Author Topic: Ozmillsy's KillerDAC  (Read 87217 times)

Offline ozmillsy

  • Hero Member
  • *****
  • Posts: 2249
  • Liked: 277
Re: OzM's KD-Plessey
« Reply #100 on: July 26, 2014, 08:32:56 PM »
Hi Oz, do you like the Plessey better than the Dubilier?
Listened for about an hour, and sounds promising.  The Plessey is brand spanking new,  and will likely benefit from some running in.

But I am definitely edging closer to my reference (the r2r).
It's all about the music,, not the equipment.

Offline Jehuty

  • Hero Member
  • *****
  • Posts: 1403
  • Liked: 111
Re: Ozmillsy's KillerDAC
« Reply #101 on: July 27, 2014, 12:40:45 AM »
Good news Oz. Now, you only need a Studer chip I reckon  :D
Not all that matters can be measured, not all that can be measured matters.

Offline ozmillsy

  • Hero Member
  • *****
  • Posts: 2249
  • Liked: 277
Re: Ozmillsy's KillerDAC
« Reply #102 on: July 27, 2014, 08:17:56 AM »
I would like a Studer r2r. Thats what I really need.   Tape is where it's really at.  8)

A studer chip would be nice,  but not essential anymore, following recent changes.
It's all about the music,, not the equipment.

Offline zenelectro

  • Hero Member
  • *****
  • Posts: 825
  • Liked: 177
Re: Ozmillsy's KillerDAC
« Reply #103 on: July 28, 2014, 10:35:23 AM »
Nice collection of E88CC there Oz.

You might want to speak to Bryan as I remember he uses different output valves for his KillerDAC. Not sure what valves but who knows, you might end up liking the sound  :)

AFAIK Brian is using 6C45 Pi  Russian super tubes on his DAC.
These are somewhat like the venerable WE437A. They are pretty tricky to implement and love to oscillate.

They have higher OP and more drive than ECC88 etc but if you have Russian tube phobia then best avoid.

I have not compared them back to back against ECC88 as they are different pinout and run very different currents etc.

Z

Offline ozmillsy

  • Hero Member
  • *****
  • Posts: 2249
  • Liked: 277
Re: Ozmillsy's KillerDAC
« Reply #104 on: July 28, 2014, 12:32:41 PM »
Hey Z,  if i was to try them, do I need to change other stuff around the tube?

Things are starting to come together now, without going for that option.
It's all about the music,, not the equipment.

Offline zenelectro

  • Hero Member
  • *****
  • Posts: 825
  • Liked: 177
Re: Ozmillsy's KillerDAC
« Reply #105 on: July 29, 2014, 03:10:03 PM »
Hey Z,  if i was to try them, do I need to change other stuff around the tube?

Things are starting to come together now, without going for that option.

Yes mate, you need to re wire the socket, and run a bit more current through the tube.
I think they also run better choke loaded, (instead of a plate resistor) but haven't tried it.

There are quite a few mods that might get you some more top end clarity.

What I-V resistor are you using?

What regs are you running on your KD?

What bit bypass caps (14x) are you running?

Lot's of options.

cheers

Offline ozmillsy

  • Hero Member
  • *****
  • Posts: 2249
  • Liked: 277
Re: Ozmillsy's KillerDAC
« Reply #106 on: July 29, 2014, 06:00:46 PM »
Hey mate,

I've come a long way, with the mods I have documented in this thread thus far.

I am getting to the point where I run the risk of going too far.   I have a bit more reference comparisons to do (with the r2r), to be sure.
It's all about the music,, not the equipment.

Offline rab

  • Administrator
  • Full Member
  • *****
  • Posts: 232
  • Liked: 36
Re: Ozmillsy's KillerDAC
« Reply #107 on: August 06, 2014, 10:44:56 PM »
the DAC (in comparison to my other front end components) has been rolled off all along.

Hi Andrew, i have only just caught up with this thread.

Just FYI, below are some frequency response measurements i did, comparing my killerdac with an Auralic Vega DAC. I won't have to tell you which curve is which! The rolloff seems consistent with the classic NOS "droop". I wonder whether that is what you are hearing, or whether there is some other factor...?

Anyway, it seems to me that one possible way to address that is to -- shock, horror -- apply a compensating filter to our digital files so that they are effectively flat when played back on a NOS dac like the killerdac. Since we often burn our own CDs, that should not be difficult... or even those of us dabbling with .. gulp ... computer transports.


Offline ozmillsy

  • Hero Member
  • *****
  • Posts: 2249
  • Liked: 277
Re: Ozmillsy's KillerDAC
« Reply #108 on: August 07, 2014, 07:39:22 AM »
Hey rab, that is an impressively flat graph (roll off aside).  Some people have commented that the Killerdac seems to have a midrange push, but this is not what your graph shows.

Yes, the natural rolloff of the tda chip (in NOS mode) could definitely be part of it, no doubt.    But also, my dac had a bunch of parts in it that were rich and almost bordering on thick,  and this was robbing it of some top end extension (there are always trade offs).

I'm happy with where it is at now, and am nearly ready to demo it at the next GTG I organise.  My tannoy system has always been a bit thick in delivery, and I had incorrectly assumed it was the tannoys.   The penny dropped when I was doing some vinyl to digital comparisons (digital recordings of my vinyl rig used for the comparo),  then I did some R2R comparisons (using same method, digital recordings taken from tape).    The dac was the problem.

But the changes I have made, are worth trying (1 at a time), it sounds soooo much more extended now.

In summary,
* chokes  (massive change)
* resistors (allen bradley changed to 2w shinkoh)
* output tubes (changed miniwatt SQ tubes)
* cap (dubilier to plessey)

I have also changed my rectifier to a brown base KBZ type, which I havent documented here yet.

More possible changes are,  chip rolling and there is another set of AB resistors (100k) on the output sockets.   But I dont think I need to go there.

When you are next in sydney, come visit me rab, the system is radically different to when you last heard it (in a different house and room, with any number of other changes).
« Last Edit: August 12, 2014, 11:14:12 AM by ozmillsy »
It's all about the music,, not the equipment.

Offline rab

  • Administrator
  • Full Member
  • *****
  • Posts: 232
  • Liked: 36
Re: Ozmillsy's KillerDAC
« Reply #109 on: August 11, 2014, 10:33:40 PM »
Great stuff Andrew. I dip into this stuff from time to time as I want to improve the sound of my Tram2 preamp and Modjeski DD amp. Always chasing more body (Steve's 'flesh and bone'), changing caps, resistors, etc.

Thanks also for the invitation to visit: i am long overdue for a visit to the Blue Mountains and would love to hear what you have done. However, i am preparing for a big OS trip in September so it will be a while... maybe next year.

- r.

Offline zenelectro

  • Hero Member
  • *****
  • Posts: 825
  • Liked: 177
Re: Ozmillsy's KillerDAC
« Reply #110 on: August 12, 2014, 12:08:34 AM »
Hey rab, that is an impressively flat graph (roll off aside).  Some people have commented that the Killerdac seems to have a midrange push, but this is not what your graph shows.

Yes, the natural rolloff of the tda chip could definitely be part of it, no doubt.


There's no roll off of the TDA chip. All 0 x oversampling DAC's (at 44.1KHz) will have this roll off due to the absence of digital filter.
It's about -2.5 to 3dB at 20kHz and it's the sinx/x response.
I would also expect any other 0 x OS DAC to be flat in pass band unless they are broken. Any midrange warmth or boost is just added
harmonics / distortion. 

cheers

Terry

Offline ozmillsy

  • Hero Member
  • *****
  • Posts: 2249
  • Liked: 277
Re: Ozmillsy's KillerDAC
« Reply #111 on: August 12, 2014, 11:13:53 AM »
Yes,  I should have stated TDA in nos mode.  Have fixed that.   Cheers
It's all about the music,, not the equipment.

Offline vitavoxdude

  • Beauty is in the ear of the beholder
  • Hero Member
  • *****
  • Posts: 1175
  • Liked: 71
  • Caring and sharing
Re: Ozmillsy's KillerDAC
« Reply #112 on: August 12, 2014, 02:35:10 PM »
:) Just want to mention that IMV having a flat FR measurement on front end eqp bears very little effect in the big picture for the total room response which is what we hear.  Try measuring any speakers 3 or so meters away in the listening seat for a real shock.  Peaks and troughs introduced by the room and speaker combination will totally dominate any FR measurements.

What Oz has experienced by changing out components for the same or similar value devices is differences to tonality and hf extension.  Great write up by the way on the effects of changing various components.  Keep it coming.

V
We all like different things so lets all agree to disagree and if any common ground is found then worship it.  Mine is the KD hence being present on this forum.

Offline ozmillsy

  • Hero Member
  • *****
  • Posts: 2249
  • Liked: 277
Re: Ozmillsy's KillerDAC
« Reply #113 on: September 09, 2014, 09:43:47 PM »
:) Just want to mention that IMV having a flat FR measurement on front end eqp bears very little effect in the big picture for the total room response which is what we hear.  Try measuring any speakers 3 or so meters away in the listening seat for a real shock.  Peaks and troughs introduced by the room and speaker combination will totally dominate any FR measurements.
I've done that (measuring HF at seating position), and totally agree.    It's why people measure HF near field.       Am using my ears in this room,   I'm not trying to measure 'flat'.     I am shooting for similar HF extension that I am getting from my R2R,   same room, same recordings (xfer'd), different front end.

Quote
What Oz has experienced by changing out components for the same or similar value devices is differences to tonality and hf extension.  Great write up by the way on the effects of changing various components.  Keep it coming.
Changing the parts I have done so far in the dac is having quite a dramatic effect.   I'm really happy with where it is sitting,  but there is a smidge more to wrestle out of the system.

Tried some KLEI IC's between dac and VC last week,  and what I got was a dramatic change in transparency,  night and day stuff.   Can hear the difference in the other room.  ;)   But, with these IC's,  sibilance was more pronounced.    So,,,,, does that mean retain the IC's and dial back some changes in the DAC,   or stick with my darker IC's?    Also have the treble settings on the Golds, that are in the mix.    Fun, fun.      :-\

I feel like the KLEI's are letting more through,  and therefore I could have gone a smidge too far with the dac.   It's getting to the pointy end of dialling in the dac (on this system/room) now ,,,,,,,,,, and then I'll move house again.  ???   

« Last Edit: September 09, 2014, 09:59:55 PM by ozmillsy »
It's all about the music,, not the equipment.

Offline zenelectro

  • Hero Member
  • *****
  • Posts: 825
  • Liked: 177
Re: Ozmillsy's KillerDAC
« Reply #114 on: September 09, 2014, 10:10:33 PM »

Tried some KLEI IC's between dac and VC last week,  and what I got was a dramatic change in transparency,  night and day stuff.   Can hear the difference in the other room.  ;)   But, with these IC's,  sibilance was more pronounced.    So,,,,, does that mean retain the IC's and dial back some changes in the DAC,   or stick with my darker IC's?    Also have the treble settings on the Golds, that are in the mix.    Fun, fun.      :-\

I feel like the KLEI's are letting more through,  and therefore I could have gone a smidge too far with the dac.   It's getting to the pointy end of dialling in the dac (on this system/room) now.   

Rule 101, if you change something and get more sibilance then that usually goes hand in hand with 'detail' 'transparency' etc etc.

So then the challenge is to identify whether:

a/ there is actually more real transparency and the sibilance is being 'let through' as a result,  or
b/ the added sibilance is fooling you into believing there is more transparency through added 'fake' detail.

If you can get a handle on these types of scenarios, you've got good ears, or just a lot of experience.

To me, the right decisions when doing this sort of analysis is more important for moving forward in the big scheme of things than lots of floury
audiophile descriptive terms that don't usually mean a lot.   :)

One thing I'd like to know is are these interconnect shielded?

Z



Offline ozmillsy

  • Hero Member
  • *****
  • Posts: 2249
  • Liked: 277
Re: Ozmillsy's KillerDAC
« Reply #115 on: September 09, 2014, 10:25:26 PM »
Rule 101, if you change something and get more sibilance then that usually goes hand in hand with 'detail' 'transparency' etc etc.

So then the challenge is to identify whether:

a/ there is actually more real transparency and the sibilance is being 'let through' as a result,  or
b/ the added sibilance is fooling you into believing there is more transparency through added 'fake' detail.

I havent been able to pin point a or b just yet.    I will say that the clarity I was getting didnt feel fake,  there was inner detail in the music that didnt sound artificial to my ears, some very nice texture coming through, throughout the range.    There was just the very top end wart, that drew attention to itself.    I need to experiment some more. 

Quote
To me, the right decisions when doing this sort of analysis is more important for moving forward in the big scheme of things than lots of floury
audiophile descriptive terms that don't usually mean a lot.   :)
hehe,   :D  , I dont mind floury terms.     For example,  ask Steven an opinion and he often says   "it doesnt sound like real music !!"    :-\  ???   ::)   Half the fun is trying to derive meaning from the words of wisdom.   8)      [he didnt say that about the dac this time :)]

Quote
One thing I'd like to know is are these interconnect shielded?
They were a new KLEI design that Craig bought around.   Details of what is in them are not shared by KE,  but they felt light, I had to be very delicate with them.  My guess is no shielding.  They were only here for his visit.     I'd like to try the stock gZero6 cables,  and play around a little further. 
It's all about the music,, not the equipment.

Offline ozmillsy

  • Hero Member
  • *****
  • Posts: 2249
  • Liked: 277
Re: Ozmillsy's KillerDAC
« Reply #116 on: September 13, 2014, 10:19:47 AM »
On Rawls visit last week, he gave me a useful tip.    He was peering into the dac, looking at the changes I had made,  and I could see the cogs cranking away,  and then he came out with the tip to try.

The way I had run the i2s wires from input sockets to receiver chip socket on the dac board could be improved.     I had run twisted pairs from each socket around the top of the case,  and down into the dac board.  At the dacboard end I had twisted the ground wires together and plugged them into the dac board.   (did all this with Steven)   
Here's how it was.....

You can see the i2s sockets in the top middle,  the wires running up,  then they pop out on the right going into the dac board.

What Rawl suggested was;
- dont run twisted pairs
- only use 1 ground (from the 3 x sockets, because they are already joined at the transport end)
- take them direct from input sockets to the dac board in free air across,   dont run them around the outside.

I dont have alot of time for tuning/tinkering activities,  my days/nights are consumed doing other things.   But I happened to fall out of bed an hour early this morning, and decided to try his change.   

Here is how it looks now,,,,,,

The picture is a bit of an optical illusion.  The i2s wires are not as close to the tubes, as they look in that photo.   They are a good couple of inches away.

Everything still works.    Is it better?   yeah,   in the usual i2s way, more coherency and clarity.   

Thanks Rawl !     :)
« Last Edit: September 13, 2014, 11:02:15 AM by ozmillsy »
It's all about the music,, not the equipment.

Offline ozmillsy

  • Hero Member
  • *****
  • Posts: 2249
  • Liked: 277
Re: Ozmillsy's KillerDAC
« Reply #117 on: September 15, 2014, 09:18:36 AM »
  I'd like to try the stock gZero6 cables,  and play around a little further.
Ok, gZero6 cables are in.  And they sound good.  I needed to do a little bit of tuning,   but there is a nice balance now.

1 new problem is there is more noticable system hum.  Given the feel of these cables, i'm pretty sure they are not shielded.  Could be either hum that was there, and the cables are more transparently letting it through,  or they are adding to it by picking up environmentals.    The IC's they replaced are shielded.

There are always pros/cons.   I'll try running them via a different path.
« Last Edit: September 15, 2014, 03:16:28 PM by ozmillsy »
It's all about the music,, not the equipment.