DX90. 2X Sabre,1st page: Downloads, info&inst. . ! Lurker0 FW Mod link 1st page !!. .NEW FW! 2.3.0 . . . . .
Apr 9, 2014 at 10:35 AM Post #2,371 of 14,084
   
Afraid you are quite wrong, both the diagrams actually show exactly how many opamps the x5 and dx90 use. The x5 uses a single dual channel opamp for LPF and another for AMP, while dx90 uses two single channel opamps for each, the dx90 is technically the better design, using two extra opamps over the x5 to completely separate the channels.
 
The reason why the x5 and dx90 have two dual channel opamps after the dac is because both dacs output a 2 channel current for each left and right channel (so 4 channel total), so either 2 dual channel opamps are needed, or 4 single channel opamps, after this the signal is converted to 2 channel and you need either 1 dual channel opamp or 2 single channel op amps the rest of the way down the chain.

No way to know that for sure so I didn't get too involved beyong a cursory check. The last 1612 of the X5 can't feed 2 channels from one output so it's impossible to know the exact chip count because the diagram doesn't actually make sense. The biggest difference is the balanced use of the DAC ind DAC itself of the DX50 along with the summing or differential input to single out in the LPF stage. If you think a balanced Sabre is better than a single 1792 it's better if not, not. Last time you said the 1612 is better than the 1602. Is one 1612 better than two 1602? The answer will always be, it depends.
 
There are the same number of stages doing the same functions. How it relates to performance is speculation. There is no better or worse in those 2 paths. It will come down to execution. 
 
Apr 9, 2014 at 10:45 AM Post #2,372 of 14,084
  No way to know that for sure so I didn't get too involved beyong a cursory check. The last 1612 of the X5 can't feed 2 channels from one output so it's impossible to know the exact chip count because the diagram doesn't actually make sense. The biggest difference is the balanced use of the DAC ind DAC itself of the DX50 along with the summing or differential input to single out in the LPF stage. If you think a balanced Sabre is better than a single 1792 it's better if not, not. Last time you said the 1612 is better than the 1602. Is one 1612 better than two 1602? The answer will always be, it depends.
 
There are the same number of stages doing the same functions. How it relates to performance is speculation. There is no better or worse in those 2 paths. It will come down to execution. 

 
Yes we do know for sure. James from Fiio confirmed the x5 uses 7 opamps, the 6 you see in the diagram plus another one for the lineout, while the dx90 clearly uses 9 opamps. I'm not sure you understand the difference between a single channel opamp and a dual channel opamp in the diagram...  A single dual channel opamp can replace two single channel opamps, lessening cost but also some performance (if both of same design) Eg, The 1612 used by x5 is a dual channel opamp, while the 1611 in dx90 is the same chip in a single channel version (exact same sound sig but better specs). Similarly it can easily be said the 1612 is better than the 1602, because their manufacturer Texas instrument clearly states the 1602 is the lower cost and lower performance version of the 1612. They essentially sound the same, but technically one has extended frequency response, lower distortion, faster response time etc, all important factors when a chip is converting a dac signal. By the way the 1612 and 1602 are dual channel chips (can feed two channels), while the 1611 is a single channel chip (need two of them for left and right channels).
 
Apr 9, 2014 at 10:52 AM Post #2,374 of 14,084
I am surprised at times at the performance of dual op amps. They can be very good. Having said this, I also like single op amps for the reasons stated in this thread. They allow for even better separation and just a little more power supply performance, which often leads to better dynamics and a better crosstalk number. I designed the pseudo dual power supply that Woo Audio uses for this very reason. They were using one power supply for R and L channel. The PDPS increased performance for a minimal cost. Sometimes dual op amps are used for space saving also and simpler layout. It is all in the hearing/listening though. I look forward to impressions once the DX90 starts hitting the doorsteps. 
 
Apr 9, 2014 at 10:56 AM Post #2,375 of 14,084
Yeah the use of 1602 instead of the 1612 is kind of confusing for me too. I hope Ibasso has a good explanation for it. I'm sure it's not about cost. The DX90 would have sold just as well if it was $20 more expensive or whatever it takes to use the better chip. I hope Ibasso did some listening tests and preferred the signature from the 1602 or there are other technical reasons we simple-minded can't understand. I hope one day we get to hear the real reason.
 
Apr 9, 2014 at 11:00 AM Post #2,376 of 14,084
   
Yes we do know for sure. James from Fiio confirmed the x5 uses 7 opamps, the 6 you see in the diagram plus another one for the lineout, while the dx90 clearly uses 9 opamps. I'm not sure you understand the difference between a single channel opamp and a dual channel opamp in the diagram...  A single dual channel opamp can replace two single channel opamps, lessening cost but also some performance.

Good he let us know but it doesn't matter. Yours is a narrow view. Unless the're the exact same device running under the same conditions, there are no absolutes in this compare. Something as simple as a few pF near the the chip can make a huge difference in stability performance. The amount of feedback or gain clearly changes parameters beyond what we're discussing. This parts in a box view is ill conceived. I'm certainly hoping the DX90 is executed better as I ordered one.
smile.gif
 
 
If all is the same I tend to prefer separate chips as well but all isn't the same here. We can't say one is better in actual use just because we see a block diagram. Ask an engineer that's toiled over circuits and ask him how much he can tell of the actual performance of a circuit from a block diagram. Same number of stages is electrically the same with the only difference being the required summing in one of them from the balanced use of dual Sabres.
 
Apr 9, 2014 at 11:09 AM Post #2,377 of 14,084
  Good he let us know but it doesn't matter. Yours is a narrow view. Unless the're the exact same device running under the same conditions, there are no absolutes in this compare. Something as simple as a few pF near the the chip can make a huge difference in stability performance. The amount of feedback or gain clearly changes parameters beyond what we're discussing. This parts in a box view is ill conceived. I'm certainly hoping the DX90 is executed better as I ordered one.
smile.gif
 

 
Oh I completely know that, in fact a several pages back I mentioned that the implementation is what counts and for all we know the x5 could have better circuitry than the dx90. It is impossible to say which is better impemented by looking at a bunch of opamps. I was simply only looking at the opamp point of view as thats all I got at the moment, and opamp choice and layout is still a huge factor in the daps sound.
 
Apr 9, 2014 at 3:26 PM Post #2,380 of 14,084
Just ordered mine
biggrin.gif

wonder how long it will take to ship
popcorn.gif

 
Apr 9, 2014 at 3:27 PM Post #2,381 of 14,084
  I am surprised at times at the performance of dual op amps. They can be very good. Having said this, I also like single op amps for the reasons stated in this thread. They allow for even better separation and just a little more power supply performance, which often leads to better dynamics and a better crosstalk number. I designed the pseudo dual power supply that Woo Audio uses for this very reason. They were using one power supply for R and L channel. The PDPS increased performance for a minimal cost. Sometimes dual op amps are used for space saving also and simpler layout. It is all in the hearing/listening though. I look forward to impressions once the DX90 starts hitting the doorsteps. 

I look forward to the impressions too (and will probably post my own), but I have to admit that your impressions are what convinced me to pull the trigger on the DX90 in the first place.  I've read your reviews and impressions about other components on Head-Fi and take great stock in your opinions.  You've never let me down and I don't expect you will this time either.
biggrin.gif
  Definitely looking forward to the DX90!
 
Cheers and All the Best!
beerchug.gif

-HK sends
 

Users who are viewing this thread

Back
Top