Review and Measurements of Topping D10 DAC

[Krunok]

We must to have almost zero output impedance until at 100Mhz.

"We must have" is not an answer, you should explain why.

Why would you need 100MHz when DAC chip doesn't output frequencies even remotely close to that? Not to mention that in the next stage after I/V you'll do low pass filtering to block alias frequencies..

And why "almost zero output impedance" when you will have output buffer to lower impedance - which btw is again OPA2134 in D10? Speaking of that, are you familiar that Amir did try rolling the op amps in place of output buffer in D10 and OPA1612 didn't perform any better than OPA2134?

 

[Krunok]

@777 give it up, @Krunok just wants to annoy you..

Not at all - I just want to see your photos and measurements and hear his explanations. Assuming two of you can provide those, of course..

 

[guenthi_r]

If there's no measureable difference, what do you suppose you're hearing / determines the better bass control?
op-amps could marginally improve (or break) amp implementation, but rolling op-amps in a DAC? Eh..

I only want to play with the D10, thats the reason i buyed the Topping.
It has no deeper meaning, dont worrying

 

[Veri]

I only want to play with the D10, thats the reason i buyed the Topping.
It has no deeper meaning, dont worrying

fair enough

edit: Krunok.. he can indeed come across as arrogant and almost aggressive, was hard to follow follow past conversation since I had him on ignore

 

[777]

"We must have" is not an answer, you should explain why.

Why would you need 100MHz when DAC chip doesn't output frequencies even remotely close to that? Not to mention that in the next stage after I/V you'll do low pass filtering to block alias frequencies..

And why "almost zero output impedance" when you will have output buffer to lower impedance - which btw is again OPA2134 in D10? Speaking of that, are you familiar that Amir did try rolling the op amps in place of output buffer in D10 and OPA1612 didn't perform any better than OPA2134?

Man, I do not have to explain nothing. You must read, a lot.

 

[Krunok]

fair enough

edit: Krunok.. he can indeed come across as arrogant and almost aggressive, was hard to follow follow past conversation since I had him on ignore

Oh, you again.. I still remember your expertise with interconnects when a poor guy bought totally wrong cable based on your advice not so long ago.

Btw, I see you kept your old style - brave as hell with giving personal accusations while hiding behind your keyboard. Were you reaised like that or you developed that behaviour on your own?

 

[Veri]

Oh, you again.. I still remember your expertise with interconnects when a poor guy bought totally wrong cable based on your advice not so long ago.

You mean that time where the seller of the cable fixed the order and there was no problem of said miscommunication, whatsoever?

Btw, I see you kept your old style - brave as hell with giving personal accusations while hiding behind your keyboard. Were you reaised like that or you developed that behaviour on your own?

??? What do you think this is, a fight club? Try and get a hold of yourself dude, you're embarrassment to yourself and to the forum with that attitude of yours.

 

[zenon]

If this is a real D10 schematic - the OPA1612AID opamps would be more suitable for both I / V and differential summing positions.
You will probably get lower distortion and will adversely affect dac sound quality.
Using OPA1612 that are more tolerant of RF is a good start. And they can supply a little bit higher output current which is necessary for use with ESS9018K2M.
With OPA2134 - D10 is designed to measure well first, and do so at a low price point. That means sound quality is not a consideration for anything, only the numbers that will be measured matter. What you get is what the numbers alone sounds like.

Yes, this is a real D10 I/V schematic, and no, OPA 1612 is not the best choice for ES9018K2M I/V stage (look at THD when R input is high value).
For best results:
Swap SPSU (XMOS, OPamps) to analog (1V; 8V), make separate PSU for XMOS PLL according data sheet, make separate AVCC_LR PSU. AVCC_LR PSU is critical for sound quality. Swap to fast (+low settling time), low noise, low THD OP amp.
Basically, every power supply needs to be of good quality!
I am currently exploring different options for AVCC, and if/when I find the optimal one I will write here.

 

[guenthi_r]

Yes, this is a real D10 I/V schematic, and no, OPA 1612 is not the best choice for ES9018K2M I/V stage (look at THD when R input is high value).
For best results:
Swap SPSU (XMOS, OPamps) to analog (1V; 8V), make separate PSU for XMOS PLL according data sheet, make separate AVCC_LR PSU. AVCC_LR PSU is critical for sound quality. Swap to fast (+low settling time), low noise, low THD OP amp.
Basically, every power supply needs to be of good quality!
I am currently exploring different options for AVCC, and if/when I find the optimal one I will write here.

Can you show me the diagram where the OPA1642 have high THD at R input please?
I only find a diagram for the output...

 

[JohnYang1997]

Isn't the old reference picture of selecting opamp for iv stage is for over sampling R-2R dacs? What does the output of direct output of 9018 look like? Maybe then we can analyze and make choices accordingly. And remember real world measurement results are always more powerful than calculations and circuit analysis. We often overlook stuff or overthink stuff. Where real world measurements don't lie(they do sometimes but still).

 

[JohnYang1997]

Can you show me the diagram where the OPA1642 have high THD at R input please?
I only find a diagram for the output...

I tried on smsl m3 a few months ago but not on d10. I tried opa2189 lmh6643(stock), opa2156, opa1656, opa1612 opa1642 opa2140 etc. opa1642 is actually really good. I think lmh6643 and opa1656/opa2156 have the lowest distortion at higher frequencies, 6643 probably even lower. But 6643 has too much noise in the low frequency also has higher low frequency distortion. The thing can be really convoluted. When you talk about thd at 1khz it's a easy task. Then you add in, noise, low frequency distortion at 20hz 60hz, high frequency distortion 6000hz 10000hz, and CCIF IMD, SMPTE IMD.... Sometimes you think you optimized it but then realized that you lost a lot more in other things. Designing a good dac is easy but trying to do the extreme for the last mile or inch it is hard.

 

[zenon]

Can you show me the diagram where the OPA1642 have high THD at R input please?
I only find a diagram for the output...

My mistake. R source.
Page 8 - 1612 datasheet.

 

[JohnYang1997]

My mistake. R source.
Page 8 - 1612 datasheet.

Opamp distortion performance with 10kohm r source.
Source - Johnc from Ti

 

[guenthi_r]

Opamp distortion performance with 10kohm r source.
Source - Johnc from TiView attachment 31657View attachment 31659

This is why i selecting OPA1642.

 

[zenon]

Rout 9018K2M is 806 ohm +/- 11% according data sheet.

 

[guenthi_r]

Rout 9018K2M is 806 ohm +/- 11% according data sheet.

781 Ohm from Es9018k2m + 820 Ohm from D10 Board.
Is my calculation correct?

 

[zenon]

781 Ohm from Es9018k2m + 820 Ohm from D10 Board.
Is my calculation correct?

From DS 9018K2M
"*Differnetial (+ or -) current output is equivalent to a differential (+ or -) voltage source in series with an 806ohm +/- 11% resistor...
...."

 

[guenthi_r]

From DS 9018K2M
"*Differnetial (+ or -) current output is equivalent to a differential (+ or -) voltage source in series with an 806ohm +/- 11% resistor...
...."

So we have 1.6kOhm Rs (Source impedance)?

 

[JohnYang1997]

So we have 1.6kOhm Rs (Source impedance)?

I think not. Just 781.5.

 

[guenthi_r]

I think not. Just 781.5.

So we dont adding the 820 Ohm resistor from the D10 board?
If this is correct, we really need an OPAmp with bipolar input....