• Welcome to ASR. There are many reviews of audio hardware and expert members to help answer your questions. Click here to have your audio equipment measured for free!

SMSL H400 MK2 Preamplifier & HP Amp Review (NTTY)

Rate this Preamp & HPA

  • Terrible (*)

    Votes: 1 1.1%
  • Mediocre (**)

    Votes: 2 2.2%
  • Good (***)

    Votes: 10 11.2%
  • Excellent (****)

    Votes: 76 85.4%

  • Total voters
    89

NTTY

Major Contributor
Forum Donor
Reviewer
Joined
Apr 27, 2018
Messages
1,875
Likes
11,873
Location
Switzerland
Hello Everyone,

This is a review and detailed measurements of the SMSL H400 MK2 Preamplifier and HP Amplifier.

SMSL H400_001.jpg


It was kindly sent to me by SMSL (H400 MK2 product page). Current price is around $540.

I must say It is always great pleasure to review SMSL devices, for their measured performances, and - spoiler alert - this one will set a precedent.


Presentation

The SMSL H400 is a pure analogue preamplifier, and is the bigger brother of the H200 that I recently reviewed. It too offers only two inputs (Balanced and Single Ended), with two rear outputs (XLR and RCA) but three front headphones output (as opposed to two for the H200).

Rear inputs must be selected from the menu, making it a 2 inputs preamplifier, again one being balanced and the other unbalanced. Both rear outputs will route the signal when they are active, else it goes to front outputs, meaning there too you have to chose between line or headphones outputs.

This device integrates well with the "400" series from SMSL, but not the "200" since it is wider.

The back is as simple as it can get, and note the USB input for potential updates:

SMSL H400_002.jpg


You can get more details about this device from SMSL themselves who announced the release of the H400 MK2 here on ASR.


User experience

The interface is simple and nicer than the one of the H200. It serves to select the input (RCA or XLR), and outputs (rear or front). There's one addition and that is the gain which is selectable between High/Mid/Low and it impacts rear outputs too (same as the H200).

There is a remote control that repeats all functions available from the front knob, but I did not use it.

Basically, we have only two analogue inputs, one Balanced and one Single Ended. This is not much, but in a digital world, the DAC is likely to be the real source selector. No phono input here.

I'll repeat the same comments that I made when reviewing the H200: Looking at it from far away, it looks like you get only a volume knob, especially if you think about it as a preamplifier. But it is a little more than that. To start with, the analog volume is digitally controlled to offer perfect channel balance, and a significant part of a preamplifier’s performances rely on the quality of the volume knob. The big difference between the H400 MK2 and the H200 is the internal power supply and of course max output power too which is 4 times more (balanced) than what the H200 can deliver (a crazy 18W into 16ohms for the H400 MKII vs 4W for the H200).

There are 3 selectable gains that influence the rear outputs too. You are used to the "Low" gain setting which is useful for sensitive IEMs to increase the dynamic range by lowering the internal gain of the amplifier. It means you'll need to crank up the volume by several notches when in low gain, and that's the purpose so as to increase the DR.
I measured the real gains (from XLR output) as below:
  • Low Gain : -5.37dB gain (claimed -5.5dB)
  • Medium Gain : 5.93dB gain (claimed 6dB)
  • High Gain : 19.74dB gain (claimed 19.7dB)
Of course RCA output is 6dB gain less, and that means with "Medium Gain", from RCA input to outputs, basically no gain (0dB) is applied, ie it's unity gain.

The reason I'm mentioning this, in the user experience section of the review, is because it's practical to get selectable gains at the output. Indeed, and similar to the IEM reasoning for wanting low gain, I do like to have high gain at the output so I can use older devices such as a Cassette Deck but also some older CD Players (HDCD ones are often limited to 1Vrms when not playing a HDCD disc). These older devices can go way below our standard 2Vrms output, let alone 4Vrms from XLR, but modern amplifiers are in demand of higher input voltages. So having a "high" 14dB (RCA) or 20dB (XLR) gain at the output of the preamp will help drive an amplifier to its max power with low output voltage sources.

The "Medium Gain" choice helps with higher output voltages sources and, with a gain of 6dB, will double the output voltage to feed an amplifier. I think this is the most general use case with modern DACs and amplifiers. Theoretically, on top of that, we get here a high precision volume attenuation that should beat the one of a DAC done in digital domain before conversion. I'll report on that.

The volume display does not show the standard attenuation in dB but a rather unusual % of maximum, from 0 to 99%, like the H200. I could hear some light relay sound when going one level up or down.


SMSL H400 MK2 - Measurements (Preamplifier)

All measurements were performed with an E1DA Cosmos ADCiso (Grade 0), and the Cosmos Scaler (100kohms from unbalanced input) + Cosmos APU, for capturing analog outputs of the SMSL. I used a high performance DAC, the SMSL DO400, as the generator with REW Software from a laptop. All these devices run on batteries, except the DO400, to prevent as much as possible any ground loops.

Both XLR and RCA outputs of the H400 MKII are non inverting, and Left & Right channels are perfectly balanced at any volume (very good).

----

Let's tart with the very standard SINAD measurement (or THD+N in my case) with 1kHz sine input at 4Vrms input, with the H400 MK2 set at medium gain and volume at maximum (99%):

SMSL H400_1kHz-4Vrms_LR_XLR.jpg


Oh, yeah, this is a very impressive 125dB SINAD. This result is driven by noise, not distortion that is vanishingly low.

Pushing it to 5Vrms input (the max I can get from the DO400) improved the SINAD by 2dB:

SMSL H400_1kHz-5Vrms_LR_XLR.jpg


Oh wow, 127dB SINAD, that is crazy good.

----

Wait a minute... I need to process this.

Actually, and knowing that the DO400 (that I used to feed the H400 MK2) can do 128dB SINAD at best (at 5Vrms output), all driven by noise, not distortion (way below at -140dBr), if the H400 MK2 would be at the same 128dB SINAD noise driven, the sum of the two would be 3dB higher (ie 125dB SINAD). That is because distortion and noise add up from one device to the other.

If we don't have this level of SINAD here, it means the H400 MK2 is incredibly quiet. Indeed, for me to measure that 127dB SINAD, the maths say that the H400 MK2 must achieve a SINAD of almost 134dB!!!

There's no way for me to directly measure that performance since no DACs I could use as a generator currently exist to feed the H400 MK2. But this indirect measurement is reliable.

Of all the devices that I reviewed, this one sets a precedent of performance. This is much better than the H200 I previously reviewed, although being extremely good too. It means that the analog input/output stages of the H400 MK2 are flawless, way beyond what can be measured. This is a true pleasure to witness.


----

Now, how can I continue this review, knowing that I'm likely to either hit the max performance of the DO400 or my analyzer? But let's go for it anyways.

And first let's go back to some more reasonable / real life case, starting with unity gain at 4Vrms input/output (volume set to 92%):

SMSL H400_1kHz-4Vrms_UnitiyGain_LR_XLR.jpg


Unity gain here means that the volume is attenuated by 6dB, but we did not lose 6dB of SINAD, only 2.3. Note the harmonic distortion at -135dBr max...

Let's try even more realistic case of 1Vrms input and 2Vrms output:

SMSL H400_1kHz-1Vrms_LR_XLR.jpg


That is 12dB less output from the SMSL DO400, and at this attenuation, its performances are down to 116dB SINAD. So, we lose only 0.2dB by going through the H400 MK2, when its volume it set to maximum! Did anyone mention transparency?


----

With RCA input/output, and again using the DO400 as the generator (122dB SINAD), I get 121.5dB SINAD at the output of the H400 MK2, meaning again almost no loss.


----

The one key element of this preamplifier is its volume control. Now that I reported a crazy good performance at full output (I mean 133dB+ SINAD!), time has come to push the H400 MK2 into volume attenuation.

I already did the same with the H200, so I'm repeating the same test: What I did is that I compared lowering the volume from the SMSL DO400 (which has a digital volume attenuation via its high end ES9039MSPRO) and attenuating the same way via the H400 MK2. Here below are the results from XLR in/out (THD+N, THD and ENOB are provided):

1793394864234.png


As you can see, when the volume goes down, it is preferable to use the one of the H400 MK2 rather than the digital one of the DO400 (XLR). Since a preamplifier is almost always used to attenuate the signal rather than to increase the gain, I think the above is very good. It is comparable with the results of the H200, only a little better at higher output levels. I guess at one point, the volume attenuation becomes the limiting factor.

The perfect level matching between the two channels at any output volume is also to be recognized.


----

Ok, back to more standard measurements, likely to be at the limit of the generator and/or my ADC.

I suppose you saw 0 PS leakage (50Hz and harmonics):

SMSL H400_PS.jpg


This is a view that Stereophile likes very much to share. Here nothing can be seen from the power supply (50Hz in Europe), in low or medium gain. No PS related spikes down to -160dBr. What a silent power supply!


----

Since I'm here, let me add the 50Hz @0dBFS into 600ohms, for compatibility with Stereophile measurements:

SMSL H400_PS_Rear-vs-Front.jpg


I've overlaid the same measurements from rear and front balanced outputs. Actually, the limiting factor here is my ADC. The H3 you see at -120dBr or so, this is the Cosmos not liking high level input at low frequency. Setting the level a little lower make that H3 disappear...


----

No surprise, the bandwidth is extremely flat (measured up to 80kHz):

SMSL H400_BW.jpg



The two channels are perfectly matched, and if you get the impression there's only one channel represented here, there are two! It is the same perfection at any volume (here it is -12dB by the way) and with unbalanced in/out too.


----

Let's continue with the famous Multitone that a lot of you, including me, like so much:

SMSL H400_MT.jpg


Roughly 22bits of distortion free range all the way through frequencies, and again knowing that the DO400 acting as the generator is also adding its own. I forgot to take note of this measurement, but from memory it's at unity gain with the H400 (Medium Gain, that I'm sure).

----

Oh, yes, another measurement that you can use to compare with other devices. It is the THD+N ratio vs frequency, with an extended bandwidth to capture all distortion and noise components up to 90kHz:

SMSL H400_THDNvsFreq_LR.jpg


See the plot and bottom results at 1kHz. This is 3dB better than the H200, and again a little more than the DO400 alone (-118dBr) but meaning that the H400 is there too, -118dB of THD+N (with noise up to 90kHz) at worst. This is also the limit of what I can reliably measure with my ADC. The rising of THD at low and high frequencies is from the ADC, not the H400...


----

The below IMD SMPTE sweep vs input level is not comparable with the same test from Amir, because it is excluding noise:

SMSL H400_IMDvsLVL_LR.jpg


I overlaid the results at unity gain and +6dB (with gain set at "Mid Gain"). It reaches a best IMD of -110dB from -25dBFS or so, meaning it "suffers" from some intermodulation distortion, at -110dBr :p


----

Let's go to the dynamic range, but should I share? Will you believe the below?

SMSL H400_DR_CCIR.jpg



SMSL H400_DR_AWeight.jpg


Speechless, I am. Let me grab a whiskey.


SMSL H400 MK2 - Measurements - Headphones

The H400 is extremely powerful (4 times more than the H200 with balanced output). But I basically measured the same performances from the front outputs, so I won't repeat what you saw before.

I can't measure its full output power, it's way too much for my interface and load. I'll still try measuring it with my usual 32ohms load, but I'm very likely to burn it doing, so...

Anyways, let me jump directly to one that Amir always publishes and that's the SNR measured at 50mVrms output (unbalanced headphone out, 2,5Vrms input from RCA of the DO400, volume of the H400 MK2 set at 37% to get 50mVrms out):

1793641958999.png


This 95.3dB SNR puts the H400 Mk2 at the top of @amirm's chart of the same measurement. It is the same performance if I use back XLR inputs, only that I need to get the volume down to 31% instead of 37%. Oddly enough, the low gain mode (need to increase the volume to 49% with RCA input of 2.5Vrms) does not improve this measurement as I see SNR of 94dB, still top of the chart though.

Using the more powerful balanced output means an increased noise floor:

SMSL H400_MG_50mVrms_FrontXLR.jpg


We still get a good 88dB SNR which puts the H400 MK2 in the comfortable zone of @amirm's chart. As a matter of facts, going low gain does not improve this measurement, it's a similar 87.6dB that I got. And this is the same I measured from the H200.


Conclusion

It is always a great pleasure for me to review an SMSL product, to witness the level of engineering

After the H200, I was eager to test the H400 MK2. I got that great opportunity only to reach the limit of what I can reliably report as pure measured performances. I loved it.

As a preamplifier, this is state of the art, a first for me. As an HP Amplifier, it's also in the best in class category, and very powerful. Basically, I doubt I measured any harmonic distortion from this device, a little of IMD yes, just enough not to get bored :)

Like with the H200, we get a super high quality analog volume control. For when listening at low or very high volume, it is the guarantee to get a real resolution difficult to equal from the best DACs, let alone amplifiers and loudspeakers/headphones.

Straight wire gain? Hell, yes!

I guess the next step for SMSL is to go down chase the Johnson noise, ie the thermal agitation of charge carriers inside an electrical conductor. You can do it guys, I trust you :cool: In the meantime, very well done on this one.

I hope you enjoyed this review and I wish you a lovely weekend.

Flo
 
Last edited:
Thanks for this review.

A good analog attenuator, like this one, has (almost) always lower noise than the DAC's digital attenuation.
The drawback is in the channels matching, which is more difficult to nail.

I'm surprised about the 50mV noise.
I expected better.

EDIT: Now you added the unbalanced 50mV measurement, which is better indeed. Thanks
 
Last edited:
Thank you NTTY for an impeccable review. In spite of the general increase in the cost of living, it makes me happy to know how $ 540 buys you today a SOTA preamplifier. I still remember my undiluted joy when, back in 1976, I listened to the Mark Levinson JC-2, and I knew it was unobtainable for me at that selling price.
 
Another really great job. Congratulations and thank you.
It shows that today it is possible to build a perfectly transparent preamplifier at a reasonable price.
However, I find the number of inputs too limited to make its use universal. Ideally, there should be twice as many. On the other hand, having a balanced input at this price is nice.
Here is a product that is technically state-of-the-art but that I wouldn't buy due to practical limitations.
 
Just wish SMSL could produce power amplifiers with similar state-of-the-art results !;)

Great measurements for DACs and Preamplifiers, now complete the line-up with great measurements for speaker amplifiers.
 
> Straight wire gain? Hell, yes!

Be careful with this often repeated analogy. It literally means that the output impedance of the source is carried through to the output of the amp. For example, consider a power amp driven by a pre-amp with an output impedance in the typical 100-600 Ohm range. If the power amp is truly a "wire with gain," its output impedance will also be 100-600 Ohm, which will make it unsuitable for effectively drive a 4-8 Ohm speaker. The amp also needs to be an output impedance converter.
 
O sinal conectado à entrada RCA pode ser transmitido pela saída XLR?
 
Thanks for this review.

A good analog attenuator, like this one, has (almost) always lower noise than the DAC's digital attenuation.
The drawback is in the channels matching, which is more difficult to nail.

I'm surprised about the 50mV noise.
I expected better.

Me too, and I tested several times to check I did not make a mistake.

In fact, If I perform the same test with the DO400 alone, that is to output 50mV, then I get 84.1dB SNR in Low Gain and 72.3dB in high gain.

Keeping the same attenuation with the DO400 but feeding the H400 and measuring at the output of front XLR from the H400, I get also 84dB in LG or MG, meaning the DO400 is the limiter.

And I performed again the same with the H400 in high gain, again at 50mVrms output (volume at 19% for an XLR input of 4Vrms), I got a very decent 82.7dB SNR. So the performance is stable accros all gains, which is good news, but 88dB SNR is the best I got in all these tests.
 
88dB SNR is the best I got in all these tests
Technically, this should be run with DAC at 0dB attenuation and attenuation done on the preamp.

Also, @amirm 's measurement is (also) done on unbalanced out - which should give better results - with a 600 ohm load.
And I think he's using CCIR-2K weighting for this one as well...

Also: depending how the DAC is built, it's possible the DO400 will give better results on 4V range, or even wirh unbalanced 2V output, for this test.
 
Last edited:
> Straight wire gain? Hell, yes!

Be careful with this often repeated analogy. It literally means that the output impedance of the source is carried through to the output of the amp. For example, consider a power amp driven by a pre-amp with an output impedance in the typical 100-600 Ohm range. If the power amp is truly a "wire with gain," its output impedance will also be 100-600 Ohm, which will make it unsuitable for effectively drive a 4-8 Ohm speaker. The amp also needs to be an output impedance converter.
Straight wire with gain: the gain makes the difference to a simple wire. It means both voltage and current gain, so the output impedance is not related to the input impedance.
 
If you dont need a HP amp its just a volume control, (no eq., source switching, phono input, ballance) ? Lower noise than a dac? I dont hear any noise from my dacs so who cares.
No thanks, Ill keep using my dacs volume control.
 
Technically, this should be run with DAC at 0dB attenuation and attenuation done on the preamp.

Also, @amirm 's measurement is (also) done on unbalanced out - which should give better results - with a 600 ohm load.
And I think he's using CCIR-2K weighting for this one as well...

Also: depending how the DAC is built, it's possible the DO400 will give better results on 4V range, or even wirh unbalanced 2V output, for this test.
Balanced 4Vrms output XLR from the DO400 into the H400 Mk2, front balanced output into 200kohms load of the Cosmos scaler, unweighted, were the parameters for that measurement. I can run the same unbalanced indeed.
 
thanks for the review, it's nice to see such a high performing device, but, as French Meloman also says, one doubt remains, why only two inputs?... the scope of a preamp is very often also that of acting as a hub for more sources. I doubt adding more inputs would raise costs dramatically. Then again, being purely analogue, it is to be assumed the digital inputs are all relegated to a dac that preceeds this preamp in the audio chain. Not that dacs have a plethora of inputs though...
I'm a little bit puzzled (but you will excuse me coming from old school hifi) how is this machine supposed to be used? what is its typical place in an audio chain?
 
Back
Top Bottom