The Zishan dsd's corner
Oct 6, 2019 at 1:32 PM Post #2,491 of 3,711
Google if ad8397 is unity gain stable (gain=1), I've read it can oscillate, if so it's not a good candidate for buffer, but I never tried it.
From what I read in theory according to datasheet it is stable, in practice it might oscillate. Have you other suggestion for other >2 channels opamp with low distorsion and current output >100mA ?
 
Oct 6, 2019 at 3:06 PM Post #2,492 of 3,711
Ivan recommend me these 2:
https://www.head-fi.org/threads/the-zishan-dsds-corner.826185/page-140#post-14824903

How to configure them as buffers:
https://www.head-fi.org/threads/the-zishan-dsds-corner.826185/page-123#post-14789479
"bridge 1-2 and 6-7"

I had opa1688 in the socket and had removed the discrete transistors buffer.
Tried with opa1622 configured as buffer instead of the transistors and I didn't noticed better or worst sound (compared to not using buffer), only different sound signature. Maybe my 300 and 400 ohm earbuds don't need more current than opa1688 in the socket provided. I'll take another look at my buffer options, but I don't have much time right now to make tests and look into that.

Maybe you could consider both options in BO to see if you like any of them.
Take a look at datasheets as I don't remember their current output.
A simple test board could be built with sockets to test both opamps as buffers outside DSD.
When inside DSD, opa1622 pcb resistors can be removed and it's ground can be connected to DSD ground. Also increasing it's local power reserve caps in it's pcb can improve bass.
I didn't search for other options.

I also tried AD45048 (in other daps) but not as buffer (another name for AD8397) and it's temperature was hot (sound was normal, but I preferred others). I'll try it again when I have some time.
 
Last edited:
Oct 7, 2019 at 11:58 AM Post #2,494 of 3,711
Have you tried other opamp ?

many.
opa2134, 2228, 1612, 1688, 1692, 2209, 189, 1692.. ad823, 8620, muses8920, lme49990, 4562, lm6172, a few intersil and linear opamps too..

edit: please note that, I tried those opamps as lpf, in zishan4495 version. newer 4497 has a quite different schematic from former..
 
Last edited:
Oct 11, 2019 at 3:04 AM Post #2,495 of 3,711
Ivan recommend me these 2:
https://www.head-fi.org/threads/the-zishan-dsds-corner.826185/page-140#post-14824903

How to configure them as buffers:
https://www.head-fi.org/threads/the-zishan-dsds-corner.826185/page-123#post-14789479
"bridge 1-2 and 6-7"

I had opa1688 in the socket and had removed the discrete transistors buffer.
Tried with opa1622 configured as buffer instead of the transistors and I didn't noticed better or worst sound (compared to not using buffer), only different sound signature. Maybe my 300 and 400 ohm earbuds don't need more current than opa1688 in the socket provided. I'll take another look at my buffer options, but I don't have much time right now to make tests and look into that.

Maybe you could consider both options in BO to see if you like any of them.
Take a look at datasheets as I don't remember their current output.
A simple test board could be built with sockets to test both opamps as buffers outside DSD.
When inside DSD, opa1622 pcb resistors can be removed and it's ground can be connected to DSD ground. Also increasing it's local power reserve caps in it's pcb can improve bass.
I didn't search for other options.

I also tried AD45048 (in other daps) but not as buffer (another name for AD8397) and it's temperature was hot (sound was normal, but I preferred others). I'll try it again when I have some time.

Thanks for all the suggestions, bought 2x 1622 in dip sockets and 2x tpa6120 (just to test them, they are cheap and I don't have multi BA iem) and an oscilloscope to better verify the results. Now I have to wait the shipping from China (usually 1 month)
 
Last edited:
Oct 16, 2019 at 4:16 PM Post #2,497 of 3,711
Question: with all the mods discussed here and in various related threads (which I have read and taken notes on, great work by all the contributors), is there any value in aiming to replace all the main passive components (resistors, capacitors etc) with 'non-fake' versions from places like Mouser, Arrow etc?

I have heard many say that when they replace a delivered ChiFi opamp with the 'same' opamp, but a known genuine version, it sounds better.

Does that apply to many of the parts? Of course I realize the cost would very quickly rise above the cost of the player itself, so its possibly not rational, but just having fun with them.

I have Z1,Z2,Z3, T1, DSD, and Walnut V2 and V2s - love listening to them, want to 'play/tinker'...:)
 
Oct 17, 2019 at 4:12 AM Post #2,498 of 3,711
is there any value in aiming to replace all the main passive components (resistors, capacitors etc) with 'non-fake' versions from places like Mouser, Arrow etc?
Depends on what ear/headphones you are using and who you are.

As you should know I pretty much replaced most of components in at least one of my DSD's.

It was a good experiment and a great experience and I can confirm that utilising thin film 0.1% 0805 resistors, PPS SMD caps, sandwiched 0.1/10uF x7r bypass caps, lo-noise LDO's and better inductors and even quality transistors improves things a lot... Funny enough my heavily modded DSD is great, but I prefer very basically modded one for certain 3-4 way BAs.
But replacing components for better quality (I don't really subscribe to this notion of "fake") components can be a can of worms you don't want to open unless you are ready to deal with the consequences ($ and time and a really high chance of destroying your DSD or two in the process).

But this "improves things a lot" should be properly contextualised as a manifestations of a law of diminishing returns: stock DSD is like 80-85, maybe even 90% there sound-wise, but to squeese those extra 10% becomes an expensive and time consuming affair...

Is it worth it? If you are me (fanatical sound purist, DIYing my own IEMs and having acute auditory perception and not willing to enjoy music unless it sounds JUST RIGHT) - definitely YES.

If your game is having great sound on a budget...

Change LDOs, OpAmp and maybe output caps and budget audio nirvana has been attained. The key word is budget, and it is an audio nirvana... But a budget version of it.

PS: let me know if you need assistance with picking quality components should you choose take a plunge I'm sure @Merlin-PT will be happy to help too, provided he is back from his travels.
 
Last edited:
Oct 17, 2019 at 8:07 AM Post #2,499 of 3,711
Thanks! I have followed your extensive mods with great interest, not that I could do all of them :) I appreciate your thoughts.
 
Oct 17, 2019 at 8:15 AM Post #2,500 of 3,711
Depends on what ear/headphones you are using and who you are.

As you should know I pretty much replaced most of components in at least one of my DSD's.

It was a good experiment and a great experience and I can confirm that utilising thin film 0.1% 0805 resistors, PPS SMD caps, sandwiched 0.1/10uF x7r bypass caps, lo-noise LDO's and better inductors and even quality transistors improves things a lot... Funny enough my heavily modded DSD is great, but I prefer very basically modded one for certain 3-4 way BAs.
But replacing components for better quality (I don't really subscribe to this notion of "fake") components can be a can of worms you don't want to open unless you are ready to deal with the consequences ($ and time and a really high chance of destroying your DSD or two in the process).

But this "improves things a lot" should be properly contextualised as a manifestations of a law of diminishing returns: stock DSD is like 80-85, maybe even 90% there sound-wise, but to squeese those extra 10% becomes an expensive and time consuming affair...

Is it worth it? If you are me (fanatical sound purist, DIYing my own IEMs and having acute auditory perception and not willing to enjoy music unless it sounds JUST RIGHT) - definitely YES.

If your game is having great sound on a budget...

Change LDOs, OpAmp and maybe output caps and budget audio nirvana has been attained. The key word is budget, and it is an audio nirvana... But a budget version of it.

PS: let me know if you need assistance with picking quality components should you choose take a plunge I'm sure @Merlin-PT will be happy to help too, provided he is back from his travels.
I changed LDOs, opamp and output caps and that was a pretty good upgrade, on my other dsd I simply changed the opamp for a Burson v5i and that is a good improvement too but they are different versions of the dsd, I might do more but with the cost going up quickly, will probably just wait for a better diy player, in my opinion, swapping everything for better parts would be a bit overkill and wouldn't make much sense budget wise, maybe easier and overall better to buy a better player from the start
Unless you just want to learn, experiment and practice, then the dsd is a great base
 
Oct 17, 2019 at 11:52 AM Post #2,501 of 3,711
Great input thanks, I will focus first on the main caps, opamps and ldo's...if my dsd survives this I may try one or two smd caps just for 'fun' :)
 
Oct 17, 2019 at 3:33 PM Post #2,502 of 3,711
If you change capacitors, my suggestion is to cut it’s legs off and use those as your soldering points. Chances are, you may end up changing caps more than you think you will. The biggest danger when making multiple passes on a soldering point on an electronic board is lifting the leads (which is more trouble to fix than you want). The leads on these devices are easy to lift if you are not very diligent with solder cleanup and patience. Even then, multiple passes will lead to a lifting lead.
 
Oct 19, 2019 at 12:24 AM Post #2,503 of 3,711
HTB1CWiEqTtYBeNjy1Xdq6xXyVXa7.jpg


Is the balanced output of Zishan 4497 has lower Z-out than the 22Ohm SE?
 

Users who are viewing this thread

Back
Top