Not really,  SS power supply, no galvanic isolation, no dedicated RBCD clock and you're stuck with a S-D DAC.  Not interesting to me at all.  
 
The only USB protocol that is potentially superior to the modded Hiface is the Ayre because it hase galvanic isolation from the USB before the I2S output.
 
 
 
Quote:

davebsc said:

Right, but a modded EVO would give you the option to use Toslink or whatever digital connection you choose, without the need for any adapters. I'm curious how the modded Hiface would stack up against a DAC with built in 24/192 asynch USB like the W4S DAC-2. That would be an interesting experiment.

Click to expand...

Anybody know diddly about the ST optical output included on the Evo. It seems a strange move as only EMM & Wadia DACs use it (but I guess the upcoming M2tech DACs will also). But apart from it being a different mechanical optical connector is there any sonic advantage to it?
 
The only bit that I could find was from here: http://www.stereophile.com/cdplayers/189/index10.html
 
Quote:

These are glass guys. Off-the-shelf integrated glass optical transmitters combine the driver circuitry and laser transmitter into one unit. But according to Wadia, such transmitters are intended for use in long-distance telecommunications; when used in short audio lengths, they overload the receiver and cause jitter. So Wadia designed a discrete glass optical transmitter tuned for shorter lengths which, they claim, lowers jitter. They say the improvement can be heard in any D/A converter with a glass-fiber input.

Click to expand...

This seems to indicate that a proprietary driver is need to avoid the high jitter of standard optical transmitters/receivers! Any ideas - is there a proprietary ST chip in the Evo?

Are you sure it doesn't have galvanic isolation? I was looking at the 6moons preview, and they said all digital inputs other than Toslink are isolated. The DAC is the Sabre 9018 which isn't exactly a bad one. What is the Ayre using?

I've received my new DIT chip and repaired my Hiface.   What are the thoughts on the external linear regulator +5V and +1.8V power supply?   I'm thinking that since it shares ground with the USB a big ground loop is immenant,   I'm thinking about forgoing this part of the mod?

This 1.8V LDO regulato rMCP1700T-1802E/TT works fine with a 3.3V input from the batteries,  I'm thinking about forgeting about the +5V mod and just using this reg on the batteries for the 1.8V.  Jenky you may want to consider this as its only a $1 cost adder?
 
I'm  going to start tonight,  any opinions appreciated.

Hey Regal, Jkeny, etc.
 
I think it would be interesting for you guys to design and build your own USB converter. Have you ever considered this?  
 
I dont mean this in a negative way, because I think your hiFace mods are incredible, but I would love to see a custom design from the ground up with your knowledge.  

Impossible,  it takes months to program an asynch converter,  huge expenses.  Its nothing like the PCM2707 USB recievers of old.
 
 
Quote:

zorlac said:

Hey Regal, Jkeny, etc.
 
I think it would be interesting for you guys to design and build your own USB converter. Have you ever considered this?  
 
I dont mean this in a negative way, because I think your hiFace mods are incredible, but I would love to see a custom design from the ground up with your knowledge.  

Click to expand...

Yes Regal, I too am concerned about ground loops although maybe the galvanic isolation on the output of the Hiface mitigates this?
 
I believe that the 5V PS is a worthwhile target as it powers the DIT4192 chip that generates the SPDIF signal! My experiments have shown that there is a sonic benefit of about 5% when compared to the sonic benefit of 95% for the 3.3V mod.
 
I'm interested in your 1.8V experiments - I didn't do this as I reckoned that it was just powering the core logic of the Xilinx chip. I reasoned that the main benefit might be in removing the Hiface on-board 1.8V regulator & whatever noise it might be putting out. But, I'm open to being wrong & interested in the outcome of your experiments.
 
When I look at regulators, the criteria I look for in digital circuits is the noise figure. looking at the MCP datasheet is not very encouraging - it's noise Vs freq plot is abysmal, up around 8/9 mV/SQRT Hz  at 0.01KHz. I think we need as low a noise figure at low Hz. I would prefer something like LT1763 at 20uV from 10Hz to 100KHz.

Quote:

jkeny said:

If you can find a quality TX & RX  for coax please tell - even more so for Toslink 

Click to expand...

You don't like the quality of the current units on the HiFace?
As for toslink, it seems that many attribute the issue with that connection to the TX and RX used.
What is your opinion on that?

Quote:

robscix said:

You don't like the quality of the current units on the HiFace?
As for toslink, it seems that many attribute the issue with that connection to the TX and RX used.
What is your opinion on that?

Click to expand...

I don't know about the Evo .
That would seem to be the issue alright.
 

Quote:

jkeny said:

Yes Regal, I too am concerned about ground loops although maybe the galvanic isolation on the output of the Hiface mitigates this?
 
 

Click to expand...

If you look at the stereophile measurements on transport jitter,  they all show a hump in jitter at 60hz,  this tells me the jitter is there from a ground loop issue before the pulse transformer and won't go away.
 

Quote:

regal said:

If you look at the stereophile measurements on transport jitter,  they all show a hump in jitter at 60hz,  this tells me the jitter is there from a ground loop issue before the pulse transformer and won't go away.
 

Click to expand...

But what if you run a laptop on batteries as your source? I would expect this to kill the ground loop issue. So now we get into a specific computer for music playback which is no bad thing as I don't believe we should be using our general purpose computer as our music transport but rather a dedicated & optimised one.
 

I am finally done with this project.  I ended up replacing the stock BNC because it didn't have the same connector as a true amphenol 75 ohm BNC.  I tried the linear power supply for +5V and 1.8V and heard no improvement so I removed the external powersupply to make it simple.  The battery holders are nice and the form factor is very small to allow hooking straight into the DAC.  The sound is really really phenomonal.
 
 

thats an idea,  I'm not a laptop person so someone else will have to try,  definately worth looking into.  I'm moving back to building tube amps.  Desoldering/installing that DIT just about did me in for SS/digital projects for a while
 
I highly recommend anyone doing this mod to use flexible small wire,  they make it for the model train hobbiests.
 
 
Quote:

jkeny said:

But what if you run a laptop on batteries as your source? I would expect this to kill the ground loop issue. So now we get into a specific computer for music playback which is no bad thing as I don't believe we should be using our general purpose computer as our music transport but rather a dedicated & optimised one.
 

Click to expand...

I'm not terribly surprised that you didn't notice any improvement with the 1.8V external supply for the reasons I gave above but I thought the 5V would have seen some benefits?
 
Anyway, it's great that you are happy with the sound - it truly is high-end, I think you'll agree!
 
You gave up the ideas of changing the output stage?
 
Your next adventure, & this should appeal to you as an admitted  frugophile, is to try attenuators on the output!!