Fosi Audio GR70 Quad Tube Headphone Amp / Preamp

[solderdude]

Hi everyone,

I’ve recently ordered an Ultra Low Noise Linear Power Supply, scheduled to arrive on March 27th. My goal was to find a high-spec LPS that offers the best "performance-to-dollar" ratio to see if further gains can be squeezed out of my current desktop chain.

While my Fosi already performs admirably with the stock switching supply (subjectively zero noise floor), I wanted to test whether a specialized power solution could offer any measurable or audible improvement.

Technical Specifications of Interest: The ZeroZone LPS-1706A caught my eye due to its ripple specifications. The manufacturer quotes:

Audio Bandwith Ripple (20Hz–20kHz): 1uV at rated current (12V/4A)

Full Bandwith Peak Ripple (1MHz): 4.8mV at rated current (12V/4A)

Internal Design & Architecture: The unit utilizes a MOS regulator and a significant capacitor bank—50 x Nichicon FW series "Gold" audio-grade electrolytics. The design philosophy here appears to be using a large array of lower value uF capacitors to achieve ultra-low ESR and faster charge/discharge cycles compared to a few massive cans.

The Test Plan: I’ll be performing some A/B listening tests between the stock wall-wart and the ZeroZone. I am particularly interested to see if the low-impedance configuration and fast power delivery impact low-frequency layering or soundstage transparency.

Beyond the Fosi, I also plan to test this with my Apos X Druid R2R DAC, as R2R ladder networks can often be sensitive to power supply fluctuations and noise.

I’ll report back with my impressions once the unit has settled in.

And you connect that via a cable which adds resistance and inductance to the otherwise low output R of the power supply ?
A bit like having an amp with DF of 10000 and using 5m of cable and having an inductor in series with the woofer.

 

[antcollinet]

Hi everyone,

I’ve recently ordered an Ultra Low Noise Linear Power Supply, scheduled to arrive on March 27th. My goal was to find a high-spec LPS that offers the best "performance-to-dollar" ratio to see if further gains can be squeezed out of my current desktop chain.

While my Fosi already performs admirably with the stock switching supply (subjectively zero noise floor), I wanted to test whether a specialized power solution could offer any measurable or audible improvement.

Technical Specifications of Interest: The ZeroZone LPS-1706A caught my eye due to its ripple specifications. The manufacturer quotes:

Audio Bandwith Ripple (20Hz–20kHz): 1uV at rated current (12V/4A)

Full Bandwith Peak Ripple (1MHz): 4.8mV at rated current (12V/4A)

Internal Design & Architecture: The unit utilizes a MOS regulator and a significant capacitor bank—50 x Nichicon FW series "Gold" audio-grade electrolytics. The design philosophy here appears to be using a large array of lower value uF capacitors to achieve ultra-low ESR and faster charge/discharge cycles compared to a few massive cans.

The Test Plan: I’ll be performing some A/B listening tests between the stock wall-wart and the ZeroZone. I am particularly interested to see if the low-impedance configuration and fast power delivery impact low-frequency layering or soundstage transparency.

Beyond the Fosi, I also plan to test this with my Apos X Druid R2R DAC, as R2R ladder networks can often be sensitive to power supply fluctuations and noise.

I’ll report back with my impressions once the unit has settled in.

Please make sure your "impressions" are derived from well controlled blind listening comparisons.

Otherwise your reports of what you hear (and I don't doubt you will "hear" a benefit) will be valueless.


EDIT : you are new here and that probably sounds harsh, so a little background:

This site is focussed on the science and engineering of audio. Part of that is a recognition that when we listen to stuff our brains are quite capable of creating "impressions" that do not exist in the sound reaching our ears. This always happens when we compare two devices knowing which we are listening to. It happens to all of us - and we are not able to turn it off. It is called perceptive bias.

So listening comparisons where you know what you are listening to are always likely to be tainted by this bias - as such they have little to no value for anyone else. This is especially the case when you are trying to derive an "improvement" to a device that you aready admit is audibly perfect (you say you cannot hear any noise), or when the device is known from its measurements to be audibly perfect.

 

[Deleted member 90885]

I’ll be relying on the quoted ripple and noise figures for this one rather than running my own bench tests. My goal for the S.S-LPS-1706A is purely a subjective experiment. I want to see if I detect any audible shifts in bass authority or the soundstage when pairing the Fosi GR-70 with a linear supply. I already have a completely black background with the factory wall-wart, so I’m not hunting for a lower noise floor; I'm just curious if a more robust power delivery impacts the presentation.

 

[Deleted member 90885]

Please make sure your "impressions" are derived from well controlled blind listening comparisons.

Otherwise your reports of what you hear (and I don't doubt you will "hear" a benefit) will be valueless.


EDIT : you are new here and that probably sounds harsh, so a little background:

This site is focussed on the science and engineering of audio. Part of that is a recognition that when we listen to stuff our brains are quite capable of creating "impressions" that do not exist in the sound reaching our ears. This always happens when we compare two devices knowing which we are listening to. It happens to all of us - and we are not able to turn it off. It is called perceptive bias.

So listening comparisons where you know what you are listening to are always likely to be tainted by this bias - as such they have little to no value for anyone else. This is especially the case when you are trying to derive an "improvement" to a device that you aready admit is audibly perfect (you say you cannot hear any noise), or when the device is known from its measurements to be audibly perfect.

I appreciate the perspective on bias. My goal here is a personal exploration of how this LPS interacts with my specific chain, particularly the R2R DAC. I'll share my subjective findings for those interested in the 'real-world' desk experience, and then I'll leave the lab-grade double-blind testing to the experts. Happy listening!

 

[solderdude]

In that case it is almost guaranteed the box WILL bring an audible improvement. Happy 'testing' !

 

[antcollinet]

In that case it is almost guaranteed the box WILL bring an audible improvement. Happy 'testing' !

Looks like they decided this is not the place for them.