Discussion in 'General Audio Discussion' started by atomicbob, Sep 28, 2015.
@batriq there is an Eitr and Gungnir Multibit available from @Ray if the FS section.
Yeah, I saw. I was hoping to buy the Gungnir Multibit with Analog2 if/when it comes out.
Here's a pic BTW
For what it's worth I find Modi Multibit to be very excellent with SEX 3.0 and my cans, and yet Gungnir Multibit with that combo is even more excellent. Both of them I consider pretty sweet devices, particularly the latter which originally opened my eyes (ears) to the multibit sound. Many thanks to the kind local enthusiast who loaned me his personal unit!
Eventually I bought the Modi MB - but given enough time I went and got myself a Gungnir Multibit of my own. I am very pleased with the result, it is a solid upgrade. Modi Multibit does amazingly well for it's price, but it sounds somewhat lesser in comparison to it's big brother. The manner in which Gungnir Multibit readily serves up low-level detail and microdynamics on a smooth, and spacious stage is just addicting as hell.
Thanks @slankoe. It's great to hear that the SEX scales with the Gungnir Multibit. BTW, I put in my L-pad, which allows me to use more of the volume pot. Here are the instructions (it's been a while so I think my images were taken down(?): https://bottlehead.com/smf/index.php?topic=7252.0
I was in a hurry and didn't create the mousepad cutouts necessary to avoid blocking ventilation holes on the dac1421. Result after brief photo session was black rubber marks on the dac1421 top surface. Took awhile to clean off. Custom cut mouse pad separators now created.
They are 5.5 cm in diameter. Need to leave a minimum of 0.5 cm between them. It might be too close for your wonderful sounding Liquid Glass.
Those Psvane 6SN7-SE tubes aren't going to fit. Not enough space behind and to the sides for them to fit. Negative clearance. Unless I find some 6 cm high tube socket risers...
I have been thinking about trying the Psvane CV181 tubes.
Sorry @Cspirou , I somehow missed this. Yes, there will be measurements sometime. Not sure they will happen before my multi-week break.
No prob Bob. I just like knowing what's in the measurement queue no matter where it's prioritized.
atomicbob's nuclear schiit
Listening to this system for several weeks. Was demonstrated at the Seattle Summer Can Meet.
Important to note Modi MB is v2 firmware in this system.
Hardware block diagram:
IVS setup for this system:
Digital flutter management (jitter):
Digital wow management (wander):
Expanded y-axis view of LiveClock with - without - with Rubidium reference disciplining:
All that can be measured during FS725 disciplining of LiveClock (before and after) is HP 5384 LSB toggle. Middle section is LiveClock running on internal clock alone. Not too shabby but not as good as with FS725 engaged. HP5384 frequency counter time reference was obtained from a Jackson Labs Fury GPSDO with Allan Deviation an order of magnitude more precise than the Rubidium time reference.
This fellow is doing something similar in the world of karts - a visual analogy.
For those with a burning desire to know system BOM:
The proof is in the listening. I've been very pleased with the results of this system's synergy.
Before anyone contemplates replicating this system, I can justify the FS-725 Rubidium time reference for lab frequency counters, Antelope LiveClock provides house clock to project studio as needed, and the MC-3+ was already in lab inventory.
Consider the Eitr, Modi MB v2 firmware, Lyr3 + PS Vane 6SN7-UK, HD800 modded to suit personal preferences and my IVS will deliver a high percentage of this system's performance.
Those short term vs long term comparisons of the signal stability are most intriguing.
It makes me wonder what identifiable acoustic aspects result from those series of tweaks.
And the Fury reference master clock is intriguing to contemplate as well.
Thanks for that in depth look see.
I have been more and more pleased with the combination of Gungnir Multibit (recent purchase, with Gen5), SEX 2.1 that I built, and stock HD650. I no longer feel the need to tweak or change my headphone setup. Most recently my son kicked me off the listening chair:
The little dongule is a voltage divider I built since the SEX is too powerful.
Clock wander updated - measured with
Keysight 53220A + Jackson Labs Fury time reference (stratum 1 ref)
Clock wander zoom for Antelope Live Clock vs FS725 rubidium time reference
Clock wander - Mutec MC-3+ internal clock after 4 hrs warmup
A transparent, engaging system. With such great time references I seem to lose track of time during recreational listening sessions with this combination.
Nuclear ZDT Jr.
System Architecture Diagram
Now that I have the residual hum below audible thresholds, ZDT Jr. joins the illustrious headphone amp collection worthy of the atomic treatment. When I am really wanting to lose track of time, putting the final icing on the cake, I substitute yggdrasil A2 for the Modi MB v2.
I just had to laugh.
Enlarge the pic and it's all topped of by a ModiMB .
It's like the least expensive except for the cables and s/w, part of the whole stack.
Yeah reducing hum on tube gear can be 'problematic'.
I just tested my BIG eyerrrrrnnnnn amp after its latest re-wire job, and the hum is now WAY better.
Still a ways to go yet but round 3 is already underway.
Rubidium Clock --> a Modi
You are an artist.
Thank-you, it is a balance between science and art.
Modi MB v2 firmware inferred jitter - with Eitr
Modi MB v2 firmware inferred jitter - with Eitr + Mutec MC-3+USB
Modi MB v2 firmware inferred jitter - with Mutec MC-3+USB + Antelope LiveClock + FS725 rubidium time reference
Noise floor looks lower but the trace was shifted downward by 9 dB so it would display nicely on the next screenshot. Noise floor is actually the same as the previous two measurements.
Comparison of the three reclocking / clock management methods
Note the blue trace has been shifted downward by 9 dB for visual clarity
Comparison of the three at 350 Hz BW zoom
Modi MB v2 firmware responded well to lowered phase noise of FS725 + LiveClock disciplining of MC-3+ reclocking spdif.
This is an excellent jitter response, one few other DACs achieve, even with atomic clock disciplining.
Ah, the most affordable Dante file delivery yet -- just two channels and PoE power, right?
@atomicbob Is there a reason you don't feed the 10MHz ref clock to the Mutec directly, is its PLL performance poor?
Correct. And easy to substitute LPS for SMPS on appropriate PoE adapter cable if desired. I used the supplied SMPS and my Noise Nuke. Worked well.
Mutec MC-3+ didn't perform as well directly connected to 10 MHz reference as when the LiveClock was used to down convert 10 MHz to word clock, typically 88 KHz for me.
Separate names with a comma.