• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required as is 20 years of participation in forums (not all true). Come here to have fun, be ready to be teased and not take online life too seriously. We now measure and review equipment for free! Click here for details.

Hardware Teardown of Pro-Ject Pre Box S2 Digital DAC

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
20,620
Likes
25,441
Location
Seattle Area
#21
The DAC/ADC is the interface between the two domains, so has to straddle the boundary.
Ah, I misread your argument as saying it should be on the digital side. :) Yes, I agree straddling it would be the right thing to do if the pinout allows it such.
 
Joined
Mar 29, 2018
Messages
21
Likes
13
Location
Bilbao, Spain
#22
On capacitors, they take a while to fail so I would use it during the warranty period as is. After that, open it up and see if any of the capacitors are bulging or anything leaking out of them. If so, then replace them immediately with a Japanese brand capacitor. As a precaution you could replace the cap on the digital side alone after that warranty period.
Would actually changing the capacitors improve sound quality (or the quality of the current they deliver) or you meant a measure to prevent failure? If so, which specs would be best?
On power supplies, most linear ones available on ebay for audio and RaspPi are 25VA 3A 5V DC ones. Are they suitable for the 1A requirement of the S2?

Thanks Amirm for your detailed rundown.
 
Joined
Feb 9, 2018
Messages
148
Likes
77
#23
Would actually changing the capacitors improve sound quality (or the quality of the current they deliver) or you meant a measure to prevent failure?
They are low ESR power supply decoupling capacitors. Amirm says there is experience of them failing. The specs look pretty decent, in terms of ESR, which is main spec of interest here. I'm assuming they are backed up in the usual manner, with smaller, higher frequency caps to give a broad spectrum decoupling. I suspect the concern is for reliability, rather than sound quality.

On power supplies, most linear ones available on ebay for audio and RaspPi are 25VA 3A 5V DC ones. Are they suitable for the 1A requirement of the S2?
The supply needs to be 5V, at a minimum of 1A (or a little less, considering it is supplied with a 1A SMPS, and therefore the unit must take less than 1A). Thus, your 3A linear supply will be more than adequate. Make sure it's a linear regulated PSU, and not just a transformer, rectifier and capacitor.
 

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
20,620
Likes
25,441
Location
Seattle Area
#24
Would actually changing the capacitors improve sound quality (or the quality of the current they deliver) or you meant a measure to prevent failure?
Hi there. Welcome to the forum. I meant failure there. I would not mess with capacitors for difference in sound.
 
Joined
Mar 29, 2018
Messages
21
Likes
13
Location
Bilbao, Spain
#25
Which transformer power would be optimal or more efficient for the S2? I have found now 25VA (3.5A), 15VA (2.5A) and 9VA (1.2V) linear regulated PSUs at 5V, all deliver over 1A of the requiered input. I guess one of them would make the transformer work its best in its parameters.

As for the capacitors, since the designer of the S2 complained about their change on its specs on the first batches of production, I thought some could have better performance at the analog power stage since users agree to note the sound as a bit lean and lacking dynamics possibly due to the low 5V current and 2.05 Veff output voltage.

I am no expert but since they may have tried to cut down on costs and sometimes diy people change capacitors in amplifliers for better sound, as I would really like this product to sound its best possible! I was thinking maybe with linear regulated psu (which improves sound) and better capacitors it could reach its potential.
 
Last edited:

trl

Addicted to Fun and Learning
King of Mods
Joined
Feb 28, 2018
Messages
672
Likes
435
Location
Iasi, RO
#26
[...]Would the headphone amp perform better with a higher supply voltage (12V)? A "high power mode" would of course require an external PSU. [...]
9602 amplifier is getting power from a "single +3.3V power supply", per ESS wrote, so output power has nothing to do with increasing the input voltage in this case.

However, there are 2 boost or buck converter inside that either boost or lower the original 5V to a different voltage and perhaps does virtual ground as well...can't say for sure. Maybe it splits the power rails, but only a Pre-Box S2 owner with a DMM or a scope could tell us more (or maybe some macro/close-up pics?). It would be interesting to find out why does this combo needs these boost or buck converters for.

Here's a decent view with lot of explaining too: https://www.computeraudiophile.com/...42am.png.654fe81e8c50a0979b702f38b0e5c79e.png. Notice the boost/buck converters created by the FET switching and the big coils.
 
Joined
Feb 9, 2018
Messages
148
Likes
77
#27
Switching supplies are probably for the digital processing ICs.

The inductors look far bigger than they need to be, given the sort of current requirement for the digital devices present (they're all small, unheatsinked devices, therefore cannot be taking much power).
 

DonH56

Technical Expert
Technical Expert
Patreon Donor
Joined
Mar 15, 2016
Messages
3,011
Likes
3,587
Location
Monument, CO
#29
The switching regulators on the PCBs we use are small compared to the inductors but are outputting 5 - 15 A of current albeit around 1 V. The high efficiency of modern SMPS ICs is pretty amazing and allows heat sinking through the pad under the device so they can handle the load without an external heat sink. Alas, the parts we design need heatsinks, often with a fan attached...
 

trl

Addicted to Fun and Learning
King of Mods
Joined
Feb 28, 2018
Messages
672
Likes
435
Location
Iasi, RO
#30
Switching supplies are probably for the digital processing ICs.
[...]
.
I would anytime choose for regular low drop-out and very low-noise TPS 5V to 3.3V/2.5/etc. converters and not for buck converters. Most likely there're some negative rail over there....

Switching supplies are also needed for anything that requires higher voltage/negative voltage than what is coming in. This is needed usually for the headphone amp.
Seems that a low-noise & low-drop-out 5V to 3.3V regulator would do for the headamp chip, based on http://www.esstech.com/index.php/tw/products/audio-amplifiers/.
 
Joined
Feb 9, 2018
Messages
148
Likes
77
#31
I would anytime choose for regular low drop-out and very low-noise TPS 5V to 3.3V/2.5/etc. converters
For the analogue electronics, yes.

Digital electronics doesn't need a low noise supply.

The choice between linear and switching for digital devices is down to issues of efficiency, and supply load. You won't find a linear regulator used for the core voltage supply for a CPU, for instance.

The switchers may indeed be used for generating a negative rail.

Those big inductors may even be supply filters...

Without the schematic, or probing/tracing the cct, we're all just guessing...
 

trl

Addicted to Fun and Learning
King of Mods
Joined
Feb 28, 2018
Messages
672
Likes
435
Location
Iasi, RO
#32
[...]You won't find a linear regulator used for the core voltage supply for a CPU, for instance.[...]
That's because it would be totally inefficient and nonsense for a CPU.
On the other hand, I've seen 80-100VA power consumption from valve headamps delivering 2 x 2.5W/channel. Audiophiles are like crazy sometimes. :)
Also, I own a Class-A transistors headamp eating up to 30VA from the mains, while delivering at most 2x4.5W.
 
Joined
Mar 29, 2018
Messages
21
Likes
13
Location
Bilbao, Spain
#33
What would you think of replacing the 13 supposedly low quality Lelon electrolytic capacitors with these premium ELNA audio-specialised silk-fiber electrolytic ones, same specs (6.3V, 1,000 uF)? (Since other people are complaining about these caps, and also they have pointed out a lean, low dynamic sound)
https://www.ebay.com/itm/182131651068 (2.27€ each)
Even if they are used for analog power stage, would they benefit with better sound quality, in terms of better current or low ESR? Maybe ceramic capacitors?
What about the other Wima capacitors used in alalog part?

(Excuse me if I'm a noob, I wonder if this product's sound quality would improve with better components)
 

Wombat

Major Contributor
Joined
Nov 5, 2017
Messages
3,883
Likes
2,174
Location
Australia
#34
What would you think of replacing the 13 supposedly low quality Lelon electrolytic capacitors with these premium ELNA audio-specialised silk-fiber electrolytic ones, same specs (6.3V, 1,000 uF)? (Since other people are complaining about these caps, and also they have pointed out a lean, low dynamic sound)
https://www.ebay.com/itm/182131651068 (2.27€ each)
Even if they are used for analog power stage, would they benefit with better sound quality, in terms of better current or low ESR? Maybe ceramic capacitors?
What about the other Wima capacitors used in alalog part?

(Excuse me if I'm a noob, I wonder if this product's sound quality would improve with better components)

I don't understand why gear is purchased if it doesn't perform to expectations. Cut your losses. Sell what you have and buy something better. Modding is usually like putting lipstick on a pig. :oops:
 

DonH56

Technical Expert
Technical Expert
Patreon Donor
Joined
Mar 15, 2016
Messages
3,011
Likes
3,587
Location
Monument, CO
#35
What would you think of replacing the 13 supposedly low quality Lelon electrolytic capacitors with these premium ELNA audio-specialised silk-fiber electrolytic ones, same specs (6.3V, 1,000 uF)? (Since other people are complaining about these caps, and also they have pointed out a lean, low dynamic sound)
https://www.ebay.com/itm/182131651068 (2.27€ each)
Even if they are used for analog power stage, would they benefit with better sound quality, in terms of better current or low ESR? Maybe ceramic capacitors?
What about the other Wima capacitors used in alalog part?

(Excuse me if I'm a noob, I wonder if this product's sound quality would improve with better components)
Are we still talking about the Pro-ject S2? I t did well in the tests, I would not feel a strong need to mod....

Those capacitors are probably decoupling the power supplies, not in the signal path, so I am suspicious if anyone said they make it sound "lean". Lower ESR is sometimes of benefit, but can also lead to noise peaking depending upon the power supply circuit.

Silk, or silk and manilla hemp per the website, would not be my choice for electrolytic capacitor dielectrics. Hard to produce without pinholes that reduce performance (and ultimately cause failure), and using a material that can absorb moisture and thus increase leakage seems like a poor choice.

Ceramic capacitors make nice small high-frequency decoupling capacitors but their value changes with voltage making them a poor choice for coupling capacitors in the audio path. Electrolytics are not very good in the signal path either, of course.

I don't think this pig needs lipstick.
 

trl

Addicted to Fun and Learning
King of Mods
Joined
Feb 28, 2018
Messages
672
Likes
435
Location
Iasi, RO
#36
Those "supposedly low quality" are Lelon OCRZ ULR polimer series or X-CON ULR polimer series (I've seen same DAC with both ver. of caps)

Lelon or X-CON vs. Silmic II (same 1.000uF/6.3V):
2.000h
of operation instead of 1.000h
105C instead of 85C (overkill anyway for this DAC)
6.100mArms/X-CON or 5.500mArms/OCRZ vs. 710mArsm of rated ripple current (quite a huge difference!)
8/12mm vs. 12.5/20mm, so Silmic II will not fit

I would vote anytime with the polimer caps, no doubts here. Also, those WIMA are probably MKS for the low-pass filters...however, are just fine planted there on the PCB, they're not adding noise or harmonics or anything like that. This DAC combo measures very well, not sure there's a real need for any upgrade, unless really want a linear PSU added on the backside.
 

amirm

Founder/Admin
Staff Member
CFO (Chief Fun Officer)
Joined
Feb 13, 2016
Messages
20,620
Likes
25,441
Location
Seattle Area
#37
What would you think of replacing the 13 supposedly low quality Lelon electrolytic capacitors with these premium ELNA audio-specialised silk-fiber electrolytic ones, same specs (6.3V, 1,000 uF)? (Since other people are complaining about these caps, and also they have pointed out a lean, low dynamic sound)
My comments are strictly about long-term reliability, not sound. If measurements are good, which they are in this case, there is no reason to change them when they are new.
 

trl

Addicted to Fun and Learning
King of Mods
Joined
Feb 28, 2018
Messages
672
Likes
435
Location
Iasi, RO
#38
Oh, now I understand why Diego Humanes was asking about caps replacement...it's related to your first post.
BTW, you should see the Cap-Top how "good" they are in the PSU. Flower after flower... :)
 
Joined
Feb 9, 2018
Messages
148
Likes
77
#39
That's because it would be totally inefficient and nonsense for a CPU.
And equally nonsense for the supply of a digital signal processing element (the XMOS chip)...

Low noise linear PSU for the DAC analogue side? Certainly.
Low noise linear PSU for the DAC digital side? Possibly; may help to reduce coupling of noise from digital to analogue side (even though the digital side is implicitly noisy, simply due to the digital input and any upsamplig/interpolation filtering it might be doing).

But, as you say, audiophiles are a bit crazy...
 

trl

Addicted to Fun and Learning
King of Mods
Joined
Feb 28, 2018
Messages
672
Likes
435
Location
Iasi, RO
#40
[...]

But, as you say, audiophiles are a bit crazy...
Designers already know that. :)

Just my 2 cents here, very low-noise regulators are a must for the digital part, no matter it's about the XMOS transport chip or the DAC or ADC chips. This will reduce noise floor and jitter as well (see AKM DAC chips specs, but also on the PCM chips the recommended layout and design specs). I think I wrote some design webpages from the DIYINHK and found nice compare table between the regulators and also how to route the PCB traces correctly and why regulators for the digital part need to be very -low noise.

As for the analogue part...simple old school regulators could do the job pretty well, because most of today opamps are having a CMRR/PSRR of 110-140dB, so ripple or noise injection from the power supply should not be a problem (unless the PSU is really bad or the PCB is totally crappy).
 
Top Bottom