Fosi Audio Box X5 Phono Preamp Review

[amirm]

Fosi itself is still showing $109.99 inclusive of tax plus shipping.

 

[Heinrich]

Fosi itself is still showing $109.99 inclusive of tax plus shipping.

Fosi Audio Box X5 Phono Preamplifier

Fosi Audio MM Turntable Phono Preamplifier. Compact design with 3.5MM AUX and RCA output.

fosiaudio.com

From that I get said US$ 131, rounded off. Maybe there’s a geolocator at work (Spain here). The „google“ result, embedded in the headline it tells it to be 110, though. Amazon listing, Europe again is 134, currency conversion considered.

 

[amirm]

Oh, you must be seeing something else as it shows USD $109.99 for me. Maybe they added VAT?

 

[SIY]

Oh, you must be seeing something else as it shows USD $109.99 for me.

Me as well.

 

[Heinrich]

Oh, you must be seeing something else as it shows USD $109.99 for me. Maybe they added VAT?

Looks like that. Only recently I discussed what a bad idea VAT is, but, obviously Europe insists on it. And so follows Fosi, once they know from my browser‘s ID

Sorry for the fuss.

 

[renaudrenaud]

The parcel arrived today on the morning. Just plugged and confirmed it was running.

There was some 50Hz hum with the ground cable connected but when I disconnected it no more hum.

I hope tonight I will have some time to listen to some records.

 

[JeyElDee]

I just received one today and already sent it back. Power supply has a very loud hum.
Also imaging is off center to the right channel. Replaced all cables twice. Issue remains.

 

[Skeeter]

I have yet to seen it done "correctly." By far the main contribution to SINAD is mains hum (aggravated by RIAA equalization) . What you propose increases the propensity of measuring just that instead of any input contributions with long wires to a cartridge hanging in the air. The input signal level is incorrect resulting SINAD just being SNR, dominated by said mains hum. I used to use higher input voltage to get 2 volts nominal but folks insisted on 5 mv so here we are, completely ignoring distortion products.

Use of any cartridge raises the question of why that instead of another. And how the results are comparable to people using other cartridges.

This is on top of the fact that groove noise, etc. dominates actual use. Not input current, etc.

At the same time as all of this, folks then ask for shorted input noise and such which is worse than what I use currently.

Bottom line, I am not proud of this suite of phono tests but is what I have settled down on. It seems to differentiate products from each other and no one has shown actual uses that disagree with these findings.

With unbalanced inputs at very low levels (5mV) hum ingress is always going to be an issue. But, a big part of that problem is cable dressing between the turntable and the preamp, and less than optimal cable management between preamp, line amp and the mains loops. That said, it is possible to have an exceedingly quiet phono input wrt mains hum if grounding rules are followed carefully.

I am critical of MM phono measurements that do not include a real-world source because they ignore the biggest noise source outside of mains hum which is input noise current. As noted in my earlier post, physics says the thermal noise voltage of a typical cartridge is 3.1uV unweighted and that sets the noise floor. You could of course use resistor cooling techniques to perhaps extract another 1-2 dB SnR, but that adds complexity. Designing for low thermal noise is as challenging as designing or low hum ingress, and this is how a purchaser can ascertain whether sufficient engineering effort has been put into that aspect or not. Unfortunately just shorting the input misses all this stuff.

The 500mH + 1350 Ohms is representative of a typical MM cartridge and should cause no issues as a measurement standard.

From the 10 carts I looked at above, they all range from 400mH through to 800mH (excl. the mono Ortofon 2M) and 500 mH + 1350 Ohms is the spec of the iconic Shure V15. It is the cartridge inductance at HF that really causes the issue - the DC coil resistance much less so. As frequency rises, the cart Z rises and the LF shunting effect of the cartridge inductance plus coil resistance on the 47k load resistor at low frequency disappears, so the amplifier input sees 15-20k source impedance in the upper octave.

To see how important noise current is in an MM amplifier, consider two opamps, the highly regarded AD797 from ADI featuring 0.9nV/rt Hz input noise voltage (designed by ADI Fellow Scott Wurcer who used to frequent this forum and sadly no longer with us RIP) and a GP JFET input opamp, an OPA1641 or 1642 which features an input noise of 5.1nV/rt Hz (1kHz spot noise figure quoted for both opamps) i.e. more than 5x the AD797. The AD797 incidentally featured in some of the LIGO instrumentation because it was at that time the quietest opamp available anywhere. However, the SnR ref 5mV on the AD797 used in an MM phono preamp comes out at just 63.78 dB while on the OPA1641/2 it is 73.15 dB, a difference of nearly 10 dB. The reason for this discrepancy is the input noise current. On the AD797 the figure is 2pV/rt Hz, while on the OPA1642 it is 0.0008 pV/rt Hz, in other words, pretty much negligible. The AD797 is a bipolar input opamp, while the OPA1641/2 is a JFET input opamp. Now, if you were designing an MC phono preamp, you would NOT choose the OPA1641/2, but the AD797 because the MC source resistance is so low (typically about 10 ohms) that the noise current term contribution to SnR is negligible. The type of device required for low Z sources is very different to that required for high Z sources and that's why the AD797 was used in the LIGO instrumentation - it was exceedingly quiet with low Z sources and of course, featured class-leading linearity.

If we shorted the input to our MM phono amp and measured SnR, the AD797 would measure 89.5 dB ref 5 mV and the OPA1641/2 would measure 76.7 dB, a massive 12 dB swing in favour of the AD797, but a completely misleading result. That is a difference between the real-world source figure of 26 dB in favour of the AD797.

I don't want to get into a battle over this, but these are the engineering facts, and why I would encourage you, and John Atkinson at Stereophile, to measure SnR on MM preamps with a real-world source. For future measurements, you could always show both the shorted input figure and the real-world figure so manufacturers submitting gear would not feel aggrieved at having the lower real-world figure up against formerly measured gear that showed optimistically high SnR because they were measured with the input short (this was Marcel v. d. Gevel's idea of how to make it fair moving forward).

(NB, I've had to edit this a few times because I cross posted myself numerous times like an idiot. Hopefully, I got it making sense now )

 

[SIY]

With unbalanced inputs at very low levels (5mV) hum ingress is always going to be an issue. But, a big part of that problem is cable dressing between the turntable and the preamp, and less than optimal cable management between preamp, line amp and the mains loops. That said, it is possible to have an exceedingly quiet phono input wrt mains hum if grounding rules are followed carefully.

I am critical of MM phono measurements that do not include a real-world source because they ignore the biggest noise source outside of mains hum which is input noise current. As noted in my earlier post, physics says the thermal noise voltage of a typical cartridge is 3.1uV unweighted and that sets the noise floor. You could of course use resistor cooling techniques to perhaps extract another 1-2 dB SnR, but that adds complexity. Designing for low thermal noise is as challenging as designing or low hum ingress, and this is how a purchaser can ascertain whether sufficient engineering effort has been put into that aspect or not. Unfortunately just shorting the input misses all this stuff. To give a good example of how this stuff affects the noise in an MM preamp, take an AD747 which is an expensive but very highly regarded opamp (it was used in some of the LIGO instrumentation and designed by ADI Fellow Scott Wurcer RIP) and features 0.9nV/rt Hz input voltage noise and put it up against a GP JFET input opamp like an OP1641 or OP1642.

The 500mH + 1350 Ohms is representative of a typical MM cartridge and should cause no issues as a measurement standard.

View attachment 421274


From the 10 representative carts I looked at above, they all range from 400mH through to 800mH (excl. the mono Ortofon 2M) and 800 Ohms + 1350 Ohms is the spec of the iconic Shure V15. Its the cartridge L at HF that really causes the issue - the DC coils resistance much less so. As frequency rises, the cart Z rises and the LF shunting effect of the cartridge inductance plus coil resistance at low frequency disappears, so the amplifier input sees the 47k shunted by typically 15-20k in the upper octave.

To put my measurement into perspective, the Stanton cartridge I used has an inductance of 530mH and DCR of 900R, so pretty much middle of the pack.

Input noise current is indeed an issue in some preamps, but Johnson noise from the cartridge resistance and the unshunted input resistance set a hard physics limit to S/N. Here's a calculator to run the numbers.

 

[Borko]

I’m at a loss as to how SnR can be quoted at 92 dB on a phono amp or 83 dB SINAD. Straight physics says that is not possible. The thermal noise of a 47k MM load resistor shunted by a typical MM cart (taken as 500mH in series with 1.3k Ohms) is 3.1uV. This sets the absolute best case SnR to 76 to 77 dB ref 5 mV. To achieve that level of performance, the input devices will have to be carefully selected for lowest noise voltage and importantly, very low noise current. Any old opamp or discrete device(s) don’t cut it.

Amir, I do wish you’d measure these things correctly. To assess the noise performance you could have a metal box into which you mount a typical cartridge, bring the connections out to some RCA plugs. To get the true SnR, connect the box to the DUT and get the integrated noise output over 20Hz to 20 kHz on the AP. Calculate unweighted SnR from the output voltage which you get in step 2 which is to remove the box and inject 5 mV from the AP to get an output signal level (this will automatically account for gain variance from product to product).

Why is this important? Because injecting a signal into a phono amp from a low Z source like a AP completely ignores the biggest noise contributor in an MM phono amp: input noise current. BTW John Atkinson from Stereophile has also been taken to task for not measuring MM phono amps correctly.

How could it be that SNR is rated 92 dB ? If we look in in the manual, for 38 dB gain, 5 mV input, and 37 uV noise floor, a simple calculation reveals 80.6 dB ?

 

[Skeeter]

Dunno - you'd have to ask Fosi. Could be they have quoted the noise in dBV and not reference 5mV or whatever their input signal was. If they for example were measuring the noise floor at the output and the preamp was outputting say 300mV, but they quoted the SnR in dBV, that would gain them 10.4 dB improvement in the quoted SnR. But, that is not the correct way to do it. If the preamp was outputting 300mV, the SnR must be quoted with reference to 300mV.

There was a guy who used to frequent another forum that tried to get away with a commercial product wherein he quoted the SnR at the maximum output of the preamp which was about 10V. So he had these outrageous SnR figures on MM phono inputs of 120 or 130 dB which of course was absolute nonsense.

Caveat Emptor.

 

[Crosstalk]

seems like Fosi is on fire! Hope there will be a higher power mono block from them which matches the SNR of the Topping b100s. Topping, SMSL, Fosi pushing the boundaries is a great thing for the consumers

 

[Ecaroh]

I just received one today and already sent it back. Power supply has a very loud hum.
Also imaging is off center to the right channel. Replaced all cables twice. Issue remains.

I’ve used mine for five days. No hum, no balance problems, extremely quiet.

 

[JeyElDee]

I’ve used mine for five days. No hum, no balance problems, extremely quiet.

Lucky you. I have no idea why they include such an ondfashioned transformer heavy bulky power supply.

 

[amirm]

Lucky you. I have no idea why they include such an ondfashioned transformer heavy bulky power supply.

This is a very common approach in these products.

 

[Ecaroh]

Lucky you. I have no idea why they include such an ondfashioned transformer heavy bulky power supply.

In my experience, AC-AC transformers tend to be heavy. I'm not sure they needed one rated at a full 1-amp, but it's convenient that it isn't an acutal wall-wart, i.e. the plug-prongs are not embedded in the PSU.

 

[lini]

(...) Use of any cartridge raises the question of why that instead of another. (...)

For MM you could consider a dummy load as in the IHF-A-202 standard (chapter 3.12.), i.e. a 500 mH inductor in series with a 1 kOhm resistor plus a 125 pF capacitor in parallel.

Greetings from Munich!

Manfred / lini

 

[cfwotzka]

I’ve used mine for five days. No hum, no balance problems, extremely quiet.

My Box X5 is now heavily engaged since 3 weeks w/o hum and really really quiet. using only MC cartridges. the sound quality still leaves me speechless every time I'm listening to it. Bravo Fosi and many thanks amirm for testing this little machine!
I'm relatively sure the ones who buy it will keep it even if they are used to much more expensive phono pre-amp gear.

 

[JP]

For MM you could consider a dummy load as in the IHF-A-202 standard (chapter 3.12.), i.e. a 500 mH inductor in series with a 1 kOhm resistor plus a 125 pF capacitor in parallel.

Greetings from Munich!

Manfred / lini

Good to see you.

 

[lini]

Good to see you.

Nice to see you here too, JP!

Well and what shall I say? I guess, for technically oriented hifi fans there hardly is a way around joining ASR sonner or later.

Greetings from Munich!

Manfred / lini


P.S.: Did anyone already contact Fosi regarding the yet unknown MM input capacitance, btw?