Fosi Audio Box X5 Phono Preamp Review

[StevenSGT98]

We, audiophiles, do not engage in psychoacoustics, covering up the blunders of manufacturers.) Amir's measurements at 20 kHz are a disaster for this device. How could the developers allow this? They apparently did not control how the result of their work sounds.I'll find time on the weekend and take a photo of what they did there)

I encourage you to read some of Amir’s other phono stage reviews to compare.

 

[mike70]

i encourage you to listen to 20khz ... if you can, you're my hero

 

[StevenSGT98]

i encourage you to listen to 20khz ... if you can, you're my hero

I don’t know about anyone else but me? Mayyybeee 15k?? Not bad for almost 60 year old ears and everything I’ve done to them over the years.

 

[lini]

That chart already includes the RIIA curve - it is velocity encoded on the record - which already has RIIA applied.

I'm aware of that. Nevertheless one should keep in mind, that the RIAA cutting curve to a certain degree counters the usual roll-off towards the higher frequencies in typical music. And I'd guess, that especially checking 45 rpm maxi singles would yield quite a few more examples of pretty hot cuts than found in Shure's old study. And one could also argue, that a good phono stage should also provide a sufficient overload margin for cartridges with pretty hot output (like for example the Pickering NP/AC or the Ortofon Q.Bert...).

Greetings from Munich!

Manfred / lini

 

[StevenSGT98]

I'm aware of that. Nevertheless one should keep in mind, that the RIAA cutting curve to a certain degree counters the usual roll-off towards the higher frequencies in typical music. And I'd guess, that especially checking 45 rpm maxi singles would yield quite a few more examples of pretty hot cuts than found in Shure's old study. And one could also argue, that a good phono stage should also provide a sufficient overload margin for cartridges with pretty hot output (like for example the Pickering NP/AC or the Ortofon Q.Bert...).

Greetings from Munich!

Manfred / lini

This is from (partially-expanded upon) Shure study.

I’m not really sure how many play hot old 45’s anymore. The theories point to the 3.5-8khz as the sweet spot. Most recent studies point to 6db of overload as being sufficient. There is another thread or 2 floating around here discussing this very thing.

 

[abrcdbra]

Well, people. I took the device apart. Here's a photo. Sequentially from input to output. Later, in a few days, I'll draw and analyze the diagram of this device. But specific problems are already visible.
NE5532 in the device is made by TI. Don't pay attention to the 5532 made by Segnetics. I have now replaced it, since I prefer its sound. Naturally, this is an original Segnetics, and not a modern 50-cent fake.

 

[abrcdbra]

View attachment 441979

Yes, an explanation for those who do not understand this picture. The nominal level of the cartridge is given by manufacturers at a speed of 5 cm/sec. For MM it is 3...5 mV. The picture shows that the maximum levels at high frequencies are in the range of slightly more than 50 cm/sec. That is, the amplifier must withstand an overload of more than 20 dB at high frequencies without audible distortion. Many Japanese manufacturers have adopted an acceptable overload of 26 dB for their devices. And this is correct, taking into account the content of the real music signal. And this has been confirmed by many decades of experience of humanity. There is no need to invent other meanings for yourself. ))

 

[abrcdbra]

I'll continue, so that the general sentence is clear.) That is, taking the cartridge level of 5 mV, taking into account the RIAA curve, we get the following values of input voltages. At 20 Hz - 0.5 mV, at 1 kHz - 5 mV, at 20 kHz - 50 mV. And taking into account the overload of real musical signals of 26 dB accepted by human experience, it turns out that the amplifier must withstand without audible distortion at 20 Hz - 10 mV, at 1 kHz - 100 mV, at 20 kHz - 1 V. And there is no getting away from this, if you want to take into account all sorts of factors of a real musical signal, including clicks. Well, this is the very beginning of the difficult path of building a sounding RIAA amplifier for a real signal and a reactive-active signal source.)

 

[Balle Clorin]

But have you taken into account the spectral/frequency content of music ? Also the cutting lathes cannot cut high levels as high at 20khz as at 1khz, also at some point the cutting stylus simply cannot move fast enough without the backside hitting the already made groove… that is why a frequency sweep test record is made with no treble emphasis or at a very much lover level.

Here is the frequency’s and level distribution of some music genres..luckily the level drops significantly at higher frequency, allowing the vinyl recording and reproduction to work with reasonable “mechanics”

https://hifisonix.com/wp-content/uploads/2018/01/Disc-Recording-Equalization-Demystified.pdf

Stereo Lab - Maximum recorded velocities on gramophone records

Stereo Lab Maximum recorded velocities on gramophone records

pspatialaudio.com

Here is a specific vinyl track recorded at 96khz average level is -20db in this particular set up

1-5kHz us 8.6db lower and 5-10k -21db lower and 10-15k 27.6db lower and 15-20k 34.4db lower

Take a closer look at the scale we hava a drop of 25-40db from 1 kHz to 10 and 20kHz

This music( big band - spider and the fly-) drops off more than 20db per decade, that is more then twice as much as pink noise and more than the RIAA treble emphasis when cutting the record.

So no cartridge cannot and will not ever reach 1Volt

 

[antcollinet]

it turns out that the amplifier must withstand without audible distortion at 20 Hz - 10 mV, at 1 kHz - 100 mV, at 20 kHz - 1 V

Are you aware of any phono pre-amp at all that can tolerate 1V at 20kHz in it's input without clipping? Or even 1/5th of that?

Here is $2k device that manages 20mV

Lehmannaudio Decade Phono Stage Review

This is a review and detailed measurements of the Lehmann audio (one word?) Decade Moving Coil and Moving Magnet phono preamplifier. It is on kind loan from a member and costs US $2099. As you see, it comes in two boxes with the one on the left being the power supply and on the right, the...

www.audiosciencereview.com

A device at a slmilar price to the fosi that can manage about 10mV

Lehmannaudio Decade Phono Stage Review

This is a review and detailed measurements of the Lehmann audio (one word?) Decade Moving Coil and Moving Magnet phono preamplifier. It is on kind loan from a member and costs US $2099. As you see, it comes in two boxes with the one on the left being the power supply and on the right, the...

www.audiosciencereview.com

Another in the similar price bracket : 50mV

Lehmannaudio Decade Phono Stage Review

This is a review and detailed measurements of the Lehmann audio (one word?) Decade Moving Coil and Moving Magnet phono preamplifier. It is on kind loan from a member and costs US $2099. As you see, it comes in two boxes with the one on the left being the power supply and on the right, the...

www.audiosciencereview.com

A better device but still only getting to 80mV

Vincent PHO-8 Phono Stage Review

This is a review and detailed measurements of the Vincent PHO-8 Phono stage/preamplifier. It is on kind loan from a member. The PHO-8 costs US $249 on Amazon with free shipping. The PHO-8 comes in a two-box configuration with separate linear power supply which I did not expect to see in this...

www.audiosciencereview.com

And how are you getting to 1V when the more or less absolute limit of 50cm/s is about 200mV with your cartridge? And if I'm reading things correctly, that 200mV is a peak, so around 140mV rms.

 

[antcollinet]

But have you taken into account the spectral/frequency content of music ? Also the cutting lathes cannot cut levels as high at 20khz as at 1khz, also at some point the cutting stylus simply cannot move fast enough without the backside hitting the already made groove… that is why a frequency sweep test record is made with no treble emphasis or at a very much lover level.

Here is the frequency’s and level distribution of some music genres..luckily the level drops significantly at higher frequency, allowing the vinyl recording and reproduction to work with reasonable “mechanics”

https://hifisonix.com/wp-content/uploads/2018/01/Disc-Recording-Equalization-Demystified.pdf

Stereo Lab - Maximum recorded velocities on gramophone records

Stereo Lab Maximum recorded velocities on gramophone records

pspatialaudio.com

Here is a specific vinyl track recorded at 96khz average level is -20db in this particular set up
View attachment 442042

Useful links, thanks.

 

[KSTR]

Here is the frequency’s and level distribution of some music genres..luckily the level drops significantly at higher frequency, allowing the vinyl recording and reproduction to work with reasonable “mechanics”

Same issue as with the other poster, those are large size FFT spectra, which do not account for sort transients and their peak levels.

Say, you have a 16k point FFT and a single 5-cycle full-scale burst at 10kHz in that block which is just a few dozen of samples. The FFT will NOT reflect that you had a full scale signal.

Now consider the same 5-cycle transient at 100Hz. It covers a much larger area of the block size and will thus give a higher readout, 100x higher.

==> Large FFTs are generally useless to examine peak levels. And short FFTs (like 8..32 samples) lack the frequency resolution.

 

[lini]

I'll continue, so that the general sentence is clear.) That is, taking the cartridge level of 5 mV, taking into account the RIAA curve, we get the following values of input voltages. At 20 Hz - 0.5 mV, at 1 kHz - 5 mV, at 20 kHz - 50 mV. And taking into account the overload of real musical signals of 26 dB accepted by human experience, it turns out that the amplifier must withstand without audible distortion at 20 Hz - 10 mV, at 1 kHz - 100 mV, at 20 kHz - 1 V. And there is no getting away from this, if you want to take into account all sorts of factors of a real musical signal, including clicks. Well, this is the very beginning of the difficult path of building a sounding RIAA amplifier for a real signal and a reactive-active signal source.)

This would appear overdimensioned to me - or respectively "doppelt gemoppelt", as we'd say over here. I.e., I'd assume those 50 cm/s velocities to already represent signal peaks, so that another 26 dB headroom on top wouldn't seem necessary and a 6 dB reserve should do - in which case roundabout 100 mV should suffice for regular hifi cartridges and roundabout 200 mV for DJ/broadcast models with pretty hot output.

Are you aware of any phono pre-amp at all that can tolerate 1V at 20kHz in it's input without clipping? Or even 1/5th of that?

According to the diagram in the brochure the Technics SU-9600 could digest almost 5 V @ 10 kHz and 0.01 % THD at its phono 2 input in the 3 mV input sensitivity setting. Quite impressive.


Greetings from Munich!

Manfred / lini

 

[Heinrich]

Well, people. I took the device apart. Here's a photo. Sequentially from input to output. Later, in a few days, I'll draw and analyze the diagram of this device. But specific problems are already visible.
NE5532 in the device is made by TI. Don't pay attention to the 5532 made by Segnetics. I have now replaced it, since I prefer its sound. Naturally, this is an original Segnetics, and not a modern 50-cent fake.

Signetics invented it, but to say TI was doing bad on the 5532 is a bold statement. Especially if only founded on subjective listening.
Later you calculate headroom. Do you know about and consider the limitations of a) the cutting, b) pressing and finally the pickup process? Let alone the cable losses with that high voltages!

 

[Rischa]

Just installed a new X5 this morning. I'm not detecting any noise at all through either channel with my ear right next to the tweeter, but this is with my low-sensitivity Ascend Acoustics; might be a different story with my JBLs.

For old records the lowest gain setting seems to mitigate pops and clicks best, but the 48dB setting sounds great with new pristine vinyl (using an AT VM540ML).

The build quality is exceptional for the price -- way better than the $500 Rega fono-mm I had at one point. Pretty nice for $110 shipped.

 

[Balle Clorin]

My RIAA has peak/clip detection and never warn about or register clipping at MC 59db on music … overload margin is mediocre but still no problem. The problem stated here is imaginary

 

[Laika2]

I don’t know about anyone else but me? Mayyybeee 15k?? Not bad for almost 60 year old ears and everything I’ve done to them over the years.

Same here, still enjoy the hobby but in 10 years my system may sound like my TV speakers, hope not but I'm expecting it. Oh well, if AM radio sounds as good as my system to my aging ears the equipment sure is cheaper.

 

[Elgered]

Over the last ~30 years I had a lot of different phono stages and for some pickups this is a really nice alternative.
It totally seems to like modern MCs and lower output MMs.
In my setups it is a nearly perfect match for my Hana SL, the old Denon DL-110 and the Technics EPC205CII-L /SAS.
Others, like Elac STS444-12 or Ortofon (Super-)OM40 find better partners elsewhere.

Noise or noticable distortion are not an issue here, but somehow with the "louder" MMs it is loosing dynamics, imho sounding a bit compressed compared to other phono stages with similar amplification levels.

 

[abrcdbra]

I was confused by the value of the second resistor of the correction circuit when the first one was 7.32 kOhm. So I tried to measure the values of the capacitors in this circuit. Without desoldering them. But it’s clear that I didn’t succeed. Therefore, during the analysis, I will assume that the capacitor in series with the 1.5 kOhm resistor is equal to 0.3 μF. Because it is made up of three identical-looking capacitors.

 

[abrcdbra]

But I am a stubborn old experimenter. And I succeeded. Well, yes, close to the calculated values for the 7.32 kOhm resistor.

The last photo shows the values of the decoupling capacitors. There are 6 of them in the device, 3 per channel, and they are all identical.