[mowglycdb]

A WW platinium seems too expensive D:  a Chroma seems more logic, given that if he's going to use the minimum length that's 0.3m there shouldn't be much diference.

 

[drez]

Yeah the shorter the better for signal integrity, cheaper also.  WW Platinum are very expensive (same as Audiophilleo 2), but performance is the best I have heard so far.  For value the Audiophilleo is probably better, actually a USB card is better also, but the performance/price does at some point catch up I think, like when you already have the best DAC, AMP, USB transport etc. and still want more SQ.  
 
Say if there are 2 DAC's and one is better than the other, the WWPS probably won't make the lesser DAC better than the more expensive one, but might close the gap a little.  Whether than makes sense depends on the gap in prices.
 
Does it make sense for Master 7.  Hmm not sure, not enough experience here to say.

 

[preproman]

  So I've decided to give the DI-2014/HDMI out to M7/HDMI. Placed my order for DI-2014 with TXCO and HDMI out upgrade and the HDMI input board for DIY install into my M7.
 
Rest of config is Audio-GD Master 10 amp into Magnepan 3.7R. Double Helix USB and ACSS cables. Source is dual I5 CAPS PC's running WS2012, Audio-optimiser 1.30 and jplay. Currently using Audiophilleo 2 and Purepower for USB to BNC into M7 so will be interesting to compare to DI-2014/HDMI.
 
This just leaves a decision to be made on HDMI Brand/Model and length in time for the arrival of the kit from Audio-GD. Double-Helix do not do a HDMI cable otherwise I would have stuck with that brand for my cables. Any nominations for HDMI cable ?
 
Frizzup

 
Very nice rig you have there Frizzup.
 

 

[frizzup]

Thanks for the feedback guys. 
 
The Wireworld Platinum is NZ$1100 for 0.5M. I can get the Red Starlight for NZ$279 for 0.5M. Local dealer happy to let me trial 1M lengths off the shelf across Chroma, Ultraviolet and Starlight. Then having chosen the model he would have to order 0.3M lengths as they are not held in NZ stock....they are about 15% less than the 0.5M lengths across the Wireworld range..
 
The trial with the Di-2014/HDMI to M7 setup may not improve on my 2 year old AP2/Purepower to M7 setup but wanted to experiment with both with I2S and 192khz content. The improvements by Audio-GD with the latest USB32 firmware and drivers, the release of the DI2014 and then the release of the i/o HDMI boards made for an attractive total package that had me wondering how that might sound with the M7. Certainly a cheaper experiment than Offramp5 +options +cable +import taxes ~ NZ$2750.  
 
I can always stick with the AP2/PP on my M7 and use the DI-2014 as the converter to my home office Audio-GD Reference 5.2/C2.2 Headphone rig. The Reference 5.2 only has the TE8802 USB chip input so the DI-2014 will be a step up. Upgrading the Reference 5.2 to USB32 was going to cost about US$80.  
 
So should have this new setup in a couple of weeks. Burn-in time for the DI-2014 and HDMI boards + up to 3 different HDMI cables before settling on one then ordering the shortest length, then some comparisons (subjective)....report back in 3 months or so 
 
Frizzup

 

[DACLadder]

However, I'm confused by your statement that each circuit (channel) ground (return) is connected by (in?) the cable.  That would seem to defeat to whole purpose of eliminating common mode interference in balanced signal cable.  Do you mean at the connector(s)?  I would think that the individual channel ground conductors would be grounded together in either the transmitter (e.g. DDC) or receiver unit (DAC), and leave the individual channel cable grounds insulated from each other over the entire exposed cable length.  What am I missing?

OK I will give it a shot.  In differential HDMI LVDS there is no ground return on one of the two signal pairs as needed with single ended schemes like RJ45 I2S.  The LVDS return current is through the opposite signal wire of the pair - both above ground.  Each of the four signal pairs are shielded and these shields are connected to ground on both sides of the cable in HDMI I2S (pins 2, 5, 8, 11) plus pin 17 which is defined as digital ground in HDMI world.  The LVDS receiver chip needs the ground reference of the transmitter in order properly decode the state of the presented LVDS signal.  LVDS is fast with low jitter due to a large part the signal swing is only +/-350mV around a bias point of about 1.2V above ground.   In single ended LVTTL as with RJ45 the single swing is 10 times as much at about 3V.  Lower signal swing means less ambiguity in decoding a logic state change which results in lower jitter. Texas Instruments has a great publication on LVDS and in the back of the document talk about jitter with emphasis on HDMI application.   http://www.ti.com/lit/ml/snla187/snla187.pdf

 

[DACLadder]

  So I've decided to give the DI-2014/HDMI out to M7/HDMI. Placed my order for DI-2014 with TXCO and HDMI out upgrade and the HDMI input board for DIY install into my M7.
 
Rest of config is Audio-GD Master 10 amp into Magnepan 3.7R. Double Helix USB and ACSS cables. Source is dual I5 CAPS PC's running WS2012, Audio-optimiser 1.30 and jplay. Currently using Audiophilleo 2 and Purepower for USB to BNC into M7 so will be interesting to compare to DI-2014/HDMI.
 
This just leaves a decision to be made on HDMI Brand/Model and length in time for the arrival of the kit from Audio-GD. Double-Helix do not do a HDMI cable otherwise I would have stuck with that brand for my cables. Any nominations for HDMI cable ?
 
Frizzup

I have a couple of Wireworld HDMI cables - Starlight 5 and Silver Starlight 6.  The main reason for purchasing was that they are the shortest cables I could find at 0.3m (1 ft).  Honestly I don't hear any difference between these two cable types probably because they are very short.  Cable quality becomes an increasing factor in performance with longer cables.  So use the shortest cable first and then best quality you can afford.  But HDMI I2S is way less tweaky versus RJ45 I2S as far as cables.
 
Let us know how the DI-2014 works out.  Very interested now with the HDMI I2S option.

 

[jacal01]

  OK I will give it a shot.  In differential HDMI LVDS there is no ground return on one of the two signal pairs as needed with single ended schemes like RJ45 I2S.  The LVDS return current is through the opposite signal wire of the pair - both above ground.  Each of the four signal pairs are shielded and these shields are connected to ground on both sides of the cable in HDMI I2S (pins 2, 5, 8, 11) plus pin 17 which is defined as digital ground in HDMI world.  The LVDS receiver chip needs the ground reference of the transmitter in order properly decode the state of the presented LVDS signal.  LVDS is fast with low jitter due to a large part the signal swing is only +/-350mV around a bias point of about 1.2V above ground.   In single ended LVTTL as with RJ45 the signal swing is 10 times as much at about 3V.  Lower signal swing means less ambiguity in decoding a logic state change which results in lower jitter. Texas Instruments has a great publication on LVDS and in the back of the document talk about jitter with emphasis on HDMI application.   http://www.ti.com/lit/ml/snla187/snla187.pdf

 
So beyond my equating the channel signal return with ground, I was more or less on the right track, methinks.  And your "Each circuit ground is directly connected together by the cable." was referring to the separate reference ground conductor (pin 17) connecting transmitter and receiver?  You weren't talking about the shielding grounds, were you?
 
The 1.2 Vdc signal bias then is kept isolated from ground throughout?  Would there be some logic pull down resistor involved at some point?
 
Thanks for the further reading, BTW.

 

[sandab]

  OK I will give it a shot.  In differential HDMI LVDS there is no ground return on one of the two signal pairs as needed with single ended schemes like RJ45 I2S.  The LVDS return current is through the opposite signal wire of the pair - both above ground.  Each of the four signal pairs are shielded and these shields are connected to ground on both sides of the cable in HDMI I2S (pins 2, 5, 8, 11) plus pin 17 which is defined as digital ground in HDMI world.  The LVDS receiver chip needs the ground reference of the transmitter in order properly decode the state of the presented LVDS signal.  LVDS is fast with low jitter due to a large part the signal swing is only +/-350mV around a bias point of about 1.2V above ground.   In single ended LVTTL as with RJ45 the single swing is 10 times as much at about 3V.  Lower signal swing means less ambiguity in decoding a logic state change which results in lower jitter. Texas Instruments has a great publication on LVDS and in the back of the document talk about jitter with emphasis on HDMI application.   http://www.ti.com/lit/ml/snla187/snla187.pdf

Are you sure HDMI is DC around a 1.2V offset?  I always thought the pairs were AC with opposite polarity.  Depends on whether the CML rails are +/- or + and GND, but the whole idea of CML is to avoid switching noise by running a fixed current (3.5mA, which over 100ohm at 340MHz translates to 350mV).  Also, looking at eye graphs they're always +/- and always indicate the two signals crossing.  Maybe I'm misreading...  Actually, it may not matter and the 1.2V offset is just the particulars of the output stage, but the two signals need to cross or you'll need a reference ground.

 

[DACLadder]

 
Quote:

  Are you sure HDMI is DC around a 1.2V offset?  I always thought the pairs were AC with opposite polarity.  Depends on whether the CML rails are +/- or + and GND, but the whole idea of CML is to avoid switching noise by running a fixed current (3.5mA, which over 100ohm at 340MHz translates to 350mV).  Also, looking at eye graphs they're always +/- and always indicate the two signals crossing.  Maybe I'm misreading...  Actually, it may not matter and the 1.2V offset is just the particulars of the output stage, but the two signals need to cross or you'll need a reference ground.

Yes, HDMI LVDS has +1.2V offset nominally and not AC coupled.  The reason is a single positive voltage power supply can be used.  LVDS is also a constant current 3.5mA scheme but requires no pullup resistors with the benefit of using less power.   See Fig 1.1 in the T1 document linked above.  The "eye pattern" charts do not reflect the DC offset in any case.  The "crossing" you are referring can be set any place DC wise within range of the chips involved.   The receiver chip only looks at current flowing in one direction through the termination for a logic 1 (or +350mV across the resistor) and in the opposite direction for a logic 0 (-350mV).  Here's a chart that illustrates common differential logic with their respective offsets and swings.

 

[jacal01]

Nice guitar. 

 

[gsr108]

The pink faun should be coming in a week or so. I'm running long lines (15 feet), so not optimized. I can try to compare it to the coax out of my HT claro II. Can also try to compare it vs USB but the USB drivers have been crashing my desktop. I probably don't have the best ears or source material (Google streaming mostly) so my review may be lacking. I would possibly be willing to send it to a respected member with a OR5 for comparison.

So the pink faun arrived yesterday.  Setup was simple - plug it into a PCI-E slot, attach molex connector.  Uses windows drivers.  It doesn't seem to like the "redmere" HDMI cable I had from monoprice.  Replaced it with a another generic HDMI cable I had laying around and everything works now.  I was able to move my system around so am only running a 6 ft HDMI cable vs the 15 ft cable I was running earlier.  Quick impressions of the pink faun vs coax (HT claro II) and USB is imaging seems improved.  Initially I thought voices were more upfront but I think that is due to them being more centered making placement easier.  The soundstage seems more 3D whereas USB/coax seems more compressed horizontally.  It seems like I can pick out instruments, finger's snapping, etc easier but I need to listen to more music to make sure.  I've never heard my OR5 so take my opinions with a grain of salt. 
 
One issue though - the left and right channels are switched.  Simple enough to switch the cables around but I'm also using the HT claro II for games - would be a hassle to constantly move cables around.  I'm emailing the seller to see if I can get that fixed. 
 
Will try to update with more impressions as I get them. 

 

[DACLadder]

 
One issue though - the left and right channels are switched.  Simple enough to switch the cables around but I'm also using the HT claro II for games - would be a hassle to constantly move cables around.  I'm emailing the seller to see if I can get that fixed.  

Thanks for the review.  You are having a similar experience as I have with using HDMI I2S.  The only solution I have to offer for reversed left and right channels is to swap the HDMI L/RCLK connections on one side of the HDMI cable.  This will invert the L/RCLK signal.  Would require a cable where you can remove the HDMI connector shells and then resolder the wires.  Strange Pin Faun would design it this way but understandable with no standard for HDMI I2S.  But Pink Faun claims that the PS Audio HDMI I2S method is incorrect in the PMs you sent to me.
 

It doesn't seem to like the "redmere" HDMI cable I had from monoprice.  Replaced it with a another generic HDMI cable I had laying around and everything works now.

Did it sound bad or just not work at all?  I use generic 6 ft. cables on my A/V setup and haven't had a problem.  Searching AVS forum seems Monoprice  had an issue with Redmere cables in 2012.  They claim the issue has been since corrected.

 

[Articnoise]

Is the pin out on the HDMI input kit fitting the Offramp 5 without changing the pin out and on the homepage it says that on Master 7 no soldering is needed, is this correct?

 

[DACLadder]

 
Is the pin out on the HDMI input kit fitting the Offramp 5 without changing the pin out and on the homepage it says that on Master 7 no soldering is needed, is this correct?

You do not need to swap pinout when used with Off Ramp 5, PS Audio, Wyred4Sound, CI Audio, DI-2014 HDMI mod, etc.  Just use a standard, off-the-shelf HDMI cable.  Only Pink Faun seems to be rogue with inversion of the L/RCLK from the PS Audio method. 

 

[gsr108]

 
Thanks for the review.  You are having a similar experience as I have with using HDMI I2S.  The only solution I have to offer for reversed left and right channels is to swap the HDMI L/RCLK connections on one side of the HDMI cable.  This will invert the L/RCLK signal.  Would require a cable where you can remove the HDMI connector shells and then resolder the wires.  Strange Pin Faun would design it this way but understandable with no standard for HDMI I2S.  But Pink Faun claims that the PS Audio HDMI I2S method is incorrect in the PMs you sent to me.
 
Did it sound bad or just not work at all?  I use generic 6 ft. cables on my A/V setup and haven't had a problem.  Searching AVS forum seems Monoprice  had an issue with Redmere cables in 2012.  They claim the issue has been since corrected.

Using the redmere cable I couldn't get any sound to play; which is strange as it worked before with my video card passing PCM to a NAD M51.  I contacted pink faun, they are looking into why the channels are switched.  But it sounds like the card can be reprogrammed to have the correct output.