio 1's, with the official specs at 8 (12 at prototype stage). So cable impedance might have some impact here.
I'm actually thinking about getting a short 4-pin XLR cable and re-solder it myself with some resistors and double the impedance. I'm actually scared to use my planars with the 789, I really don't want to break anything. I haven't even tried the Balanced out yet, only SE out with Balanced input.
4 volt peak is the warranty limit btw. Not RMS.

 

[Music1969]

Hi @amirm

What's the reason you didn't test distortion with real headphone loads with the THX 789, like you previously did with Chord Hugo2's headphone output?

Same with the JDS Lab Atom headphone amp measurements too.

Is there something faulty with the Hugo2 test you did?