Next Level of Performance DACs.

[JohnYang1997]

Are there any advantages to putting the DAC "inside" the amp and optimizing the amp's output? I am just thinking that the amp output is what the end-game is, and there is probably more progress that can be made on the amp output.

wut?

 

[HooStat]

Are there any advantages to improving the SINAD of an amp if you can control the DAC output 100%? (i.e., incorporate the DAC with the amp in one unit -- say USB or coax input and L/R amp output)

 

[JohnYang1997]

Are there any advantages to improving the SINAD of an amp if you can control the DAC output 100%? (i.e., incorporate the DAC with the amp in one unit -- say USB or coax input and L/R amp output)

Are you saying using relay controlled step volume after DAC output to reserve SNR? Yes, and it's been done. Varying in different steps, you get RME adi2dac, Matrix element x. Adding a low noise preamp serves same purpose like LA4 from benchmark.

 

[IVX]

JohnYang1997, yesterday I tried to find any offers for the leasing of APx555 in Guandong - nothing. Probably, I going to buy one, it is stupidly expensive but for the reason, I mentioned above, I have to..

 

[JohnYang1997]

JohnYang1997, yesterday I tried to find any offers for the leasing of APx555 in Guandong - nothing. Probably, I going to buy one, it is stupidly expensive but for the reason, I mentioned above, I have to..

Try making a 4* ak5578 ADC with attenuated input. It should give better performance than AP's input at 1khz.

 

[IVX]

JohnYang1997, no way, it will make a bit less noise but a lot worse THD. I have a simple analog 1kHz notch filter that outperforms any AP in all respects, however, as I mentioned above, I have to measure my products with trusted equipment. In fact, I gonna throw out $30000 only to make screenshots from APx555 application

 

[JohnYang1997]

JohnYang1997, no way, it will make a bit less noise but a lot worse THD. I have a simple analog 1kHz notch filter that outperforms any AP in all respects, however, as I mentioned above, I have to measure my products with trusted equipment. In fact, I gonna throw out $30000 only to make screenshots from APx555 application

You can't measure noise with self made notch filter right? THD should be better than -130dB but won't come close to -150dB that's for sure. But you should be able to get better overall THD+N.
Also I heard about FLEX from AP that give you the software itself that you can use with other electronics.

 

[IVX]

You can't measure noise with self made notch filter right?

who can stop me? )) THD+N of the OPA1612 way lover than -130db with low enough network impedance and good quality caps.

 

[JohnYang1997]

who can stop me? )) THD+N of the OPA1612 way lover than -130db with low enough network impedance and good quality caps.

You can give a result but I guess it needs to be calibrated or apply with external eq to make sure the noise part from the DUT isn't getting reduced. The noise measurements can be inaccurate with notch filter.

 

[IVX]

All AP analyzers do the same high-Q notch filtering, I see zero reasons to EQing noise-floor, at least for 1kHz and BW22k. The AP's problem is they using 10k resistor(hence, SNR -120db SYS2 but APx555 probably 3-5db lower, or 3-5db higher voltage) in their adjustable notch and only 2VRMS input.

 

[JohnYang1997]

All AP analyzers do the same high-Q notch filtering, I see zero reasons to EQing noise-floor, at least for 1kHz and BW22k. The AP's problem is they using 10k resistor(hence, SNR -120db SYS2 but APx555 probably 3-5db lower, or 3-5db higher voltage) in their adjustable notch and only 2VRMS input.

Alright.

 

[IVX]

or another way to froze AP down to -190C with liquid nitrogen, if it will start yet

 

[Harmonie]

JohnYang1997, no way, it will make a bit less noise but a lot worse THD. I have a simple analog 1kHz notch filter that outperforms any AP in all respects, however, as I mentioned above, I have to measure my products with trusted equipment. In fact, I gonna throw out $30000 only to make screenshots from APx555 application

Sorry maybe a candid question, but before throwing out 30 grands, can't you make such test in some audio lab who has the equipment you are looking for ?

 

[JohnYang1997]

or another way to froze AP down to -190C with liquid nitrogen, if it will start yet

Same condition as Amir. Basically what paralleling left and right channel of okto stereo can get to. AP loop back can only get to 0.000080%

Another one with 5.2V

 

[IVX]

JohnYang1997, not sure what you wanna say but I see 1.8db in levels difference and 1.8db in the THD+N reading too(-135db harmonics corresponds to 0.000017% of THD). As I told, all AP, including APx555, are too noisy to accurately measure modern DACs. Not sure if we really need that practically, especially if loaded THD+N tens times worse.
BTW, a good example https://www.audiosciencereview.com/...topping-a90-headphone-amplifier-review.13592/ No-load amp is nonsense by definition, of course, because any $1 opamp will show the same -140-150db level of 2-3rd harmonics. Let's see what happens if the amp is loaded at 33ohm, there is -115db i.e. 30 times worse. And such a miracle we may find almost everywhere, even if we pay a few grand for a DAC or amp. So, the achievement ultralow THD+N with the ideal no-load condition is a pure virtual sport to me without any practical content.

 

[JohnYang1997]

JohnYang1997, not sure what you wanna say but I see 1.8db in levels difference and 1.8db in the THD+N reading too(-135db harmonics corresponds to 0.000017% of THD). As I told, all AP, including APx555, are too noisy to accurately measure modern DACs. Not sure if we really need that practically, especially if loaded THD+N tens times worse.
BTW, a good example https://www.audiosciencereview.com/...topping-a90-headphone-amplifier-review.13592/ No-load amp is nonsense by definition, of course, because any $1 opamp will show the same -140-150db level of 2-3rd harmonics. Let's see what happens if the amp is loaded at 33ohm, there is -115db i.e. 30 times worse. And such a miracle we may find almost everywhere, even if we pay a few grand for a DAC or amp. So, the achievement ultralow THD+N with the ideal no-load condition is a pure virtual sport to me without any practical content.

We got -130db harmonics on 16Ohm load when balanced. -130db harmonics for 32ohm load when unbalanced. Also distortion performance don't degrade when using high gain setting. The things that aren't easily shown required most of the work. To compete with THX these have to be done but Magnius (and heresy) hacked the system by supplying good no load performance. And I also value 50mV SNR a lot which is not very good on THX based amps.(105dB achieved)

What I was trying to say in the last post was that there is still room for AP to measure differences in current dacs.(not accurately but differences) The last few bits actually was another 3dB increase in overall performance.
DAC+AP can have better performance than AP loopback. That's what I am trying to say.

 

[IVX]

JohnYang1997, so what is that? BTW, I also said to myself, no reason to design HPA and ask for that trivial circuit money with worse than -130db harmonics level

 

[JohnYang1997]

JohnYang1997, so what is that? BTW, I also said to myself, no reason to design HPA and ask for that trivial circuit money with worse than -130db harmonics level

Amir didn't turn on high performance mode. Hence it's limited by noise.
See here for our measurements (THD+N vs Level graphs are A-weighted, THD vs Frequency graphs are not weighed)
http://www.tpdz.net/productinfo/472128.html

 

[IVX]

http://www.tpdz.net/productinfo/472128.html now it is clear!