Simple DAC mods that will make your DAC sound like it cost twice as much.

[Blizzard]

Hi Guys,

I thought I'd start this thread because I've modded quite a few DAC's so far, and the differences that can be made with very little money can be shocking!

The 2 easiest mods you can do are:

1: Master clocks

2: Discrete Opamp's


Depending on the DAC, even more mods are possible. If you have a Lampizator for example, several easy mods can be preformed that can result in huge performance upgrades for cheap.

If your among the crowd open minded enough to embark on this fruitful mission, let me know what you have, and I'll let you know what I think can be done cheaply and easily

 

[NorthStar]

Right on! Talk about something positively beneficial to the entire audiophile community...just like @ DIY.

* I have a Rotel RCD-991 CD player with HDCD decoder and variable dither levels (user selectable). The DACs are the Burr-Brown PCM-63 in differential mode, one per channel.

 

[Blizzard]

Right on! Talk about something positively beneficial to the entire audiophile community...just like @ DIY.

* I have a Rotel RCD-991 CD player with HDCD decoder and variable dither levels (user selectable). The DACs are the Burr-Brown PCM-63 in differential mode, one per channel.

I'll see if I can find pics of the board. A CD player like that would have lots of potential

 

[Blizzard]

Wow look at this. The Great Lukasz himself even modded this unit. I'll see what I can come up with.

http://www.lampizator.eu/LAMPIZATOR/REFERENCES/Rotel RCD-991/Rotel RCD-991 CD with tubes.html

 

[NorthStar]

Right now: http://www.erji.net/simple/index.php?t1231873.html

...And I was just going to post the link that you just posted above, with all the gold inside.

 

[Blizzard]

Right now: http://www.erji.net/simple/index.php?t1231873.html

...And I was just going to post the link that you just posted above, with all the gold inside.

There's soooo much info out there on modding this CD player. Looks like everyone's tried everything. But the easiest and probably most effective mod I can think of to start would be discrete opamps. Looks like 2 have done this on the Burson site:

http://www.bursonaudio.com/rotel-rcd-991-upgrade-with-burson-hd-opamp-by-garry-norman/

http://www.bursonaudio.com/burson-hd-audio-opamp-in-rotel-rcd991-cd-player/

But having tried many discrete opamps, if your budget allows, I'd reccommend the Sonic Imagery 994's. Hands down the best discrete dual opamp I know of:


http://www.sonicimagerylabs.com/products/Model994DiscreteOpAmp-Ticha.html

That should keep you entertained for a while. Then maybe upgrade the clock

 

[NorthStar]

By the way, my Rotel 991 is a very inexpensive CD player; only $400-500 on the used market today.

You think a good mod would be less than a thou? ...Or am I better to buy a brand new DAC for about that price and that can beat a mod Rotel 991?
/// Like a AKM DAC, the Verita AK4490EQ DAC inside, two or even four of them.

About the ModWright Oppo BDP-105 with tubes (hi grade) and its own separate power supply, the PS 9.0?
? http://www.modwright.com/modifications/oppo-bdp83-and-bdp83se-mod.php
It's about 3-4 grands.

Good night.

 

[Blizzard]

By the way, my Rotel 991 is a very inexpensive CD player; only $400-500 on the used market today.

You think a good mod would be less than a thou? ...Or am I better to buy a brand new DAC for about that price and that can beat a mod Rotel 991?
/// Like a AKM DAC, the Verita AK4490EQ DAC inside, two or even four of them.

About the ModWright Oppo BDP-105 with tubes (hi grade) and its own separate power supply, the PS 9.0?
? http://www.modwright.com/modifications/oppo-bdp83-and-bdp83se-mod.php
It's about 3-4 grands.

Good night.

The Sonic Imagery 994's are only $94USD a piece. Grab 2 of them for now. This will make a huge difference. I wouldn't go overboard though. Once you get into spending too much, you might be better off going with a DAC with like you say something like with an AKM 4490 chip, and 24/384 PCM/ DSD 256 compatibility.

You'll want a hot air rework station to swap the IC opamps with sockets though. If you don't have one I'd just find a local electronics repair shop that will do it. Only like a 5 minute job.

 

[amirm]

http://www.sonicimagerylabs.com/products/Model994DiscreteOpAmp-Ticha.html

Boy, that is a pretty neat design!

 

[Blizzard]

Boy, that is a pretty neat design!

It's a killer discrete opamp. It can sink or source 150mA with THD+N of 0.0003%, and -140dB noise floor. And it runs in pure class A up to 30mA! I've tried Burson's, Dexa's, Sparko's and these. Nothing can touch these units. If your DAC uses IC opamps, swapping them out to these would be a dramatic transformation.

They aren't afraid to backup their claims with a comprehensive datasheet as well. Who else has a datasheet like theirs?

This is straight up solid engineering. No snake oil here.

 

[amirm]

The engineer nerd in me just wants to get one to play with . My own DAC uses a discrete output stage so no real application yet.

 

[DaveC]

Those would probably make a very nice preamp/buffer stage on their own...

Any recs for a Sony HAP-Z1ES? RWA and Modwright offered some upgrades like tube output stages but imo that would just be a difference. The stock SS output section on the HAP is pretty darn good with a far lower output impedance...

 

[Blizzard]

Those would probably make a very nice preamp/buffer stage on their own...

Any recs for a Sony HAP-Z1ES? RWA and Modwright offered some upgrades like tube output stages but imo that would just be a difference. The stock SS output section on the HAP is pretty darn good with a far lower output impedance...

Yes a damn good preamp at that.

Clocks and Opamps for the Sony as well. Just need to figure out what it's using now for opamps. The clocks look like 22.5 and 24.5mhz. These are the best pics I could find.

If you have one, pop the cover off and take high Rez pics of the clocks and opamps.


View attachment 23653View attachment 23654View attachment 23655

 

[microstrip]

It's a killer discrete opamp. It can sink or source 150mA with THD+N of 0.0003%, and -140dB noise floor. And it runs in pure class A up to 30mA! I've tried Burson's, Dexa's, Sparko's and these. Nothing can touch these units. If your DAC uses IC opamps, swapping them out to these would be a dramatic transformation.

They aren't afraid to backup their claims with a comprehensive datasheet as well. Who else has a datasheet like theirs?

This is straight up solid engineering. No snake oil here.

Are you sure that are not forgetting zeros in the THD spec? The Texas Instruments LM4562 audio operational amplifier has a THD + N @ 1 kHz of 0.00003%.

 

[NorthStar]

Can you hear that additional zero in a blind test.

 

[Blizzard]

Are you sure that are not forgetting zeros in the THD spec? The Texas Instruments LM4562 audio operational amplifier has a THD + N @ 1 kHz of 0.00003%.

Yes I know but is sure doesn't sound as good.

How do I know?

View attachment 23656

The problem with IC opamps are, as soon as they are subjected to real world conditions, that stellar performance spec shits the bed. The 0.0003 THD+N is the 994's rating while sinking and sourcing 150mA. The LM4562 craps out at 60mA.

Try sinking 30mA into a LM4562 biased in class A and let us know how it sounds.

 

[Ric Schultz]

Blizzard,
I think you should stick to what you know. Just tell us that the discrete one sounds better. Please do not try to justify it with specs. Seems you don't understand them: The 4562 is biased at 5 milliamps per amplifier....the 994 at 17 ma. Neither will be asked to sink more than a 1ma when used on the output of a DAC. 2 volts RMS divided by the load...say 20K is .0001 amps....or .1 milliamp. How do you get 30ma sinking?....that is 300 times the actual current being delivered into a load. And how do you bias a 4562 into class A? You can put a constant current source to ground on the output and force one half of its output into deeper class A....but this is not a very successful technique. Its nice that the 994 can deliver more current.....but most applications do not require much current drive at all. Now, if you are driving a 600 ohm load....which is what professionals like to think is the case, then current delivery is more crucial. However, you only need one amplifier on the output of the 4490 (a summer) and it would drive your amp (100K in your case) or a preamp (10-100K typically) directly. Standard output of a DAC is 2VRMS. Even if you set if for 4VRMS it would only double the current stated above. If you set the op amp to make 10VRMS peaks and you drive a 600 ohm load then you are sinking 16 milliamp peaks (not a normal home set up at all). What is nice about having more current available is that you could drive low impedance headphones directly.....this is where you would need some serious current. A normal DAC output of 2VRMS into a 10K or higher load is just not current demanding at all.

Specs are one thing and sound is another. Please do not try to prove to us that something that measures better in some ways makes for better sound. It just is not always the case. I have no doubt that you listened to both integrated circuit op amps and discrete op amps and that you find this particular discrete one the best sounding. I like discrete ones better myself.....but I would never justify it with a spec, cause there is no way to really know. If you were going to use a spec for comparison then the fact that the discrete one is biased much higher in class A (in probably all its stages....not just output stage) might make a sonic difference. But there is so much more to everything than any one spec.

 

[NorthStar]

He's not saying that better specs sound better, he replied to microstrip's (Francisco) post who mentioned one OA better (spec) with one more zero in its THD+N.

Nature, music, audio, spirit. Love is all there is!

 

[Ric Schultz]

Very few people have the skill or inclination to mod something to any serious extent. I think you are wasting your time here. I personally have mostly given up on trying to "help" people because what us tweaks talk about seems to be like talking Martian to most people....they just cannot hear it because they have no direct experience with it. There is hardly anyone listening. Changing clocks, output op amps, caps and resistors and jacks, etc. has been around forever but hardly anyone wants to do it but a few would like someone else to do it for them.....so start modding, like I do. You could mod peoples stuff with better clocks and discrete output stages and what not...that is what I do. However, if you think you will motivate anyone to make a DAC for $400 in parts that might compete with a $10,000 DAC....well.....you are just fooling yourself. Just build a great $2000 DAC all tweaked out and sell it. That is what I am planning to do. If it still beats the $10,000 DAC then everyone is happy. Hardly any one is really listening to what you say about self tweaking.......I know, I have tried for years to help people this way......you are just talking to yourself and maybe a couple of other people. No one cares about your 4490 DAC with discrete output and great clocks....no one. Why should they? They cannot get one. However, If you make one and sell it then you might have an audience. An audience is something you obviously seek.

I have modded a Sony Happy one with discrete output stages, better parts, super many tweaks and it sounded way, way better than stock. However, it was still not that great. The problem with the Sony is that it has no direct digital in and no digital out.....so you are stuck with those Burr Brown DACs and their form of upsampling, etc. What if you want to use an R2R DAC or different form of upsampling.....you cannot. Just not an open system. DACs are coming out every day....soon I will probably have one (parallel R2R with discrete outputs, shunt regs, WA quantum chips, Magic Tubes, Less Loss filter, worlds best parts, etc.).

Yes, he is saying that specs matter and is quoting them.....

And indeed LOVE IS ALL THERE IS.....Happy Thankgiving....with each breath....Joy and Love to everyone.

 

[Blizzard]

Blizzard,
I think you should stick to what you know. Just tell us that the discrete one sounds better. Please do not try to justify it with specs. Seems you don't understand them: The 4562 is biased at 5 milliamps per amplifier....the 994 at 17 ma. Neither will be asked to sink more than a 1ma when used on the output of a DAC. 2 volts RMS divided by the load...say 20K is .0001 amps....or .1 milliamp. How do you get 30ma sinking?....that is 300 times the actual current being delivered into a load. And how do you bias a 4562 into class A? You can put a constant current source to ground on the output and force one half of its output into deeper class A....but this is not a very successful technique. Its nice that the 994 can deliver more current.....but most applications do not require much current drive at all. Now, if you are driving a 600 ohm load....which is what professionals like to think is the case, then current delivery is more crucial. However, you only need one amplifier on the output of the 4490 (a summer) and it would drive your amp (100K in your case) or a preamp (10-100K typically) directly. Standard output of a DAC is 2VRMS. Even if you set if for 4VRMS it would only double the current stated above. If you set the op amp to make 10VRMS peaks and you drive a 600 ohm load then you are sinking 16 milliamp peaks (not a normal home set up at all). What is nice about having more current available is that you could drive low impedance headphones directly.....this is where you would need some serious current. A normal DAC output of 2VRMS into a 10K or higher load is just not current demanding at all.

Specs are one thing and sound is another. Please do not try to prove to us that something that measures better in some ways makes for better sound. It just is not always the case. I have no doubt that you listened to both integrated circuit op amps and discrete op amps and that you find this particular discrete one the best sounding. I like discrete ones better myself.....but I would never justify it with a spec, cause there is no way to really know. If you were going to use a spec for comparison then the fact that the discrete one is biased much higher in class A (in probably all its stages....not just output stage) might make a sonic difference. But there is so much more to everything than any one spec.

You should check out this thread. Start with post #84:

http://www.head-fi.org/t/745032/lks-audio-mh-da003/75

In my particular case, my amps have a 1.8k input impedance. So they require a lot of grunt to drive properly. 13.46v will drive them to the full rated output.


Another good article:

http://stephan.win31.de/music.htm#opclassa