G-01 & 10MHz Output Cable Impedance

[SCAudiophile]

Caelin,...there is another thread where this has come up. I hope you don't mind that I am about to put your quoted posted into that thread. It helped me alot and I'm sure it will help others.

[CGabriel]

Originally Posted by zephyr24069

Caelin,...there is another thread where this has come up. I hope you don't mind that I am about to put your quoted posted into that thread. It helped me alot and I'm sure it will help others.

Of course, no problem.

[chessman]

Caelin, thank you for bringing science to this discussion.

[SCAudiophile]

Definitely needed science as we had a number of manufacturers putting specific connectors on their devices (Esoteric, SRS, Cybershaft, others...), documenting in manuals the need for proper cabling and then coming out later and saying there was no difference; very confusing to those of us without EE degrees!

[kb7215]

First, thanks to Caelin for his technical explanation regarding 50 & 75 Ohm BNC clock cables. I'd like to add the following points. Quoting from the G-01 Owner's Manual:

1. "CLOCK OUT: Rectangular wave (TTL level/75 Ohm)"
2. "10MHz OUT: Sine wave: 0.5+/-0.1 Vrms/50 Ohm"

The CLOCK OUT (A1/2; B1/2; C1/2) BNC connectors are 75 Ohm. The 10MHz OUT BNC connectors are 50 Ohm.

The SRS Data Sheet for the PRS10 Rubidium Frequency Standard lists the following:

1. "Output Frequency: 10MHz sine wave; Amplitude: 0.5 Vrms, +/-10%"

From the G-01 Brochure:

"A 10MHz square wave can be output from any of the output terminals of A, B and C [75 Ohm BNC connectors]. In addition, two systems are designed for devices equipped with 10MHz sine wave requirements [50 Ohm BNC connectors]."

So it looks like the 75 Ohm CLOCK OUT signal is totally different than the 50 Ohm 10MHz OUT. The first one is a rectangular wave and the second is a sine wave. A close examination of the 75 Ohm BNC and 50 Ohm BNC connectors show they are totally different in construction. There is an instructive image on the Wikipedia 'BNC Connector' page clearly showing the differences. Yes, I know a 50 Ohm BNC will connect to a 75 Ohm BNC and vice versa.

My question: does it make any difference to the ongoing discussion if the 10MHz output (50 Ohm BNC) is a sine wave and the CLOCK OUT output (75 Ohm BNC) is a rectangular wave?

Ken

[CGabriel]

If given a choice and all components are compatible I would choose the 50 ohm 10Mhz sinewave. It has fewer real world issues because the signal is a sinewave and has no harmonics.

[SCAudiophile]

I'm going to wake this thread up (and the parallel thread where the same discussion is duplicated) as over the past 2 months I've been doing alot of testing that overlaps with subject of 50-ohm versus 75-ohm clock cables, their significance and their interchangeability. There is also another audiophile or 2 that I know doing similar testing for his own purposes with different equipment in terms of master clocks, music server versus player/transport/DAC, etc....

I've had the good fortune to have a 2nd OCXO-based 10Mhz Master clock here on loan that is a prototype with even better phase noise characteristics and even an improved (Schottky barrier diode based) power supply. More about why this clock has been here the last couple of months when the info can become public knowledge but for now, I can only share that the OP14 and better levels of precision and lower noise characteristics it brings are but a stepping stone to something even better and off the charts...

Well here goes...

Cybershaft 10 Mhz Clock#1 (very low phase noise) OCXO clock, 50-ohm output calibrated using Symmetricom (TSC5115A) and BVA (8607) reference units;
Phase Noise:
-111.3 dBc/Hz @ 1Hz offset *
-133.1 dBc/Hz @ 10Hz offset ***
-142.2 dBc/Hz @ 100Hz offset ***

Cybershaft 10 Mhz Clock #2 (very low phase noise) OCXO clock, 50-ohm output,
Schottky barrier diode-enabled power supply, calibrated using Symmetricom (TSC5115A) and BVA (8607) reference units;
Phase Noise:
-114.7 dBc/Hz @ 1Hz offset ***
-133.5 dBc/Hz @ 10Hz offset ***
-142.2 dBc/Hz @ 100Hz offset ***

*Dual listening test runs over a 1 month period with both Cybershaft clocks producing 10 Mhz and the OCXO clock in the Esoteric D-02
producing 22.xxx Mhz 75-ohm clock signals respectively.

*The phase noise specs. given above beat another manufacturer's Rb-oscillator based top 2 units based upon specs I was sent recently from that manufacturer of < -100 dBc/Hz @ 1Hz offset, < -130.0 dBc/Hz @ 10Hz offset and < -140.0 dBc/Hz @ 100Hz offset. This called out only to
underline the fact that the 2 Cybershaft clocks used for this test (for 10 Mhz signal production) are at least reference grade.

<Cybershaft 10Mhz 50-ohm output drives clocking accuracy baseline for D-02 DAC>

Esoteric D-02 DAC with 10 Mhz 50-ohm input from Cybershaft

<75-ohm output from D-02 DAC to P-02 Transport>

Esoteric P-02 Transport taking in 22.xxx Mhz "MCLK" frequency mode over 75-ohm connect slaved to master clock in D-02

Reference cables;

Wireworld Platinum Starlight 75-ohm BNC:BNC, Kubala-Sosna EMOTION 75-ohm BNC:BNC (1.5m)
Esoteric/MEXCEL 8N-A2000 50-ohm BNC:BNC (1.5m)
2 other 50-ohm high quality cable from 2 well-known manufacturers (1.5m, 2.0m)
2 other 75-ohm high quality cable from 2 well-known manufacturer (1.5m, 2.0m)
1 50-ohm high quality cable with a 75-ohm connector from a well-known manufacturer

Results (at a very high-level);

Please note this is NOT about which cable is better for 50-ohm or 75-ohm…

The only purpose of the tests written about here is to determine whether the thesis that use of 50-ohm or 75-ohm BNC clocking cables matters on a matching characteristic impedance circuit (or not) when the lengths are short for 10 Mhz, 22.xx Mhz (master clock frequencies) and frequency-multiple re-clocking frequencies (44.1, 88.2, 176.4, etc…)…

All of the above cables and clocks used in all possible combinations, testing occurred using a variety of audiophile reference material from genres ranging from pipe organ and taiko drumming to large ensemble and small ensemble classical, jazz, big band, vocal (male and female), hard rock and progressive rock…

Results (mine, your mileage may vary):

- use of 75-ohm and 50-ohm spec-compliant cables (i.e. cable and connector both of that characteristic impedance) judged equivalent on
75-ohm MCLCK circuit between the P-02 and D-02 in categories such as tonality, sound-stage depth, width and height, leading and trailing edges of notes, overall transients, voices and ensemble play, side to side and front to back image coherency, and musical impact of both low volume and high volume delicate as well as highly dynamic passages HOWEVER….

- 50-ohm spec-compliant cables judged superior and worked perfectly on a 75-ohm MCLCK circuit; small but noticeable advantage in pacing of complex passages (large-scale orchestral and driving big band full ensemble play showed this off best), and ‘image specificity’, i.e. the exact right size and location of a given voice, instrument, etc on the stage in all 3 dimensions as well as air around instruments and voices, etc…..….(very (nice) surprises here)

- 50-ohm cables judged superior in all categories on a pure 50-ohm circuit (10 Mhz); best result in every category including all those listed above;
best air around instruments, image and soundstage specificity, 3-dimensionality, tonality and musical realism and correctness…all around winner is to use a proper 50-ohm SOTA cable on a 10 Mhz connection (see below….)

- 75-ohm cables judged inferior and prone to pacing problems and some lack of image specificity, decrease in soundstage depth in particular and breadth secondarily on a 50-ohm 10 Mhz signal circuit. Also heard undeniable evidence of lessening of air around instruments and decrease in transients around sax, drum and cymbals,….several other areas where I did not like the result (musically) compared to better results above….

- surprise was the 50-ohm cable with 75-ohm connectors on a 75-ohm MCLK circuits. All tests worked out and showed the best of the results above in all categories EXCEPT one key area, bass & sub-bass impact and extension. For some unknown reason there was a shortfall, i.e. early roll-off of bass and lessening in intensity on several test tracks with pronounced single drum strikes of varying amplitude and frequency…No telling why and this will no doubt require alot more testing to try to nail down but it was undeniably present…

- 50-ohm cable with 75-ohm connector not testing on pure 50-ohm circuit for obvious reasons…

In short, even at shorter lengths, cable rating (50 versus 75-ohm) can make a difference but only in certain (positive or negative) ways depending upon configuration…

I’m sure this will generate some healthy dialogue and debate which is definitely appreciated!

[crwilli]

Ha, I would love to discuss and debate but I would first have to understand what you wrote. I don't even know what my BNC cable is, 50 or 75 ohm. . I remember reading a post here regarding 50ohm versus 75ohm cables for clocks but didn't understand that one either. I suppose what I get from your post is that I should try and find a 50ohm cable if mine is a 75...

I do have one question.

Have you run your system through all combinations of the Master frequencies (44, 48, 44EXP, 48EXP) and the specific outputs of 100Hz, 10MHz, 22.5MHz etc).

I am a little OCD when it comes to settings (one of the reasons I am selling my MEN220 and have packed up all my room measurement gear). My issue is that I fundamentally do not understand what the settings of the G-02 are doing so it is just alchemy to me.

[SCAudiophile]

Originally Posted by crwilli

Ha, I would love to discuss and debate but I would first have to understand what you wrote. I don't even know what my BNC cable is, 50 or 75 ohm. . I remember reading a post here regarding 50ohm versus 75ohm cables for clocks but didn't understand that one either. I suppose what I get from your post is that I should try and find a 50ohm cable if mine is a 75...

I do have one question.

Have you run your system through all combinations of the Master frequencies (44, 48, 44EXP, 48EXP) and the specific outputs of 100Hz, 10MHz, 22.5MHz etc).

I am a little OCD when it comes to settings (one of the reasons I am selling my MEN220 and have packed up all my room measurement gear). My issue is that I fundamentally do not understand what the settings of the G-02 are doing so it is just alchemy to me.

Chances are your BNC cable is a 75-ohm....is it WW Platinum Starlight or something else?

As to the other question(s); I've owned P-03U/D-03/G-0s clock setup back in the day followed by no clock for 2 years, then adding back in a G-03x followed by moving to the P-02/D-02 with no clock (except for the one in the D-02 DAC which is as good to my ears as the old G-03x was....) then almost a year later adding in the 10Mhz Cybershaft master reference clock which has superior phase noise and Allen Variance test results to everything in its price class and classes up to $20K+ units. From what I can tell, the only thing it does not eclipse is a $35K BVA 8607 unit...

With the G-0s and 03 separates I tested 44.1, 88.2, 176.4 (all multiples of 44.1 Khz) and 48Khz clock freqs, etc....I found the 48 setting only benefitted me when I was playing back some video, and overall I did not love what it did for audio or video more than my destination clock freq at the time, 176.4 Khz that was my tried and true setting with the G-0s, G-03x, D-03 & D-092 (when used without an outboard clock). Remember that clock frequency is not really about speed of playback but rather the number of times per second that the units sync from a common reference. Many Esoteric owners in Europe and Asia still use only 44.1 and prefer it; I happen to think that for soundstaging, musicality, etc....176.4 Khz is best on these older units.

The short story is that when the P-02/D-02 arrived, I had the G-03X and set the G-03X for 176.4Khz as before. Playback was as expected, even better than the 03 generation units and 176.4 Khz worked very well coming from the outboard G-03x clock. After a while kicking into my own brand of audio-nut OCD, I had to start experimenting again; what I found was that the new D-02's internal OCX clock was much more quiet and precise than its D-03 predecessor and at least equivalent at 176.4Khz. I sold off the G-03x from there and went forward very happily. Shortly thereafter, a friend also had the 02 units and found that 22.xxx Mhz was the ideal (musical) output clock frequency from the D-02 as a standalone. I tried it and there was no looking back, at least for my rig.

The only modification since that time is the (1 year+ ago) add of a ultra-high precision, low phase noise, 10Mhz master reference clock that I used to set a more solid baseline than the PLL circuit in the D-02. Still have the D-02 set to 22.xxx Mhz but the D-02 is baselined by the Cybershaft. Since that time also the number of hardcore clock-friendly guys I've spoken to using 10 Mhz clocks from SRS, Esoteric, Cybershaft, etc...is on the rise. Even more interesting is the introduction of the all-10Mhz Grandioso G1 clock. Ultimately, I'll probably go there or some lower phase noise higher precision equivalent but for now, my config is as documented for the reasons I state above and below. 10 Mhz is the best universal clock frequency out there from what my studio associates tell me so long-term I'll probably go to an all 10Mhz clocking 'backplane'.

At first all this reclocking versus master clocking mumbo-jumbo comes off as just a boatload of numbers (frequencies and bit sample rates juxtaposed); after alot of reading and trial and error over the years, I understand a bit more but am no expert. I'm happy to jump on the phone and discuss in more detail any time....

[adol290]

SCAudiophile,


Thanks for all the great information on your clock.

I have been thinking of getting a clock for my K-03x for a while now. The esoteric option is a bit out of my range, so after reading your posts here and a few other places, I am considering the Cybershaft.

Do you prefer the OCXO clock, over the RB. Would you still suggest getting a clock that has both?

Is the prototype OCXO from the same manufacture?, is it a single or dual clock?