or Connect
Head-Fi.org › Forums › Summit-Fi (High-End Audio) › High-end Audio Forum › Audio-Gd Master 7 - Discrete Fully Balanced DAC (PCM1704)
New Posts  All Forums:Forum Nav:

Audio-Gd Master 7 - Discrete Fully Balanced DAC (PCM1704) - Page 86

post #1276 of 3498

Questions most definitely welcomed and no I don't feel threatened if anyone questions what I am saying. Quite the opposite - I'm gratified that someone's paying attention enough to ask!

 

Firstly your question contains a huge 'if' which doesn't get borne out in practice. No two ground potentials are ever the same - at least never close enough for great audio performance. It might be true at DC but at AC frequencies which give us such grief (ultrasonic switching frequencies from PSUs) it can't ever be so due to inductance and skin effect. Well it could be so if there was zero AC current flowing - but that (drawing zero current) would mean we're not using the ground at all, so it might as well not be there.

 

I'm not familiar with ISO transformers so not at all knowledgeable about what benefits they might bring. Swapping the power supply in the computer is most definitely an option, but not for the faint-hearted. A linear power supply, preferably one using a split bobbin EI core transformer would be a big help - problem is its much bulkier and heavier Onot to mention runs hotter) than the unit it replaces, so might well need to be housed externally. But in reality the optimum solution is just to get some decent common-mode choking on the mains input - decent meaning ones with good rejection at the frequencies coming out of SMPSUs. For which purpose I've designed my own, not finding them in the current catalogs. More than happy to share details of the choke if anyone's a serious anorak enough to want to make one.:evil:

post #1277 of 3498
Yes and thanks for replying. Yes please explain how to make the choke as I know what it is . It's been a really long time since I did the formular to make one. But an ISO is a choke . It is a transformer witha primary and secondary but the two wininfpds are isolated and they have a one to one ratio. So in effect they act like a choke for noise comiing from the mains . But to make them code compliant you have to have a common gnd from input to output. Now what you are talking about I think is a choke but have it tuned to have a action at the freq of the noise . Well I used to know how to disign that many years ago. But unless you used a scope how do you determin the freq of the noise and it's harmonics ?

Al
post #1278 of 3498

OK so I got what an ISO is - its not actually a choke rather an isolation transformer. A choke is just like a 1:1 transformer but wired differently into the circuit (at 90 degrees schematically speaking) so the input is to both the primary and secondary terminals (one side of each winding) together. The output is taken from the other side of the primary and secondary windings. So with a transformer you get perfect isolation at DC and progressively worse isolation at AC - with a choke its the opposite. Zero isolation at DC and better and better isolation as the frequency goes up. So the best solution is a transformer, followed by a choke (or the other way around). Which is in fact what's used in ethernet connections to reduce noise - albeit ethernet transformers and chokes are very small because the frequencies are so high.

 

So about the choke - its certainly one way to design a choke if you know the noise frequency, you can 'tune it'. That's not how I've designed mine - rather I went for the broadest possible band of rejection. This is because the higher the frequency the worse the audio effect, and also the easier it couples through stray capacitances. What I did I haven't adapted to typical desktop PC power levels so far - it works on a 100W switching supply brick I use to power a chipamp. 100W is plenty enough to power a laptop computer though. It could be scaled up by using thicker wire.

 

My construction took a segmented plastic former used to make power transformers for SMPSUs. The multiple segments are needed to get good HF rejection because stray capacitance between the windings is what kills the high frequency performance of a choke. If you go to the suppliers of common-mode chokes (Murata is a good one) you'll see they make chokes in a couple of styles - normal, and segmented. The latter post better figures at HF by 10dB or more. But they only use two segments. I figured - 'why not use more segments?'. So the coils I have built use 7 segments - just because those are the formers I was able to find.

 

Having found a multiple segment former, the next step is to wind it. I experimented with using enamelled copper wire for this but did so with trepidation. And my fears were borne out - its definitely not suitable for working at mains voltages in this kind of choke. That's because I use it twisted, and twisting the wires together comrpromises the insulation - minute cracks form in the coating. Since one of my chokes caught fire briefly due to this I've gone away from this kind of wire. But it proved the concept was sound while it worked - which was for a few months.  I now have some tri-insulated wire which has three layers of insulation (rather than one) so should be safe. I've built some chokes with this wire but yet to run them up. The tri-insulated wire is thicker so I couldn't fit as many turns on the former as with the original enamelled wire, so when I do have a listen I might find I'll need to go to a bigger former to get the same performance.

 

Does any of this help much? :cool:

post #1279 of 3498
It makes sense and I follow it , why do you have to wind it. Is there now vender to buy ready made chokes of varing specs. Now if I remember any choke theory then current is what makes them work and as such you need to have a basic current value to design one. I am interested to see how it pans out for you. How about using a ups unit . Simpler to use. I know it generates harmonics but should not matter to a CPU. For me I have ps audio regenerator 2 P10 ,s and one P5. It makes a big difference even for a plasma tv you can see the change in the video very easily .

Thanks
Al
post #1280 of 3498

I have to wind them because no-one else is doing so yet. But after I've proven a few with the tri-insulated wire, I'll probably get a local factory to make a batch. I'll want to characterize the performance though first, which might mean I need a spectrum analyser (something I don't currently have).

 

A UPS is just too bulky for what I want and they'll need a choke in the offline charger part anyway.

post #1281 of 3498
I understand , but my point is , it will tell you quickly if the noise is coming from the PSU in the CPU and not the line. .
As it is generating in the ups.

Al
post #1282 of 3498

Finally have come full circle on I2S with the Off Ramp 5 and Master 7.  What a pain in the ass but I learned so much and made it a worthwhile journey.  The OR5 is repaired and HDMI I2S adapter board working reliably in the Master 7 (5 days). Sound quality is consistent as expected, signals at the M7 DSP board excellent, and no surprises. Want to thank forum members Alrainbow and jacal01 who kept me going really or would have probably given up. Steve Nugent at Empirical Audio repaired the OR5 under warranty so a big thank you to him. (Don't hot plug the I2S cables!!!!!). Also thanks to Paul McGowan, CEO at PS Audio, who answered a lot of dumb questions and is very supportive of this HDMI I2S effort.

 

While the OR5 was in Oregon on life support I got reaquainted with USB32 and the older firmware as well as the new V2 version. V2 moves the SQ bar quite a bit. The OR5 has slightly better bass, more detail, and smoother but not by a long shot as was the case prior to V2.  Perhaps still less jitter with the OR5.  The OR5 always had less silibance and more inner detail but now consistent day to day via HDMI I2S. 

 

Here's a few pictures of the M7 with the HDMI I2S adapter board (green colored) mounted in place of the RJ45 board. Luckily discovered an easy source of 3.3V/GND to power the HDMI receiver chip. There are two spare power pins on the header supplying power to the USB32 board (red/black wires). The multi-colored ribbon cable supplies I2S to the DSP board. It is temporary until something better comes along. Thanks again for your interest. I won't be offering upgrade kits for many reasons but will supply anyone interested with all documentation and knowledge.  The M7 has never sounded better and with the HDMI interface ready for the future!

  

 

 

Power picked off spare header pins (red/ black wires) that also supplies power to USB32

 

Short 0.3m Wireworld HDMI cable.  Bought at a bargain price on Fleabay.


Edited by DACLadder - 2/6/14 at 7:24am
post #1283 of 3498
Thanks for the help and better understanding
With the offramp and products. But here is some food for thought
As our quest for better sound leads on a digital journey
Donyouveonder what kingwa did to make the same USB board sound better
I would doubt he could make the clock chip better
With software. So just what did he do to make it better
And further more why wait almost two years to do.
One answer is USB is getting better so he needed to keep up with the others
But overall what software makes the change and why
Al
post #1284 of 3498
Quote:
Originally Posted by ALRAINBOW View Post

Do you ever wonder what Kingwa did to make the same USB board sound better ? I would doubt he could make the clock chip better with software. So just what did he do to make it better ? And further more why wait almost two years to do ?

But overall what software makes the change and why
Al

Hey Al, I can't be sure, but I would bet he's learning as he goes. Any good designer or engineer is going to do their best on a statement product when they try. And then hopefully if they're an even better engineer they're going to learn more skills and techniques as they work and get better and better. I bet Kingwa is always doing trial and error on his products and here with the new firmware he's really learned a new approach and come up with something great! 

 

Since Kingwa is updating and releasing the firmware for free, there's no financial incentive to release an inferior product now only to release a better version later for free.

 

That's my $0.02 anyway.

post #1285 of 3498
Quote:
Originally Posted by DACLadder View Post
 

Finally have come full circle on I2S with the Off Ramp 5 and Master 7.  What a pain in the ass but I learned so much and made it a worthwhile journey.  The OR5 is repaired and HDMI I2S adapter board working reliably in the Master 7 (5 days). Sound quality is consistent as expected, signals at the M7 DSP board excellent, and no surprises. Want to thank forum members Alrainbow and jacal01 who kept me going really or would have probably given up. Steve Nugent at Empirical Audio repaired the OR5 under warranty so a big thank you to him. (Don't hot plug the I2S cables!!!!!). Also thanks to Paul McGowan, CEO at PS Audio, who answered a lot of dumb questions and is very supportive of this HDMI I2S effort.

 

While the OR5 was in Oregon on life support I got reaquainted with USB32 and the older firmware as well as the new V2 version. V2 moves the SQ bar quite a bit. The OR5 has slightly better bass, more detail, and smoother but not by a long shot as was the case prior to V2.  Perhaps still less jitter with the OR5.  The OR5 always had less silibance and more inner detail but now consistent day to day via HDMI I2S. 

 

Here's a few pictures of the M7 with the HDMI I2S adapter board (green colored) mounted in place of the RJ45 board. Luckily discovered an easy source of 3.3V/GND to power the HDMI receiver chip. There are two spare power pins on the header supplying power to the USB32 board (red/black wires). The multi-colored ribbon cable supplies I2S to the DSP board. It is temporary until something better comes along. Thanks again for your interest. I won't be offering upgrade kits for many reasons but will supply anyone interested with all documentation and knowledge.  The M7 has never sounded better and with the HDMI interface ready for the future!

  

 

 

Power picked off spare header pins (red/ black wires) that also supplies power to USB32

 

Short 0.3m Wireworld HDMI cable.  Bought at a bargain price on Fleabay.

And DACLadder, that's absolutely awesome news! Congrats, man! I'm really happy to hear that HDMI I2s is possible on the M7 if you've got the gusto to go the distance. So, any difference between HDMI-I2s input and RJ-45-I2s input on the M7 via OR-5??

 

I've just bought a PWD Mkii, so I doubt I'll be moving to M7 territory anytime soon, but it's good to know. I think I'll be buying/demoing an Audiobyte Hydra-X+ USB->HDMI-I2s by the fall and that's cool that if I ever did make my way to the M7, which I REALLLLLLLLLY thought long and hard about when I pulled the trigger on the PWD Mkii, that the Hydra-X+ can do it's job there too :-D

post #1286 of 3498

DACLadder

 

I'm wondering if Kingwa would be wanting to change the i2s rj45 to HDMI if requested (don't think it will be for free lol)


Edited by mowglycdb - 2/6/14 at 11:18am
post #1287 of 3498
No I agree with you. Psaudio does the same thing
But most dacs there is no upgrades in firmware
Or equipment. It is good to see some makers have a better approach

Al
post #1288 of 3498
Quote:
Originally Posted by Benny-x View Post
 

[snip]

I recently got the pin-out from the PWD Mkii HDMI-I2s from PS Audio because I was looking at making up my own LVDS->HDMI I2s cable to feed from an Audio-gd DI-V3 I2s output to my PWD Mkii HDMI-I2s input. I don't know much about the format, but as soon as I saw what was going on I gave up on the idea.

 

You know, now that I've been on a voyage of discovery with DACLadder, what you're looking for I2S LVTTL to LVDS via HDMI is basically the reverse of the same schematic as DACLadder's LVDS/HDMI to LVTTL converter board, ±HDMI pin 19 grounding, and you would mount it in the ADG DI-V3 side.

 

But it is doable, and DACLadder's converter board would be a big leg up, if the circuitry and chip also works in reverse. If not, the PSA PWT I2S HDMI output schematic here should work just fine as a DIY project.


Edited by jacal01 - 2/6/14 at 8:11pm
post #1289 of 3498
Quote:
Originally Posted by DACLadder View Post
 

Also thanks to Paul McGowan, CEO at PS Audio, who answered a lot of dumb questions and is very supportive of this HDMI I2S effort.

 

Was that a not so subtle shot at me?  :wink:

post #1290 of 3498
Quote:
Originally Posted by mowglycdb View Post
 

DACLadder

 

I'm wondering if Kingwa would be wanting to change the i2s rj45 to HDMI if requested (don't think it will be for free lol)

Don't know about that change, but I ordered my Master 7 and asked Kingwa to directly solder the i2s rj45 cable into the board instead of having the port, and he didn't charge me anything.

 

Will be getting my master 7 in mid Feb :) 

New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: High-end Audio Forum
Head-Fi.org › Forums › Summit-Fi (High-End Audio) › High-end Audio Forum › Audio-Gd Master 7 - Discrete Fully Balanced DAC (PCM1704)