-
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!
Fosi Audio Box X5 Phono Preamp Review
- Thread starter amirm
- Start date
I've never seen an impact below ~1kHz from loading, and certainly nothing in the bass region. "Thin" means different things to different people. On a VM95ML at 50pF all-in you'd be about -2dB at 5.5kHz, -1dB at 10kHz, and -2dB at about 15kHz. If flat is the goal, the best compromise for that cart is about 250pF. All assume 47k Rl.
Skeeter
Member
The 1350 ohms isn't the problem. It's the 400+ inductance. The quality of construction has nothing to do with using it as a reference source.
Skeeter
Member
It is currently measured here and at Stereophile with a very low source impedance which for this discussion is the same as shorting the input and measuring the noise at the output. Measuring using a low source impedance means the biggest noise contributor in an MM phono amp - the input noise current - flows through ~0 ohms and thus does not develop a noise voltage, which IS what happens if you feed it with a real-world MM source. You therefore end up with the bizarre situation that opamps like an AD797 or OPA1611 are quieter than an NE5532 or OPA1642 and so get ranked more highly for noise performance when in fact they are 7-10 dB worse when you connect a cartridge. You have to select the correct type of amplifier depending on the transducer you are using. For MC, you want very low noise voltage (so use an AD797 or OPA1611/12) and for MM, you should choose something with very low or negligible noise current and low noise voltage - so use an OPA1641/2 or NE5532/4 (these latter two have higher noise voltage than either the AD797 or the OPA1611/12
Skeeter
Member
I don't want to see current noise specified. We are getting into technical details that probably most are not interested in unless they are designing low-noise analog circuits. I am suggesting that instead of measuring phono amps on this site directly out of an AP - which is a low source impedance - it is better to do it using a real-world MM source. @SIY has suggested just feeding the AP through the MM cart coils which seems a good approach to me. This will give a true reflection of the preamp noise performance when a cartridge is connected.
Alas, I feel that I‘ve fallen out of the ongoing discussion. But anyway, equivalent current input noise is measurable, easy in my book at least, and as such could be used to estimate the impact of any generator impedance. That would not be possible when measuring the phono pre with just one real cartridge, and would need further calculations when using more than one. Besides the technical stuff, I strongly suggest to just play around, because vinyl is about that to the best of it. I‘m not on an irony ticket, I mean it. I loved my platter once for a reason. Can‘t put my finger on it, there‘s something about personal involvement hard to describe.
Skeeter
Member
Agree noise current is easy to measure and the total noise contribution will depend on the cartridge used. But, the 500mH + 1350 ohms I proposed is a good compromise since the 500mH L bit is pretty consistent with most mainstream carts and in the big scheme of things the 1350 DC coil resistance is neither here nor there - it will not materially affect the real-world noise measurement with a standard 47k load.
Anyway, I concur there is a magic about T/Ts!
morillon
Major Contributor
- Joined
- Apr 19, 2022
- Messages
- 1,916
- Likes
- 452
This certainly makes obsolete the measurements and classification in mm practiced here until now...but offers a clear approach for the future, using a fairly standard cartdrige, and more relevant measurements...
but does not prohibit redoing a classification in mc which would seem to be globally quite relevant....
(Perhaps the opportunity arises to re-measure those who offered very optimistic measures in mm in the same way to put things in perspective.)
but does not prohibit redoing a classification in mc which would seem to be globally quite relevant....
(Perhaps the opportunity arises to re-measure those who offered very optimistic measures in mm in the same way to put things in perspective.)
Sure, it is about the feel of it. Not to far in the past (digital) measurement was science fiction for most of us. Do we want to spoil the experience of personal preference in letting numerical ranking take over? Given the not always fully understood limitations, in all their variety, is measuring one or the other isolated data reasonable? There is so much not just right, all by at least one order of magnitude well objectionable (vertical tracking angle, horizontal tracking angle, degradation with the n‘th copy from a ‚mother‘, … ), hopeless and promising at the same time.
With the recent advent of low cost, quite stabile, low voltage noise opamps, current noise became an issue. MM data is affected, sure. With the X5‘s design one might argue that it is hard on the edge. But still it won‘t be easy to declare the data presented here misleading.
If there‘s substantial criticism, I would expect it be presented by the critics beyond vague suspicion ;-) I‘ve seen nothing so far …
Add.: Look, the 1300 something Ohm DC won‘t do anything serious with noise current, the resonance from L/C at 10kHz or so is located where the preamp is 40dB below its max amplification, and it is narrow band, or low in reactive resistance. Plus the opamp is specified to be quite well behaved up to 10kOhm anyway (complicated because the reactive part dominates).
Last edited:
I experimentally showed over a 20dB difference when using a MM cartridge impedance as a source compared with a low impedance source. Noise current is an issue that can be significantly reduced with design, but Johnson noise (both of cartridge and load) can't be. A few minutes playing with the cartridge Johnson noise calculator over on my website will demonstrate the importance of that factor.
Still, although current noise can be minimized, it isn't necessarily. There's a lot of badly designed phono stages out there. Part of the issue is that devices which are optimal for MC (low base spreading resistance BJTs, for example) are not optimal for MM. And devices which are optimal for MM (JFETs, for example) are not optimal for MC. So if a phono preamp has variable gain to accommodate both, chances are that it's not optimum for at least one of them. I cringe when I see something like an AD797 (which works well for MC) ignorantly used in an MM preamp because its voltage noise is low.
Maybe in a pathological educational example, but not here fore sure. You are stressing a point not in case. What else to say?***
Anyway, vinyl is about personal preference, frankly accepting limits in favor of the content, not counting numbers for obvious reasons. Loosing a few dB of SINAD to noise or distortion doesn‘t compare to hum & buzz from a bad grounding technique, or, forgive the cruel example, loosing it to the pre-owners neglect of cleaning the record before playing, using a worn needle, embossed forever and a day into that single ‚unobtainium’ rare copy you may find in all your remaining lifetime.
The X5 is transparent regarding noise and distortion in all practical use cases. If in doubt please present some real data, example given on above mentioned 20dB—which real MM was it, and what of a real pre amp?
*** you didn‘t find your time to showcase the speculated issue with the X5, using opamp xyz, using your calculator, did you?
Quite the opposite. A very standard phono preamp with a very standard MM cartridge. In this very thread (besides my original post in a different thread a couple years ago). I don't have a Fosi in house to demonstrate this, but I would expect a very similar result since it's physics driven.
I've one on the way.
Skeeter
Member
Yes, very easy to do. I think the SINAD ranking on MM inputs on this site would change significantly. For starters, anything using an AD797 or OPA1611/2 would drop in a SINAD test to about 63dB (my calculator result ref 5mV) and a top notch JFET (opamp or discrete) would sit about 10 dB higher.
With shorted inputs, the noise floor on an AD797/OPA1642 is at -87 to -91 dB while a JFET input stage would be about 5-6 dB worse at -84 dB, which is a completely misleading result (these figures are unweighted).
With shorted inputs, the noise floor on an AD797/OPA1642 is at -87 to -91 dB while a JFET input stage would be about 5-6 dB worse at -84 dB, which is a completely misleading result (these figures are unweighted).
OP
- Thread Starter
- #356
Come back when you can show this. Don't do math without considering RIAA equalization.
Skeeter
Member
This is with math. RIAA will ‘improve’ the flat response noise by about 10dB due to the response shaping (the top 2 octaves in a MM pre account for near enough half the total thermal noise over 20 Hz to 20 kHz). LTspice models this very well but I also have a spread sheet that gives near enough identical results.
OP
- Thread Starter
- #358
I don't care about math or simulation. Bring measurements of actual phono stage including mains noise and RIAA equalization.
morillon
Major Contributor
- Joined
- Apr 19, 2022
- Messages
- 1,916
- Likes
- 452
maybe just if you still have this fosi , case which would seem to be able to be concerned, re-realize your procedure in mm by integrating in the loop just by a small archiclassic mm like an at-95 (which must be by far the best-selling modest cartdrige )
and for the futur....
It's more up to you to check your procedure....
and for the futur....
It's more up to you to check your procedure....
OP
- Thread Starter
- #360
No, it is up to whoever claims measurements change significantly. If they have never measured, then they can't make that assertion especially when simple things like mains leakage and equalization is not taken into account.
Similar threads
- Poll
- Poll
