• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required. 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!

Schiit Freya S Preamplifier Review

CDMC

Major Contributor
Forum Donor
Joined
Sep 4, 2019
Messages
1,172
Likes
2,321
I am a happy owner of a Freya S as well as an original Saga (bought on closeout). I tried the original Freya but returned it as it sounded notably inferior to the Adcom GFP-750 I was using (when I say notably, my wife heard it and wondered what was wrong). I am convinced that there was something wrong with that Freya, but never heard from Schiit after I returned it if it was defective.

When the Freya S came out, I decided to give that a go. My Adcom has worked great for the 15 years I had it, but the volume knob was getting scratchy in balanced mode and some parts are no longer available. I figured sell it while the selling is good (and was for $600) and get something that has a warranty and parts available. The Freya S checked the boxes, true balanced, the ability to run passive or active, and a reasonable price. To me the Freya S sounds great, as good as the Adcom, with perhaps a bit better imaging (but that could be due to the scratchy volume knob on the Adcom leading to distortion or imbalance).

The one thing I noted with the Freya S and my Saga is the deviation from spec in gain in the different modes. At first tried level matching by ear and found some differences in sound between the modes. Once I level matched by SPL meter, the differences all but disappeared (I think the passive sounds slightly more relaxed, but am not confident I could tell in a blind test). Here was what I found about the different gain levels and posted to SBAF:

So some interesting observations on the different modes and gain between the OG Saga and Freya S. I decided to pull out my SPL meter to check listening levels on my Saga for my desktop system, so I could note what volume positions net me 75db peak and 85db peak (using pink noise). As I will often be at my desk for hours at a time, I like to make sure I keep the volume below 75db peak. Keeping in mind I did this with a handheld SPL app on an Iphone X, so it may not be as accurate as if I ran it with room eq and the calibrated mic.

On the Saga OG there is about a 2 db gain between passive an active, 2 clicks on the volume gets you within .5 db.

The surprise was the Freya S (running from a Gungnir Multibit with balanced to a Sunfire amp with balanced). From passive to Gain 1, the Freya S jumps 7 db. From Gain 1 to Gain 4, it jumps another 13 db, or a total of 20db over passive. This is way above the rated 12db.
 
Last edited:

CDMC

Major Contributor
Forum Donor
Joined
Sep 4, 2019
Messages
1,172
Likes
2,321
That gain difference is the same as what I reported in the review. I will confirm with new sample but so far it seems it has higher gain than spec indicates.

From other reviews I have read of the Freya S, this is consistent. I know with the original Freya they reduced the gain jump at some point, but never updated the specification. I am not sure why they don't label it accurately on their website. The good thing is that they did put a delay in the new Freyas when switching gain mode so you don't accidentally blast yourself out of the room.
 

Snafu

Active Member
Joined
Dec 27, 2018
Messages
223
Likes
140
I tried the original Freya but returned it as it sounded notably inferior to the Adcom GFP-750 I was using. I am convinced that there was something wrong with that Freya, but never heard from Schiit after I returned it if it was defective.

Exactly what happened with my original tube Freya. Sound was strange, occasionally good but usually you were just wondering what's wrong without just catching it. It was hyper sensitive to touch, relay clicks came through speakers, even touching rca- or power-cables caused buzz from speakers.

Schiit/audio store (EU) said, tubes are wrong/not good, cables are not good enough, other equipment is placed too close etc. I sent them video about issues and it was returned asap. No explanation given if it was faulty or just schiit.

btw, even the remote included was damaged...

Last schitt for me :mad:
 

majingotan

Major Contributor
Forum Donor
Joined
Feb 13, 2018
Messages
1,511
Likes
1,781
Location
Laguna, Philippines
IMO, the closest thing to Freya S but on the pro market is Coleman Audio M3PHMKIII Monitor Controller. It might not win any beauty contest but it’s definitely a SOTA performing preamp if you want absolute transparency
 

Snafu

Active Member
Joined
Dec 27, 2018
Messages
223
Likes
140
^
I don't know what You mean by closest thing, but in pre amp in ~1k range i recommend Denafrips Hestia (it doesn't have head phone jack though).
 

ta240

Major Contributor
Joined
Nov 7, 2019
Messages
1,357
Likes
2,653
The JDS Atom ($99) has two inputs. In low gain mode it would serve as a perfectly fine pre-amp.

If you only need to switch between two input's, JDS also sells a passive pre-amp & switch ($45). I use one to toggle between a Sony HAP-S1 and an Oppo 105 to a pair of active speakers. The former holds my music library, the latter serves as a DAC for streaming from my computer and a Fire TV.

JDS pre-amp also switches two outs, which I use to toggle between my active speakers and the Atom headphone amp. Though, as I said, in theory the Atom is quiet enough I could have it inline to the active speakers all the time w/o degrading the signal from either the Sony or the Oppo. But since I have a passive output switch, I use that.

HTH

I keep hoping that someday JDS will come out with an actual 'preamp' in more of a wider form factor like this and with 3 inputs. As others have pointed out the reasonably priced, yet good, preamp field is pretty much empty.

And even though the Atom can be made to work with an additional switch it would still be nice to have more of a shelf form factor for some situations.
 

folzag

Member
Joined
Jul 24, 2018
Messages
99
Likes
131
I keep hoping that someday JDS will come out with an actual 'preamp' in more of a wider form factor like this and with 3 inputs. As others have pointed out the reasonably priced, yet good, preamp field is pretty much empty.

And even though the Atom can be made to work with an additional switch it would still be nice to have more of a shelf form factor for some situations.
Might be a long wait. ;) Looking at their product line it, they seem to see their niche as the desktop form factor. Me, I've made some plans... (wish me luck)
 

ta240

Major Contributor
Joined
Nov 7, 2019
Messages
1,357
Likes
2,653
Might be a long wait. ;) Looking at their product line it, they seem to see their niche as the desktop form factor. Me, I've made some plans... (wish me luck)

There is a lot to be said for them just focusing on doing their thing really really well and in reality there are probably a lot more headphone amp customers than preamp customers. I figure if I end up using a power amp where I need some sort of preamp I'll snag an Atom but I do still check their site regularly just in case.

I'm a big fan of the idea of adding modern, quality components inside vintage cases. I had an old amp I was going to try it with but I ended up scrapping it because I can picture the way things should turn out but I just don't have the skills to get them mounted, connected and assembled neatly. Some people have amazing skills in that area and some don't.
 

CDMC

Major Contributor
Forum Donor
Joined
Sep 4, 2019
Messages
1,172
Likes
2,321
Exactly what happened with my original tube Freya. Sound was strange, occasionally good but usually you were just wondering what's wrong without just catching it. It was hyper sensitive to touch, relay clicks came through speakers, even touching rca- or power-cables caused buzz from speakers.

Schiit/audio store (EU) said, tubes are wrong/not good, cables are not good enough, other equipment is placed too close etc. I sent them video about issues and it was returned asap. No explanation given if it was faulty or just schiit.

btw, even the remote included was damaged...

Last schitt for me :mad:

That is a shame. Mine showed up perfect in the box (of course it only had to go 350 miles). Operationally, it worked perfectly, it was just veiled sounding and there was no imaging. Making sure it wasn't me, I level matched between the Adcom and Freya and asked my wife to listen, without telling her what was being listened to, for, or changed. I asked her if she heard any difference between when I played the first and second time and if so, what. She (without knowing which it was) said the first (Adcom) sounded much clearer. We repeated twice with the same result. She was actually a bit disappointed as she liked the looks of the Schiit.

The good news is the Freya S has worked flawlessly since I got it and I am very happy with its features and performance.
 

MRC01

Major Contributor
Joined
Feb 5, 2019
Messages
3,424
Likes
4,030
Location
Pacific Northwest
What is the difference between a potmeter going into an amplifier or a potmeter going into another amplifier ?
Why would degradation be less ?
On further thought, I think I understand your point. The concept is that thermal noise of a resistor, as a physical phenomena, is a fixed voltage of absolute value. For example the noise of a 5 kOhm resistor at room temperature is about 1.2 microVolts. The key point here is that this voltage noise is powered by the random motion of electrons at a certain temperature, so its magnitude is independent of whatever voltage happens to be applied over the resistor.

Thus, describing voltage noise in decibels is misleading. A decibel is a ratio between 2 things. Expressing thermal noise in dB suggests that the noise level is a fraction or ratio of whatever voltage is applied over the resistor. On further reading, it seems that the dB measurement of thermal noise is based on an arbitrary reference of 1 V. That is, the actual physical phenomena of thermal resistor noise is volts (absolute fixed level powered by temperature). But by convention, it is expressed in dB relative to an arbitrary reference of 1 V.

For example, take that 5 kOhm resistor at room temperature with 1.2 microVolts of noise. That's -118 dB of noise only if you apply 1 V across the resistor. If you apply 2 V across it, then the noise is -124 dB (6 dB lower). The noise level is always the same no matter what voltage is applied across it. Within reason of course, since applying big voltages will heat it up, increasing the noise level.

Is that right?
 

617

Major Contributor
Forum Donor
Joined
Mar 18, 2019
Messages
2,403
Likes
5,296
Location
Somerville, MA

solderdude

Grand Contributor
Joined
Jul 21, 2018
Messages
15,891
Likes
35,912
Location
The Neitherlands
On further thought, I think I understand your point. The concept is that thermal noise of a resistor, as a physical phenomena, is a fixed voltage of absolute value. For example the noise of a 5 kOhm resistor at room temperature is about 1.2 microVolts. The key point here is that this voltage noise is powered by the random motion of electrons at a certain temperature, so its magnitude is independent of whatever voltage happens to be applied over the resistor.

Thus, describing voltage noise in decibels is misleading. A decibel is a ratio between 2 things. Expressing thermal noise in dB suggests that the noise level is a fraction or ratio of whatever voltage is applied over the resistor. On further reading, it seems that the dB measurement of thermal noise is based on an arbitrary reference of 1 V. That is, the actual physical phenomena of thermal resistor noise is volts (absolute fixed level powered by temperature). But by convention, it is expressed in dB relative to an arbitrary reference of 1 V.

For example, take that 5 kOhm resistor at room temperature with 1.2 microVolts of noise. That's -118 dB of noise only if you apply 1 V across the resistor. If you apply 2 V across it, then the noise is -124 dB (6 dB lower). The noise level is always the same no matter what voltage is applied across it. Within reason of course, since applying big voltages will heat it up, increasing the noise level.

Is that right?

Even when using a 5k potmeter and 10V input voltage the total power dissipation of 0.02W is not going to heat up a relatively large surface resistance track at all (far less than 1 degree in any case)

As long as there is no audible background noise from the transducer it all does not matter.
Of course doing amplification before the volume control (has its own weaknesses) and then buffer it (because of low output R) will yield measurably better results at low output voltages than potmeter followed by much amplification or a very noisy buffer.
In the end, as long as the end result is no audible self-noise at the ear then attenuation will not lower resolution over the entire range.
There is no difference between active and passive when done correctly.

In case of the Freya there seem to be balance issues and the amp stage (that adds distortion) is behind the volpot.
Certainly when using tubes on a high voltage I would put the gain stage in front of the pot and use a buffer (or not) behind it.
 

MRC01

Major Contributor
Joined
Feb 5, 2019
Messages
3,424
Likes
4,030
Location
Pacific Northwest
The key point is what powers the thermal noise in a resistor: (A) the voltage applied across it, or (B) the temperature based random fluctuations of its electrons?
If (A), then thermal noise is a fraction of the voltage applied across it. It's a constant dB relative to that voltage. If so, then you could attenuate the signal as small as you want without changing the SNR.
If (B), then thermal noise is a constant level independent of the voltage across the resistor. If so, then as you attenuate the signal, the SNR drops with it.
 

solderdude

Grand Contributor
Joined
Jul 21, 2018
Messages
15,891
Likes
35,912
Location
The Neitherlands
I think the active components before and after the volume control are more instrumental in noise the final noise floor than the contribution of the volume control itself.
 

MRC01

Major Contributor
Joined
Feb 5, 2019
Messages
3,424
Likes
4,030
Location
Pacific Northwest
My prior notion was based on the assumption that the noise in passive components is powered by the voltage across them, which makes them ideal for attenuation. But on further reading it seems that noise is constant regardless of the voltage across them, in which case the big advantage from the numbers I posted earlier disappears. A much smaller advantage still exists: (1) metal film resistors of equal value have slightly lower noise than a pot, and (2) perfect channel balance at all volume settings.

Another way to reduce noise as you turn down the volume might be to attenuate AFTER the gain-feedback loop. I see 2 advantages: (1) the volume pot attenuates the noise added by the gain stage, along with the signal, and (2) the volume pot's own noise isn't amplified by the gain stage.

I think the active components before and after the volume control are more instrumental in noise the final noise floor than the contribution of the volume control itself.
That may be true. One of the primary noise factors in the active components is the gain ratio. If the volume control reduced the gain ratio then you wouldn't need any separate attenuation. It would have to be inverting to have a negative gain ratio, but that shouldn't be a problem. In this case, as you turn down the volume, you turn down the noise with it.
 

solderdude

Grand Contributor
Joined
Jul 21, 2018
Messages
15,891
Likes
35,912
Location
The Neitherlands
(1) metal film resistors of equal value have slightly lower noise than a pot, and (2) perfect channel balance at all volume settings.

1: As said, the noise floor from the potmeter on line level is of no consequence.
Noise from resistors is relevant in mic pre-amps, phono (MM + MC) preamps, instrumentation amps and stuff like the AP analyzers and stuff.

2: yes true for stepped (and relay) volume control.

Another way to reduce noise as you turn down the volume might be to attenuate AFTER the gain-feedback loop. I see 2 advantages: (1) the volume pot attenuates the noise added by the gain stage, along with the signal, and (2) the volume pot's own noise isn't amplified by the gain stage.

It is possible to reduce the gain of amplifiers but there are snags.
1: Not all circuits like the gain to be reduced below a certain value or unity gain (non inverting amplifiers which most are) at the risk of oscillations.
Special care needs to be taken in such case.
2: You can't lower the volume below 1x gain

One of the primary noise factors in the active components is the gain ratio. If the volume control reduced the gain ratio then you wouldn't need any separate attenuation. It would have to be inverting to have a negative gain ratio, but that shouldn't be a problem. In this case, as you turn down the volume, you turn down the noise with it.

When using inverting amplifier circuits (or 2x inverting) you can go below unity gain but not all down to zero.
 

MRC01

Major Contributor
Joined
Feb 5, 2019
Messages
3,424
Likes
4,030
Location
Pacific Northwest
... When using inverting amplifier circuits (or 2x inverting) you can go below unity gain but not all down to zero.
Theoretically, the gain of a simple inverting opamp is -(Rfeedback / Rinput). Mathematically there's no limit to how low it can go. Is there a practical limit? Is that based on keeping the resistors in an ideal range?
 

solderdude

Grand Contributor
Joined
Jul 21, 2018
Messages
15,891
Likes
35,912
Location
The Neitherlands
Not only theoretically.
The question is whether it is needed.
Amps with inaudible noise floors can already be created.
It is pointless to go beyond the noise floor.
Besides, even such an amplifier design will still have a practical noise floor because of the active and passive components used.

But yes, you can build such an amp.
 

MRC01

Major Contributor
Joined
Feb 5, 2019
Messages
3,424
Likes
4,030
Location
Pacific Northwest
...
The question is whether it is needed. Amps with inaudible noise floors can already be created. It is pointless to go beyond the noise floor. ...
Some of the best DAC headphone amps measured here (SMSL M500, Topping DX7Pro) have a SNR that starts out excellent (120 dB) and drops to around 80 dB when you turn the volume down to 50 mV. That's a moderate listening level on HD600 headphones and quite loud with IEMs. Under these conditions noise may become audible, especially when listening to music with big dynamic range and quiet average levels.

Also, the overall trend in headphones is higher voltage sensitivity; low sensitivity classics like the HD600 or HE-6 are becoming the exception, not the norm.

This makes me suspect that there may be a need for a volume control that better preserves SNR at quiet levels. And how that might be achieved.
 
Top Bottom