This is a review and detailed measurements of Schiit Jotunheim DAC and headphone amplifier with its AK4490 DAC module. It is on kind loan from a member and in this configuration costs $499 plus shipping from Schiit direct. I had previously reviewed Schiit Jotunheim but that was with its multibit DAC. Some rightly objected that perhaps the performance of the unit was limited by its low performing DAC as opposed to headphone amplifier/pre-amp. In this test we remedy that two ways: 1) the included DAC uses the AKM Ak4490 DAC chip instead of multibit DAC used in the other unit and 2) I also test the pre-amp independently of the DAC.
The unit has the same bland, cost optimized packaging of other Schiit products:
The unit is mains powered so is quite hefty and sits where put. Decoding the cryptic symbols, there are USB, RCA and balanced XLR inputs. For output in addition to the latter two in the back, there are 1/4 inch and XLR headphone outputs. So from feature point of view, the Jotunheim is quite good.
As always the trick with Schiit products is how they perform and in the case of mains operated units (i.e. WITHOUT an external power supply), how safe they are. Let's get into that.
DAC Measurements
I am going to start by testing the DAC subsystem by itself using the RCA and XLR outputs. Alas, I see that they are both variable and go up to pretty elevated levels so they are really tapped from the headphone amplifier. Be careful then in turning up the volume too much if driving other audio products.
Let's start with our dashboard of RCA outputs with levels set to the typical 2 volt RMS:
Right away we see a big improvement in performance with SINAD (signal over noise and distortion) jumping up a whopping 10 dB over its multibit version. Unfortunately at 96 dB, it is not going to win any races:
It is squarely in tier 3 of performance (tier 1 being the best). So not broken but also not designed with excellence in mind.
We see that there are mains related (60 Hz and multiples) visible in our FFT spectrum. This is typical of RCA outputs in many products due to poor nature of unbalanced audio interconnects. Let's switch to XLR/Balanced outputs and see if it is still there:
Balanced inputs don't use the ground as a reference so usually they are immune to mains related currents between equipment that upsets RCA connected gear. Yet, we still see that. Let's do our prior experiment of grounding the lid of the unit with a wire to the safety mains:
As I suspected, the channel in blue is completely good now. And what remains of 60 Hz in red channel is also reduced.
As a final confirmation, let's just touch the volume knob without the grounding wire and see what happens:
Wow, see how much the 60 Hz hump increased in level. This is the classic symptom of ungrounded Schiit Jotunheim. See this thread where I explain the issue in more detail and show how to fix it: https://www.audiosciencereview.com/...otunheim-headphone-amplifier.3733/#post-89406
Note that due to lack of sensitivity of our hearing in low frequency, what I have shown is inaudible. But that is in my situation. In other cases the rise may be substantially higher leading to hearing buzz and hum. The bigger issue beyond fidelity is safety though. A metal enclosure needs to be full grounded to mains safety pin and in this case, it is not. The case is insulated due to finish on the top lid so it does not make electrical contact with the safety ground pin on the IEC outlet. Anyway, let's get back to measurements.
Jitter and noise shows elevated levels of the latter and some spurious responses:
The spikes are all too low to be audible.
Dynamic range misses the 109 dB specification:
Given the variable nature of the output, one probably could play with the volume control to get the specified rating.
The higher noise floor impacts the intermodulation+noise test naturally:
Here is the nature of THD+N with respect to source frequency:
Interesting that it peaks from 2 to 7 kHz. Our hearing is most sensitive in 2 to 5 kHz so it is unfortunate that it wanted to get more distorted there than higher up in frequency range.
Here is our 32-tone test track to simulate "music":
I did not have the chassis grounded with my jumper cable so we see the mains spikes showing up in the spectrum.
Linearity shows lack of perfection but is transparent for CD music:
Finally here is the response of the reconstruction filter measured using white noise:
Looks like a very slow filter is used which provides way too little attenuation at the "Nyquist" frequency of 22.05 kHz for CD music at 44.1 kHz sampling rate. We only get 11 dB of attenuation where the theory demands infinite level.
Headphone Output Measurements
Let's start with driving the 1/4 inch headphone jack using the USB input (i.e. going through the internal DAC):
Power level is very good but despite the much higher price, the Jotunheim fails to outperform the Topping DX3 Pro on that front. Noise is naturally higher as measured before.
Let's measure the performance by feeding the unit from its analog input:
We get good bit of reduction of noise there. Power output increases some because I am able to push the amplifier section harder than internal DAC. But then again it gets distorted so no sense in doing that. Also, once the amp distorts, it doesn't matter if you use the DAC or not. They both meet at the same place/point as you see in the graphs.
Switching the output load to 33 ohm with analog (RCA) input gets us the typical distortion curves of Schiit headphone amplifiers when stressed:
We see that two stage rise in distortion. First one is some kind of non-linearity which should have been dealt with using better circuit design. The second sudden peak is the unit running out of current and sharply distorting.
As I have shown in the graph in green, the little $99 JDS Labs Atom blows away the Schiit Jotunheim out of the water. It is essentially distortion free until it too runs out of current. That happens at 1 watts with the Atom whereas the distortion starts to rise in Jotunheim at just 50 or so miliwatts.
Since there is XLR (incorrectly called "balanced") outputs, let's measure that using 50 ohm load. And for variety use XLR input:
Same problem as before manifests itself in very early rise in distortion. As such, our other state-of-the-art, the Massdrop THX 789 leaves the Jotunheim in the dust, produce more power at a fraction of the distortion.
Finally, here is the test of channel balance versus position of volume control:
Due to use of analog control we get inaccuracies between channels. Until you get to low volumes, it is not too bad at 0.4 dB.
Listening Tests
I did not bother with my usual level matched blind AB tests. Instead I just played some music and turned up the volume to detect distortion. WIth Sennheiser HD-650 there was fair bit of power but once you went past the 2:00 o'clock position on the volume control, you could easily hear the distortion setting in. Similar situation existed with Hifiman HE-400i. Neither could pass my test of loudness without distortion which admittedly means pretty high SPLs (sound levels).
Conclusions
Use of traditional delta-sigma DAC chip significantly improves the performance of Schiit Jotunheim versus the multibit version tested before. So if you are going to go this route, that would be my strongest recommendation.
The world has moved on since the Jotunheim was designed and today we have stellar headphone amplifiers in the form of JDS Labs Atom and Massdrop THX AAA 789. Few amplifiers can get close to the performance of these units much less a Schiit headphone amplifier. So while the Jotunheim can be an enjoyable device to use versus lower powered amplifiers, it just doesn't rise up to level of objective performance I like to see in this day and age and at that price.
On the plus side though, this is a very full featured all-in-one unit. Right now I can't think of any other device I have reviewed with the same functionality.
Most important concern here is electrical safety. I continue to be dismayed that Schiit has not recalled all of these units to properly repair them after confirming that the top is supposed to be grounded. On that basis alone, I cannot recommend Schiit Jotunheim in any configuration for purchase.
------------
As always, questions, comments, recommendations, etc. are welcome.
They say you should save money for a rainy day. It rains for about 6 months here so I need plenty of money!!! Please consider donating funds using:
Patreon: https://www.patreon.com/audiosciencereview), or
upgrading your membership here though Paypal (https://audiosciencereview.com/foru...eview-and-measurements.2164/page-3#post-59054).
The unit has the same bland, cost optimized packaging of other Schiit products:
The unit is mains powered so is quite hefty and sits where put. Decoding the cryptic symbols, there are USB, RCA and balanced XLR inputs. For output in addition to the latter two in the back, there are 1/4 inch and XLR headphone outputs. So from feature point of view, the Jotunheim is quite good.
As always the trick with Schiit products is how they perform and in the case of mains operated units (i.e. WITHOUT an external power supply), how safe they are. Let's get into that.
DAC Measurements
I am going to start by testing the DAC subsystem by itself using the RCA and XLR outputs. Alas, I see that they are both variable and go up to pretty elevated levels so they are really tapped from the headphone amplifier. Be careful then in turning up the volume too much if driving other audio products.
Let's start with our dashboard of RCA outputs with levels set to the typical 2 volt RMS:
Right away we see a big improvement in performance with SINAD (signal over noise and distortion) jumping up a whopping 10 dB over its multibit version. Unfortunately at 96 dB, it is not going to win any races:
It is squarely in tier 3 of performance (tier 1 being the best). So not broken but also not designed with excellence in mind.
We see that there are mains related (60 Hz and multiples) visible in our FFT spectrum. This is typical of RCA outputs in many products due to poor nature of unbalanced audio interconnects. Let's switch to XLR/Balanced outputs and see if it is still there:
Balanced inputs don't use the ground as a reference so usually they are immune to mains related currents between equipment that upsets RCA connected gear. Yet, we still see that. Let's do our prior experiment of grounding the lid of the unit with a wire to the safety mains:
As I suspected, the channel in blue is completely good now. And what remains of 60 Hz in red channel is also reduced.
As a final confirmation, let's just touch the volume knob without the grounding wire and see what happens:
Wow, see how much the 60 Hz hump increased in level. This is the classic symptom of ungrounded Schiit Jotunheim. See this thread where I explain the issue in more detail and show how to fix it: https://www.audiosciencereview.com/...otunheim-headphone-amplifier.3733/#post-89406
Note that due to lack of sensitivity of our hearing in low frequency, what I have shown is inaudible. But that is in my situation. In other cases the rise may be substantially higher leading to hearing buzz and hum. The bigger issue beyond fidelity is safety though. A metal enclosure needs to be full grounded to mains safety pin and in this case, it is not. The case is insulated due to finish on the top lid so it does not make electrical contact with the safety ground pin on the IEC outlet. Anyway, let's get back to measurements.
Jitter and noise shows elevated levels of the latter and some spurious responses:
The spikes are all too low to be audible.
Dynamic range misses the 109 dB specification:
Given the variable nature of the output, one probably could play with the volume control to get the specified rating.
The higher noise floor impacts the intermodulation+noise test naturally:
Here is the nature of THD+N with respect to source frequency:
Interesting that it peaks from 2 to 7 kHz. Our hearing is most sensitive in 2 to 5 kHz so it is unfortunate that it wanted to get more distorted there than higher up in frequency range.
Here is our 32-tone test track to simulate "music":
I did not have the chassis grounded with my jumper cable so we see the mains spikes showing up in the spectrum.
Linearity shows lack of perfection but is transparent for CD music:
Finally here is the response of the reconstruction filter measured using white noise:
Looks like a very slow filter is used which provides way too little attenuation at the "Nyquist" frequency of 22.05 kHz for CD music at 44.1 kHz sampling rate. We only get 11 dB of attenuation where the theory demands infinite level.
Headphone Output Measurements
Let's start with driving the 1/4 inch headphone jack using the USB input (i.e. going through the internal DAC):
Power level is very good but despite the much higher price, the Jotunheim fails to outperform the Topping DX3 Pro on that front. Noise is naturally higher as measured before.
Let's measure the performance by feeding the unit from its analog input:
We get good bit of reduction of noise there. Power output increases some because I am able to push the amplifier section harder than internal DAC. But then again it gets distorted so no sense in doing that. Also, once the amp distorts, it doesn't matter if you use the DAC or not. They both meet at the same place/point as you see in the graphs.
Switching the output load to 33 ohm with analog (RCA) input gets us the typical distortion curves of Schiit headphone amplifiers when stressed:
We see that two stage rise in distortion. First one is some kind of non-linearity which should have been dealt with using better circuit design. The second sudden peak is the unit running out of current and sharply distorting.
As I have shown in the graph in green, the little $99 JDS Labs Atom blows away the Schiit Jotunheim out of the water. It is essentially distortion free until it too runs out of current. That happens at 1 watts with the Atom whereas the distortion starts to rise in Jotunheim at just 50 or so miliwatts.
Since there is XLR (incorrectly called "balanced") outputs, let's measure that using 50 ohm load. And for variety use XLR input:
Same problem as before manifests itself in very early rise in distortion. As such, our other state-of-the-art, the Massdrop THX 789 leaves the Jotunheim in the dust, produce more power at a fraction of the distortion.
Finally, here is the test of channel balance versus position of volume control:
Due to use of analog control we get inaccuracies between channels. Until you get to low volumes, it is not too bad at 0.4 dB.
Listening Tests
I did not bother with my usual level matched blind AB tests. Instead I just played some music and turned up the volume to detect distortion. WIth Sennheiser HD-650 there was fair bit of power but once you went past the 2:00 o'clock position on the volume control, you could easily hear the distortion setting in. Similar situation existed with Hifiman HE-400i. Neither could pass my test of loudness without distortion which admittedly means pretty high SPLs (sound levels).
Conclusions
Use of traditional delta-sigma DAC chip significantly improves the performance of Schiit Jotunheim versus the multibit version tested before. So if you are going to go this route, that would be my strongest recommendation.
The world has moved on since the Jotunheim was designed and today we have stellar headphone amplifiers in the form of JDS Labs Atom and Massdrop THX AAA 789. Few amplifiers can get close to the performance of these units much less a Schiit headphone amplifier. So while the Jotunheim can be an enjoyable device to use versus lower powered amplifiers, it just doesn't rise up to level of objective performance I like to see in this day and age and at that price.
On the plus side though, this is a very full featured all-in-one unit. Right now I can't think of any other device I have reviewed with the same functionality.
Most important concern here is electrical safety. I continue to be dismayed that Schiit has not recalled all of these units to properly repair them after confirming that the top is supposed to be grounded. On that basis alone, I cannot recommend Schiit Jotunheim in any configuration for purchase.
------------
As always, questions, comments, recommendations, etc. are welcome.
They say you should save money for a rainy day. It rains for about 6 months here so I need plenty of money!!! Please consider donating funds using:
Patreon: https://www.patreon.com/audiosciencereview), or
upgrading your membership here though Paypal (https://audiosciencereview.com/foru...eview-and-measurements.2164/page-3#post-59054).
Last edited: