Question: Better sound from a $40 USB 2.0 HDD external drive OR from a custom-built low noise, low vibration external USB 2.0 HDD device?

Question: Better sound from a $40 budget USB 2.0 HDD external drive OR from a custom-built (assembled) low noise, low vibration external USB 2.0 HDD device?

Covid-19 Lock Down frivolous activity experiment: Audio from 2 varieties of 500GB 2.5” HDD.

Is this of interest to this forum then I am happy to start the following experiment.

I spent some time building my own great sounding streamer (IMHO) and during lockdown just looked at the pile of stored stuff I had left over. Almost enough stuff to perform this pointless or maybe useful experiment; I can not decide myself which it might be?

Drive #1. A WD Blue HDD in an external case with a USB 2.0 connector (power & data).
Drive #2. A custom-built, low noise, low vibration external USB 3.0 HDD device configured as follows:

Using a 2 stage EFI mains power filter + standard 12V/5V DC switching power supply (SMPS) + SATA Power Noise Filter from the SMPS to the USB 2.0 interface. A 10,000 rpm WD VelociRaptor (1 Amp) HDD mounted to a mass block cooler for heat dissipation mounted on 2# separate levels of anti-vibration mounts in addition to rubber feet (so, 3 levels).

All mounted in a fancy aluminium chassis: blue LED front panel push power switch + small LED at the rear beside a fused C14 power socket. A rear panel mounted USB 3.0 Type B socket for USB to a Streamer or PC etc. Chassis air flow to be passive via holes in the base plate at the front and holes at a higher level on the back plate.

This filtered mains power supply and the SATA noise filter we hope will reduce power supply interference? to improve? audio Quality? from the hard disk and USB interface??? Cleaner power can also help protect the HDD, slightly increase efficiency and stability. The SATA filter should also help lower working temperature, buffering, current ripple, EMI interference and isolation of the HDD.
The SATA noise filter to have Audio SMD WIMA capacitors, TDK power inductor, AVX tantalum capacitors and two-way filter ground isolation.

Why USB 3.0? (updated to USB 2.0 for parity)
For this amount of co$t it should also function as an expensive backup drive also and USB 2.0 could be limiting for its transfer rate of say 500 CD’s in FLAC etc.
Why Not an SSD?
1.) I had spare mechanical HDDs 2.) cost, using most of what I had. 3.) easy to swop-out to an SSD in the future. 4.) I can not get to the bottom of which dives have less electrical noise; some say 2.5” mechanical, 3.5” mechanical, Solid-State Drive…!?!?
Why not use the spare 3.5” HDD?
I am not yet sure of any heat dissipation issues in this experiment so am playing safer with a 2.5” HDD. The spare 2.5” 10,000 RPM HDD should access audio-files at speed (not as quick as an SSD) as data bandwidth is not an issue in this application. This HDD needs 1 Amp+ so most external standard interfaces/housing will not work without an external power source.

Which will sound better? The estimated $300 custom-built, the budget drive or neither (no difference)???

Testing outline: Unscientific and not to lab conditions. 6 well-known tracks played to 2 listener’s 5 times (60 tracks in total demo. 30 each). The other listener to be the random external HDD swapper an equal amount of times per 6 track session x 5. After a benchmark listen, marking to be simple, after each track = Better, Worst, No-difference. HDDrives feed to a Raspberry Pi 4, using, Volumio/Moode maybe roon via a NUC (TBC) USB 2.0 into a PS-Audio PerfectWave mkII DAC then a Chord Electronic Pre-Power combo and Wilson Benesch ACT speakers. OR via my own custom-built streamer to a Chord DSX1000 DAC (TBC).

Is this of interest ?

If your HDD is in the same room as your listening post and the HDD itself or its fan is noisy then it will impact your listening enjoyment.

Otherwise, zero SQ difference with what type of drive your music files reside on.


OK, we start with a ‘no’. I have noticed on these forums some heated debates on the effects of vibration, power supplies and cables in the digital realm. As a HDD and a flac files are in such a realm it’s a possible light hearted test bed.

No = 1 to embarking on this experiment. Yes = to voice a yes ad yet.

For every bit of hifi there are believers and those that decry the product. There are even those who have tried and decided it doesn’t work.
This post is like many others it will just degenerate into chaos.
The only plus point about this post is that it is in tinkering.

As an example a rethink one of the more idiotic pieces of hardware suggested by the OP.

Yes, the hard drives will work and send digital audio files and yet one will have a cleaner USB signal from a better environment. Say from a good hotel against a tent in a field!

This is s hearing test not measurements. I make no presumptions on the results. I did a similar experiment with streamers and my initial thoughts of the end results where overturned. I kept an open mind and use a streamer.

OK. No = 2 to not conduct the experiment. Yes = iyet to get a vote to proceed.

It is of interest to me. Curiosity mostly. I have settled on a server with internal solid state drives as the best solution for me. Small, portable and shock resistant.

1 Like

Yes, but nowhere in your storage selection criteria was the consideration of what would produce the best sonic experience.

A primary benefit of solid state drives is silence. And SQ in that respect was the reason I chose an internal SATA drive over a much cheaper HDD.

If you’re going to go to the trouble of staging a blind test, why not make it an A/B/X test instead of A/B ?

It won’t take any longer:

  • For each track, a benchmark listen to both “A” and “B”, followed by 4 (or whatever number you like – more is better) repetitions whose identity is determined by a coin flip by the tester.
  • Listener records whether he thinks each repetition is “A” or “B”.

But this way, you’ll be able to determine whether either of you can reliably distinguish between “A” and “B” (and not get saddled with rubbish “results” like: ‘Track 1 sounds better on A, but track 2 sounds better on B.’)

If you can’t reliably distinguish between the two, it’s nonsense to try to determine which one sounds “better.” If you can reliably distinguish between the two, it’ll probably be obvious which is better.


Jacques, I like your suggested method of testing. Happy to perform that method instead. Thank you.

1 Like

Great! I’ll be eager to hear your results.

I hear you. If this thread/post as you say “…is like many others it will just degenerate into chaos…”, then I will just stop and spare you all. Life is too short for such things and there is plenty more to do at this time; when you are just planning to share the results of a possible tinkering open minded activity.

BTW: do not have a SOtM SATA power noise filter yet get your point.

There are 5 possible outcomes:

  1. No real interest from this tread = Do not do it.
  2. HDD enclosure made + tested = Sounds better (subjectively by how much?)
  3. HDD enclosure made + tested = Sounds worse (subjectively by how much?)
  4. HDD enclosure made + tested = Sounds the same/inconclusive (subjective outcome?)
  5. No real interest from this tread = I just do it anyway for tortured fun/interest and end up with the world’s most expensive external USB hard disk drive! Maybe add a raspberry pi and make it a none roon music NAS???

BTW: I did home audition the Auralic Aries G1 Streamer and used the same said $40 budget USB 2.0 HDD external drive, my initial impression was (…) I’ll let you know at the end of this experimental test should there be enough interest.

I am guessing you are going to feel that there is a difference and that the custom HDD is better.

1 Like

I really have no pre-conceived notion of the future result.

That said I learnt something a few years ago a ‘coffee cup moment’. In my office I have a HiFi and it had a NAD M50 streamer/roon endpoint/CD player/ripper (very nice, happy with it for the office system). Its power supply failed NAD would not take self-diagnoses and it had to be shipped away 6-8 weeks return.

So, pulled a raspberry Pi 3, std smps and a mid-cost Digital-to-Digital Converter together, fed into a PS-Audio PerfectWave mk II DAC in the office. My expectations where low for SQ as the NAD was a Master Series product right! I just wanted something to fill the 6-8 week gap.

As I lifted my coffee cup to my lips and clicked play for the very first time… The cup stopped as this sounded just better, I hovered for 30 seconds… it was better. NAD M50 returned for an A/B/C listen … I sold the NAD M50 after I built many Streamer mock-ups.

I have kept 3# self-built streamers 1.) Budget-ish 2.) mid-band (now in my office) 3.) a no cost build where I stopped at ‘could not hear the difference’ this ‘stop spending, you can not hear the difference’ model now sits in my lounge and now longer use the streamer side of a Chord Electronics reference DAC/Streamer DSX1000.

Point is I had an open mind and found something different, when I hear “streamers do not make a difference” I have my own empirical hearing tests/kit that say different + My #3 streamer has been home auditioned against many streamer, some $5000+ and #3 sounded no worse (blind A/B).

Many moons ago I sold a Tag Mclaren DVD 32R (its best CD transport at $5,500) as I could not hear the difference between a $250 NAS and the DVD 32R. Another surprise, again from a ‘what if’ and an open mind. I do miss that 32R with its sliding lip and exquisite space age looks etc.

Who would have thought the humble USB from the ICT world is becoming the best audio bus and the bespoke audio SPDIF feed is fighting to hold ground!

Also found that in the digital audio realm you can obtain great performance for modest money, its just getting better for lower cost all the time.

1 Like

i have a hole available for throwing your moneys in :rofl:

1 Like

OK, I’ll take that as a ‘No’ to the experiment. Little money to be spent on this as I have the vast majority of parts form previous disassemble products.

Do what makes you happy, It’s just my point of view. :+1:

1 Like

I’m with Boy Zonder, tinkering is fun. If you have the parts and time, and it gives you joy and satisfaction, why not? You’ll end-up with a very nice external storage drive that you can take pride in, regardless of any perceived or real improvements in SQ. And, my guess is that it is going to sound really great (to you) each time you use it – just as it should.

Hello, Re: Covid-19 Lock Down, frivolous activity, tinkering experiment: From existing parts (mostly).

After almost a week of starting this tread in which to get a feel for the amount of interest and insights to make it more relevant should I start.

Yet I am not feeling a roon community ground swell of interest in this topic to see a result. So, with that I will put this on my Covid-19 lock down might do/not do list!!!
If I did go option 5. of possible outcomes that being… No real interest from this tread = I just do it anyway for tortured fun/interest and end up with the world’s most expensive external USB hard disk drive! Maybe add a raspberry pi and make it a none roon music NAS???..

To @Henry_McLeod and @Jacques_Distler I will let both know the subjective outcome; thank you both for your interest/input should I ‘take 5’ (Dave Brubeck pun intended, thank you). Yet, build the external drive #2 as a USB 2.0 instead to more align with #1 drive comparison.

BTW: Have seen Mechanical Hard Disk Drives have less electrical noise than SSD and Naim & NAD prefer 3.5” mechanical HDD against 2.5” and SSD (wish I could find out more on the topic).