Skip to main content

Hi,

I'm finding that as I migrate my studio more and more to the digital end of things, I'm picking up more and more equipment with word clock BNC connectors. The problem is, some have an IN but not an OUT, so how do I chain all of them together if I want to use one clock source (as I understand I'm supposed to)?

Is there such a thing as a clock splitter? Something that has say, one BNC connector IN with four or more BNC word clock OUT connectors? Maybe even BNC 'Y' cords?

I am using an RME Digiface sound card and digital breakout box which I was intending to use as the clock source. I'm also using a Frontier Designs Tango 24 interface, a Focusrite MH-442 DCA card in a Focusrite ISA 428. To patch them all together, I'm using an M-Audio Digipatch unit. Part of the reason for that is to be able to route the output from a Tascam DA-88 and DA-38 directly into my computer without exiting the digital domain and without having to re-wire everything when I want to switch routings. Each of the Tascam TDIF-1/ADAT converters has a word clock IN but not OUT. Neither does the DA-38 . so I don't see a way to "daisy-chain" all of these devices together using one source clock.

The other question is, which of these devices would work best as the main clock source? Does it matter? I have two Aardvark Q10 interfaces as well. I know I'm going to sell one but I may keep the other if it behaves itself when it gets a new neighbor (the RME). I have heard that Aardvark has excellent word clocks. Or, do I even need to synch everything via word clock?

Any suggestions?

Topic Tags

Comments

anonymous Mon, 01/17/2005 - 12:52

I hope the following can help:

1/ In a digital rig, there must be indeed only one master clock, and all the other gear in the studio will be sync'ed (=slaved) to this clock.

2/ You have two possibilities to slave your equipment: either through WC using a proper 75 Ohms BNC cable (we'll come back to this later), or through any digital audio interface AES/EBU, SPDIF, ADAT, MADI, etc. In both cases, you need to configure the slaved equipment so that it will choose the correct clock source.

3/ Your best choice, if you can afford that, is to have a dedicated master Wordclock delivering the clock signal to all other equipment in a star topology, using good cables of the same length, properly terminated (we'll cover that later as well). Examples of such products are BigBen from Apogee, AardvarkSyncII from Ardvark and Nanoclocks from Rosendahl.

4/ Your second best choice, if you cannot afford 3/ is to use the best clock amongst your digital gear as your master clock (in your case, I am still hesitating between the Aardvark Q10 and the RME Multiface, maybe someone could chime in), and slave the other stuff using a daisy chain WC cable. In such a case, you cannot go for a star topology as you generally have only one WC out connector.
The daisy chain is as follows: on the master clock side, you need to attach a 75 Ohms T connector on WC OUT. On one side of the T you will connect a 75 Ohms resistor for impedance matching purposes, on the other side, you connect a proper 75 Ohms BNC cable. Then on the second digital gear, you will attach another T connector on the WC IN. One side will receive the BNC cable coming from the master clock, the other one will go to the next digital equipment to sync. Repeat the process until you reach the last equipment. A T connector is attached to the WC IN of this last equipment. On one side, you will connect the BNC cable coming from the previous stuff, on the other side you need to connect a 75 Ohm resistor similar to the one you have on the master clock side.
Be careful to make your BNC cables as short as possible. I do not know the recommendations by heart, but I would not go for more than 3-4 hops in the chain and no more than 4-5 meters of cable altogether.

I suggest you make a search on google to see what a T connector and a 75 Ohms terminator look like. I am sure it will be easier to understand than my lengthy explanation.

Now if you can describe more precisely your setup, I can certainly be more specific and give better advice.

Best,
Thomas

Barkingdogstudios Mon, 01/17/2005 - 13:42

Thomas,

Thank you very much for the suggestions. I did manage to find a piece of equipment by Lucid (the CLKx6) that will at least split the word clock signal if not generate it. It's a lot less money than the equipment listed in option 2/ below (about $300+). I hesitate to buy anything from Aardvark since they've gone out of business.

If I did go for choice 4/ would it be ok to have the next-to-last piece of equipment use it's WC Out to the last piece of equipment's WC In?

I'm really looking for the most "stable" solution for the lowest cost. Would you agree that using the RME (or Aardvark Q10) clock with the Lucid box do what I think (and hope) it will do? Would that solution be in line with the "star" configuration you describe?

I have read in other posts about the issue of cable length. I think I can keep most or all of them under 2m.

Thanks again.

anonymous Mon, 01/17/2005 - 16:48

I have heard good things about the Lucid, although I have not tried it personnaly. In that case, the topology would be quite simple:

Q10 or Multiface as the master WC -> Lucid -> star configuration to all other pieces of equipment. Also do not forget to properly terminate ALL your cables on both sides (T+impedance adaptor). And let us know how it goes...

Best,
Thomas

ghellquist Tue, 01/18/2005 - 09:53

Only one build on Thomas excellent description.

In my mind there should be only 1 (one) termination on each WC cable. That should go after (or at) the last IN. No termination at the OUT end then.

(I believe the termination at the OUT end will not be a large problem, but not having any termination at the last IN will be a problem. Problems in word-clock are elusive, because they may turn up as only an increase in jitter, and it might be only at certain times, say certain times of the day).

Gunnar.

anonymous Tue, 01/18/2005 - 17:21

In my mind there should be only 1 (one) termination on each WC cable. That should go after (or at) the last IN. No termination at the OUT end then.

(I believe the termination at the OUT end will not be a large problem, but not having any termination at the last IN will be a problem. Problems in word-clock are elusive, because they may turn up as only an increase in jitter, and it might be only at certain times, say certain times of the day).

Just to be crystal clear: if you go for the star topology and NOT the daisy chain approach, each BNC cable will be terminated on BOTH ENDS with a T adapter and the 75 Ohms resistor. You do not have to terminate the WC Out of your slaved equipement.

Do not hesitate to ask again if I did not make myself clear!

Best,
Thomas

ghellquist Wed, 01/19/2005 - 00:03

Thomas,
to make things crystal clear. What you are suggesting is that each WC cable should have 2 (two) terminations, one at each end.

I´m firmly convinced that there should be only 1 (one) termination on each cable, after (or at) the last receiving unit. Absolutely no termination, and no T connector at the sender (the one whos out connector you are using).

Could you convince me that I should reconsider this by pointing to some reference I could read. (I did a small bit of radio work ages ago, so I can read that kind of literature. Not sure I understand it though).

You are right in that no termination is used on unused OUT connectors.

Gunnar

anonymous Wed, 01/19/2005 - 04:12

Of course you are right Gunnar, I must have been on something. That's even more stupid of me that I have Big Ben just under my eyes (which has detectors of bad cable terminitions) and all my digital stuff is connected to it with only one termination on each cable (the receiving end).

One should not post too late in the evening. Thanks for pointing that out and my apologizes to everyone reading for the confusion... :oops:

Best,
Thomas

anonymous Fri, 01/21/2005 - 11:26

Word Clock Madness Part Deaux.

So,
I've got a pile of digital gear. Some of witch has a clock in or out via BNC, some of witch does not. How on earth do you sync them all together? I'm very familiar with a master clock system, star layout, 75ohm vs 110 ohm, etc. The gear with BNC's or any style of clock connector sync's fine to a master clock. My problem is this:
Mac G5 with toslink I/O, Presonus Digimax LT with ADAT Optical I/O, Digi 002R with SPDIF I/O and ADAT Optical I/O, A panasonic 3800 with AES/EBU I/O, a Sony Super Bit Mapping CD burner with all sorts of I/O and a sony minidisk with TOSLINK optical I/O and SPDIF I/O.... NONE of the above have word clock in OR out. They all sync via their various digital in's and outs.
Do I need to connect them all together via their various digital I/O's and set them all to sinc via those said digital I/O's EXTERNALLY and make one unit the master?
alternatively, I'm checking out the Drawmer M-Clock. It seems to be just the ticket. In the absence of a Drawmer M-Clock type of product, is there anything that can be done?
Am I stuck switching sync sources and sync master/slave settings everytime I get one digital device to talk to another? I fear this is my situation.
Thanks in advance for any and all advice. I do have a black burst generator, but, I dont think that thing gets me any closer. I still cant get sync INTO the stuff.
Tekbench.