• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required as is 20 years of participation in forums (not all true). Come here to have fun, be ready to be teased and not take online life too seriously. We now measure and review equipment for free! Click here for details.

Schiit Magni 3+ and Heresy Headphone Amp Reviews

solderdude

Major Contributor
Joined
Jul 21, 2018
Messages
4,824
Likes
8,865
Location
The Neverlands
why would sensitivity have anything to do with mOhm ?
Only varying impedances is the reason for varying FR.
 

Tks

Major Contributor
Joined
Apr 1, 2019
Messages
1,447
Likes
1,741
Location
NYC
Not necessarily. Actually I was looking at the guy who said a few cm of wires would degrade noise performance.
Then you used benchmark hpa4 to support it.
To me it's the fault of lacking information. I don't blame you guys. I didn't want to write about 500 words just to explain how this works so I just posted that.
Yeah, I wouldn't be asking basic questions on if the stepped attentuator I posted could be a simple swap and drop-in method without ruining an amp, I sure as shit wouldn't know what effects and differences there are between it, and passive pots (aside from passive ones being scratchy and sometimes bad channel balance). Just trying to learn a bit about these things and why they cost so much (I'm guessing manufacturing isn't simple).
 

solderdude

Major Contributor
Joined
Jul 21, 2018
Messages
4,824
Likes
8,865
Location
The Neverlands
But I do value channel matching, and low-level volume control.
Yes that will be paramount when you want to listen to sensitive IEMs at low volume. The stepper may have excellent L-R balance down below but usually also has bigger steps so fine adjusting the volume will be harder with the stepper most likely.
A relay attenuator may work better here.

IamHolland mentioned he would need to see measurements of someone taking the Heresy and put in a stepper in order to believe that a stepper with short wires would not be equally noisy. The wiring is inside a metal enclosure so screened anyway. No antenna there. Besides one could use screened wire. The extra solderjoint won't do that much. He should worry more about the hundreds of solder joints in the stepper.
Anyway.. I am not going to provide evidence as I don't have an AP nor a Heresy nor do I need one nor am I going to fit a stepper in there.
An RK27 would suffice.
 

JohnYang1997

Major Contributor
Joined
Dec 28, 2018
Messages
2,893
Likes
3,294
Location
China
Alright here we go.

Firstly THX 888 is said to have 1.6uV of noise. If you do calculation it's accurately 89dB of SNR in 50mV test.

Then 789 is a different circuit than 888 and has less noise. Its noise is limited by opa1602 and resistance noise as well. Hence 887 came to improve over 789 using 1612 and lower resistor values to reduce noise. It's great that the 887 has better noise performance than HPA4 but that's why.

RME ADI2 DAC's topology is opamps in parallel(heresy is very similar but more details) so inherently the noise can be very low. Plus the fact the RME really really knows what they are doing they don't put large resistors in the signal path or feedback path then the noise can be lower than 0.6uV. Which gives 98dB. Then plus noise of AP you get 93dB.

If you have 0.2uV or lower noise then the number can be even higher but only at this level it's limited by AP. under 90dB means the output stage is noisy.

Now onto the issue of changing POT and the extra 6cm wire. As we see in these amps, the mains noise is around or under 200nV. There's an issue of measuring headphone amplifier using BNC port but I won't comment on that now maybe other day. If you change a pot the only possible degradation in noise performance is interference if same value of pot is used. Now I gotta say, if you get interference by doing this, the layout has some issue. I'm not sure whether schiit's layout is on point at least it's quite low at the moment. But output stage noise is not going to be affected. As well as someone commenting on connect the shell to gnd, it is not the proper way to do it. These amps are isolated from earth, connecting shell to gnd won't do anything it's the same ground. I'm having 100nV of mains noise without it, it's not even put in a case. Hence, changing a pot with a few wires shouldn't degrade performance if the amp is well designed.
 

Tks

Major Contributor
Joined
Apr 1, 2019
Messages
1,447
Likes
1,741
Location
NYC
Yes that will be paramount when you want to listen to sensitive IEMs at low volume. The stepper may have excellent L-R balance down below but usually also has bigger steps so fine adjusting the volume will be harder with the stepper most likely.
A relay attenuator may work better here.

IamHolland mentioned he would need to see measurements of someone taking the Heresy and put in a stepper in order to believe that a stepper with short wires would not be equally noisy. The wiring is inside a metal enclosure so screened anyway. No antenna there. Besides one could use screened wire. The extra solderjoint won't do that much. He should worry more about the hundreds of solder joints in the stepper.
Anyway.. I am not going to provide evidence as I don't have an AP nor a Heresy nor do I need one nor am I going to fit a stepper in there.
An RK27 would suffice.
And quickly if you can for me, so I don't spam up this thread anymore. Any generalized differences I should know about relay attenuators vs passive pots vs stepped attenuators?

Also you mentioned steppers will be tough, but the one on that site I posted seems to be capable in one variant of 1dB increments? That sounds fine enough for me, seeing as how each step down that low wouldn't be too noticable to begin with?

EDIT: Oh an thank you as well @JohnYang1997 That was a great overview
 

solderdude

Major Contributor
Joined
Jul 21, 2018
Messages
4,824
Likes
8,865
Location
The Neverlands
And quickly if you can for me, so I don't spam up this thread anymore. Any generalized differences I should know about relay attenuators vs passive pots vs stepped attenuators?
Relay volume control takes up a lot of PCB space, needs lots of relays to make small steps, needs logic and a power supply, can be remote controlled easily. Good L-R balance. Volume control 'taper' can be shaped easily by the designer. Wide volume adjustment range is possible but requires more relays.

stepped attenuator. Limited to the amount of physical steps. Volume control 'taper' can be shaped easily by the designer. L-R tracking is excellent. Does not need power.

Potmeter has a choice of fixed tapers, can be small in size, cheap, quality can be suspect. Prone to scratchy behaviour. Does not need power nor logic.
potmeters with a motor fitted can add remote control.

LDR needs power, has distortion issues that differ depending on volpot setting. Needs adjustment for proper L-R tracking. Ages so may need adjustment after a while. The larger the input voltage the worse the distortion becomes. 'taper' can be shaped by the control voltage.

Chip volume control (analog) is a bit noisier and has pre-defined steps. Requires logic and power supply. Limited input voltage range. Easy to implement remote control. Usually very good L-R balance and adjustment range/step size. Fixed 'taper'. Could add some distortion (depending on chip).

Volume control in the digital domain depends on the implementation but can be best in L-R tracking. step sizes can be really small and 'taper' can be anything the engineer wants it to be (soft or firmware based)
 
Last edited:

Asylum Seeker

Active Member
Joined
Feb 12, 2020
Messages
287
Likes
152
Location
The Island of Bali
If you look at the graph, you will see that the imbalance is variable based on position of the volume control. Is it not constant. You could set the volume to max and then change the input level.
The channel imbalance is a small price to pay for such exceptional performance and price, no?
 
Joined
Feb 7, 2020
Messages
36
Likes
10
I have a dac that can not to drive my headphone akg k712 pro, wath's the best amp for me : schiit heresy , magni 3+,topping a50 or smsl 200 sp ?
 
Joined
Feb 12, 2020
Messages
22
Likes
11
Location
near Eindhoven, Nederland
I have understood that the output opamps in the Heresy are the same as the ones used in the RME ADI-2 DAC (I have). In the review of the RME it is shown that the THD increases with a real headphone compared to a resistor (https://www.audiosciencereview.com/...and-measurements-of-rme-adi-2-dac.2582/page-5, #86 and #96). I don't know about the differences between the rest of the circuitry of the RME and the Heresy, but, it might be that the Heresy has the same problem as the RME!?

And the Magni 3+?
 

solderdude

Major Contributor
Joined
Jul 21, 2018
Messages
4,824
Likes
8,865
Location
The Neverlands
The output opamps may well be the same, the output configuration will not be.
I assume the Heresy uses much more opamps in parallel but with series resistors so a not close to 0 output impedance.

The issue of the increased distortion numbers need to be seen in a certain light.
The plots are made with 0.5W into 60 Ohm. When you pump that into the V6 it will be pumping out 132dB SPL.
You have other problems than a -60dB distortion in the bass electrically speaking.
That headphone will behave non linear and have much higher distortion than the -60dB.

The second issue is back EMF. This is generated by V6 and is a voltage with a 80 Ohm source in this case.
That source thus creates a current (the damping current) which is mostly determined by the source (80 Ohm) and hardly changes when the output R is 0.1 Ohm or 1.0 Ohm.

BUT here is the problem of the measured rise in distortion.
When an 80 Ohm source (the MDR V6 for instance) is driven with 500mW = 6.3V and the impedance rise = 50% (from 80 to 120 Ohm) says that when the Ohmic resistance is 80 Ohm there is 80mA needed. For 120 Ohm 50mA (all rounded off) so 30mA is back EMF.
30mA across 0.1 Ohm = 3mV.
or in case of an 1 Ohm resistor 30mV.

With an output signal of 6.3V there will be 3mV (distortion) in the low output R amp. This is -66dB
30mV in the higher output R amp which is -46dB measured distortion.
The lower the output R the lower the distortion seems to be. However, the damping current in both cases is 30mA so that doesn't change.

The Heresy has 0.2 Ohm output R and the Magni 3+ = 0.25 Ohm. The effect of the seemingly increase in THD will thus be the same.
Of course, there is also the fact that the back EMF is not 180 degrees shifted in phase (only at the imp peak) so that too will have some effect depending on overall feedback and ability of the amp to always have the same output R regardless of current phase.

So moral of the story is. No audible consequences and 3+ and Heresy will be a close call in this aspect below audible levels.
There is no practical problem here. There is more of a measurement method issue (IMHO)
 
Last edited:
Joined
Feb 12, 2020
Messages
22
Likes
11
Location
near Eindhoven, Nederland
ok, it looks like that I understand your point, but:
In the RME topic it is writen that Benchmark Media "solves" (not reduces) the problem with a RC in the feedback.
RME does not do that.
Does Schiit use an RC? It is a newer design.
So it is not only output impedance that has influence on this behaviour.

And about the Magni 3+:
Around the year 1990 I measured multiple loudspeaker amplifiers on this behaviour. Unfortunately with not that good measurement equipment.
It looked like (= I am not sure) that the amplifiers with bigger problems had problems with delivering negative/out of phase current.
And most class-A (and high biased) amplifiers behaved better (the Musical Fidelity A1 went BANG!).
So, back to the Magni 3+: it is discrete, maybe biased higher, maybe a more stable feedback loop. I am interested to know.
 

solderdude

Major Contributor
Joined
Jul 21, 2018
Messages
4,824
Likes
8,865
Location
The Neverlands
With speakers currents are much higher and AB idle currents are reached relatively sooner.

You can test this by not putting any signal on the input of the amp and injecting a tone via a resistor on the output combined with say a 32 Ohm resistive load.
Then put some DC on the input of the amp (both Magni's are DC coupled I believe) and measure again as the output stage will be in class-A and measure the difference on the output of the amp.

That should tell you the amp's reaction to back EMF around the 0V line and is dependent on the feedback loop and idle current perhaps.

Don't have either amp so cannot test this nor do I know the exact schematics so it can be simulated (one could reverse engineer the schematic)

Then you could as yourself if such low amounts of distortion are audible and one could ask themselves if the measured voltage across the output (which is the diff signal that is used to generate distortion numbers) also is present in the headphone itself or is just a measurement artifact.
After all, headphones will have higher distortion than the numbers we are talking about at the reported levels (130dB and 0.5W).

Interesting for the measurement crowd, possibly not to explain reported audible differences that seem to disappear in properly conducted blind tests.
 
Joined
Jun 20, 2019
Messages
30
Likes
46
Got the Magni Heresy today. It's a very good amp, I like it a lot. Volume control knob is too close to the headphone jack as Amir already observed.
Now the major drawback: I got no roll of toilet paper with it.
Bummer! :confused::p

P.S. We need a new fancy facepalm smiley, it's ugly in comparison to the other ones...
 

Tks

Major Contributor
Joined
Apr 1, 2019
Messages
1,447
Likes
1,741
Location
NYC
Got the Magni Heresy today. It's a very good amp, I like it a lot. Volume control knob is too close to the headphone jack as Amir already observed.
Now the major drawback: I got no roll of toilet paper with it.
Bummer! :confused::p

P.S. We need a new fancy facepalm smiley, it's ugly in comparison to the other ones...
Snap a pic, always nice to see someone's new setup :]
 

Sgt. Ear Ache

Senior Member
Joined
Jun 18, 2019
Messages
332
Likes
552
Location
Winnipeg Canada
Ordered a Heresy this morning. I'd been on the verge of getting one of the THX amps for the past week - but wasn't really even aware of the Heresy till yesterday. I have a Liquid Spark already, and I doubt I'll really be able to hear any sort of audible difference, so for under $200 Canadian (had to order it from Amazon in the US as Amazon.ca doesn't have the Heresy for some reason) It's a lighter hit to the pocket (even the least expensive THX amp is still more than $100 higher in price) and appears to be pretty close to endgame in performance for me...hard to resist for the price.

it's pretty crazy what a couple hunny (cdn) can get now-a-days for headphone amp performance.
 
Last edited:

Sgt. Ear Ache

Senior Member
Joined
Jun 18, 2019
Messages
332
Likes
552
Location
Winnipeg Canada
lol...man it's got to be frustrating for companies to see 1 and 2 star reviews on Amazon for their products from buyers who seem to have no idea what they are talking about. "Its OK...doesn't have much amplification power. I returned it." Er...OK then.
 
Joined
Jun 20, 2019
Messages
30
Likes
46
Yeah, well, I can understand how some of the reviewers get the impression that these first rate amps don't deliver. I was myself in such situation twelve years or so ago. Bought new headphones: AKG K701 - presumably reference grade - and then a Violectric V100 amp. Boy, was I disappointed! Sound quality was so poor, I really was devastated! Then I discovered the brand beyerdynamic and had the opportunity to listen to some of their headphones. I finally bought a pair of DT990s and boy, the difference was night and day!

I've learned the following things:

- Headphones make the most impact concerning sound stage and sonic impression. Most people that complain about poor bass or sound stage or too little power should just switch to another headphone brand (and impedance, that is) because they probably have not the right headphone gear for their needs. Those AKG cans were not to my liking. The sound stage of the DT990s kicks ass and I still own and use them (after ten years of daily use). (But I know some people here have a more profound knowledge of headphone gear – I've never listened to planar magnetics and I have no idea if those really are audiophonic heaven or if they're just marketing hype... :D)

- A good DAC is always a good investment and should be carefully chosen - don't trust your PC sound card output, it's just not delivering :eek: but don't invest too much initially, you can get good quality for little money (as Amir's tests show)

- Headphones amp is a must if you want your headphones to really shine, not as important as no.1 and 2, but you can accomodate for that by buying gear that implement both functions, DAC and headphone amp. My personal pick in this category would be the RME ADI-2 DAC right now for endgame quality gear. (I still own the Violectric HPA-V100, it's a solid amp with marvellous workmanship (probably not endgame in these days but built like a german tank!).

So these reviews stating
"Its OK...doesn't have much amplification power. I returned it."
seem to have just picked the wrong equipment for their needs and have not considered the whole audio chain.
 
Last edited:
Top Bottom