MHDT Labs: R-2R NOS Tube DACs
Mar 25, 2016 at 3:39 AM Post #106 of 897
How does the vibration travel through part of the chassis into the cone and not board onto the rest of the components?

This is one more reason to separate the torroidal from the DAC housing itself. What started as an upgrade from the smaller torroidal to r-core transformers could transition into simply removing the toroidal from the DAC. Or perhaps swapping the bolt for a longer one and placing an isolation pad under the transformer would do.
 
Mar 25, 2016 at 4:11 AM Post #107 of 897
"How does the vibration travel through part of the chassis into the cone and not board onto the rest of the components?"
 
It does not completely, but it improves the situation: the sound of the Stockholm with the cones has more definition, it is sonically obvious. Ideally you are right and removing the toroidal from the DAC would be the best. However, I want to retain myself from going too far, as this is endless. So cones + filtering power chord + PCM56K (or AD1856K) + category "one" tube + good audio cable + tube amplifier (Analog Squared Paper TU-05 for me) will allow me to reach my goal: a truly remarkable audio system with a sound that I just love. It will always be possible to improve this or that but this also produces frustrations.
 
- chris
 
Mar 25, 2016 at 4:22 AM Post #108 of 897
Just search for PCM56K on amazon: the provider's name is utsource.
 
Once you eliminate the defective chips, I confirm the PCM56K produces slightly better results than PCM56J: better treeble (more natural), better definition on complex signals
 
Mar 26, 2016 at 7:40 AM Post #109 of 897
utsource is a Chinese supplier. I bought many components from this source. Some of them were bad quality, some of them were not working or failed after a short time of usage.
 
When I owned MHDT Havana I tested many PCM56P chips: Korean, Malaysian, Japanese, K,J,L grade.
I suppose all of them.
 
Bought many via ebay, from chinese suppliers, some of them I got from my friends.
 
but the best sonic results I achieved with desoldered ones from the old hi-end equipment.
 

 

 


 
Mar 29, 2016 at 12:14 AM Post #110 of 897
Just got a bunch of 6N3P-DR tubes in from across the pond. I'll have to compare them directly against the GE Triple Mica, but they are excellent so far.
 
They could eat American 5670 tubes for breakfast. 
 
This is a 6N3P-DR next to a GE 5670 Triple Mica (5 Star White):
 

 
Unlike the 6N2P-EV, you don't need an adapter for them. The DR version is very different internally from the standard 6N3P, so if you want to try one out make sure you get the real deal. I couldn't find any in the USA, so I ordered these a few weeks ago.
 
Apr 1, 2016 at 7:27 AM Post #112 of 897
After not hearing differences with firmware when using BNC, I pinged Jiun.

He confirmed that the firmware changes only impact USB input. My ears haven't failed me!

I picked up a Mutec MC-3+ USB to S/P-DIF converter... I planned to review and return the loaner, and ended up buying two (one per Atlantis). Yes, I spent more on a converter than the DAC itself.
 
Apr 1, 2016 at 7:44 AM Post #113 of 897
After not hearing differences with firmware when using BNC, I pinged Jiun.

He confirmed that the firmware changes only impact USB input. My ears haven't failed me!

I picked up a Mutec MC-3+ USB to S/P-DIF converter... I planned to review and return the loaner, and ended up buying two (one per Atlantis). Yes, I spent more on a converter than the DAC itself.

 
What specifically does the firmware update address for the USB input?
 
Apr 1, 2016 at 7:52 AM Post #114 of 897
It's not an update so much as three slightly different options.

The stock firmware is left justified, then there's one labeled as I2S, and one as S/P-DIF (which also makes it read as an S/P-DIF on the computer and ASIO no longer works).

The changes are more subtle than DAC chips, tubes, or even USB converters, but I prefer the I2S firmware over stock.

If you PM your email I can send you the firmware that Jiun sent me. It has versions for all the current dacs.
 
Apr 3, 2016 at 2:24 PM Post #116 of 897
Any MHDT owners who want in on my 5670 tube blowout extravaganza, PM me. I'll give you an extra discount on the tubes.
 
15 tubes that I spent >$300 on in a bundle for much less. And if you have an MHDT, I'll slip in a 6н3п-др (6N3P-DR), which is now tied for my favorite alongside the GE Triple Mica.
 
http://www.head-fi.org/t/803862/5670-2c51-396a-tube-blowout
  1. Western Electric 2C51 / 396A - No date code (1940s), 6700 / 7300 (minimum good 3150) (D [ ] Square Getter)
  2. Western Electric JW 2C51 / 396A - Date Code 513, 5900 / 5600 (minimum good 3150) (D [ ] Square Getter)
  3. RCA Command 5670
  4. CBS Industrial 5670
  5. Raytheon JAN 5670
  6. Sylvania JAN 2C51
  7. Raytheon CK5670
  8. Sylvania JHS 2C51
  9. CBS JHY-2C51
  10. Westinghouse 5670
  11. 3x Sylvania JAN 5670 (Halo Getter, packed 06/69) - Perfect lettering, all the same date
  12. 2x GE 5670 Triple Mica (my personal favorite 5670 tube

 
Apr 8, 2016 at 1:24 PM Post #117 of 897
My new board came. Jiun is extremely helpful guy. I just installed it and it works flawlessly so far.
 
 
Apr 21, 2016 at 2:20 PM Post #120 of 897
On the first post in this thread (and on their website) there is a row called "Output format" that has bit quantities. For example, the Stockholm 2 is 16bit, Atlantis 20bit and Pagoda 24bit... what output are we talking about here? Is this the down/up sampling that occurs before the actual conversion to audio? (that's the only thing that makes sense to me)
 
I have a good bit of hi-rez music, but I have *never* been able to reliably A/B the difference of a 192 file, but have been able to A/B a 96 one time only. So I'm not sure that I personally would care about much past 16bit except for the processing up or down that may be occurring that can cause issues in the conversion. 
 

Users who are viewing this thread

Back
Top