• From: Frankfurt, Germany
  • Registered: 2011-03-23
  • Posts: 56

Topic: Can different wordclock cable lentghs cause sync issues?

I am trying to find the cause for frequent crackles. Our setup: 3  HDSPe cards running 3 Multiface II, on each Multiface II there is one Octamic II. I want to rule out a wordclock cable length issue.

Sync is done via wordclock. One of the Multiface II is master, the other 5 digital devices are set to slave / external clock. The wordclock BNC cables between the 1st Multiface II (clock master) and the 2nd Multiface II and 2 of the octamics are short (< 30cm), but the bnc connection to the 3rd Multiface II and the 3rd Octamic II is 3 meters. The bnc connections are properly terminated.

As soon as we use any input or output on the 3rd Multiface II parallel to using inputs/outputs from one of the other 2 Multifaces, crackling starts (we use Cubase 5 / 6). Crackling is independent of cpu usage. Crackling happens even on test projects with very little cpu usage.

Is it possible that the wordclock cable length to the 3rd Multiface II can cause this? If this is an issue, is there any guideline about wordclock cable length in a digital setup?

Thank you in advance for input on this.

2 DAWs, each C 7/64, W7/64, Dell Prec. T5500 (2 x X5690), 2 x rme HDSPe / Multiface II / Octamic II
  • From: Houston, TX
  • Registered: 2007-05-11
  • Posts: 1,198

Re: Can different wordclock cable lentghs cause sync issues?

Does the Sync/Lock status fluctuate in each Multiface's DSP settings box?  All should show "Sync" for ADAT and Wordclock connections on all 3 Multifaces (the master Multiface obviously won't have a Wordclock Sync since it's the system WC master)

How are the BNC's routed?  Tee's or In/Out loops?  Some people are adamantly against the use of "Tee's" and a proper end-chain terminator - but I rely on Tee's w/o issues with almost as many pieces sync'd off a common clock output.  If you are using Tee's - are you sure you aren't double or triple terminated somewhere?

I also have various length BNC cables in my setup ranging from 0.5 Meter (18") to 9 Meters (25') - my gear stays sync'd just fine.  You can also try to temporarily sync over a mix of SPDIF Coax and ADAT Optical to rule out your BNC cables and termination as a possible culprit (SPDIF Coax between Multifaces, and ADAT out from MF to the Octamics)...


cool

MADIface-XT+ARC / 3x HDSP MADI / ADI648
2x SSL Alphalink MADI AX
2x Multiface / 2x Digiface /2x ADI8

3 Reply by MC

  • MC
  • MC
  • Administrator
  • Offline
  • From: Germany/Thailand
  • Registered: 2007-04-06
  • Posts: 25,779

Re: Can different wordclock cable lentghs cause sync issues?

To add to Randy's perfect answers: all PCI cards are flashed to the same firmware version?

Regards
Matthias Carstens
RME

MC's Website

  • From: Frankfurt, Germany
  • Registered: 2011-03-23
  • Posts: 56

Re: Can different wordclock cable lentghs cause sync issues?

Matthias: all 3 PCIe Cards are flashed to v18.

Randy:
- BNC is routed with T-ee's. At both ends of the BNC chain there is an Octamic with termination switched on. I checked for termination errors a million times...
- The DSP status display shows proper "sync" for ADAT and Wordclock (obviously no Wordclock sync for the master Multiface) and this status display does ot change even if I have a shower of crackles going on.
- Thank you for your report about your cable lenght, good to know this can work.

2 DAWs, each C 7/64, W7/64, Dell Prec. T5500 (2 x X5690), 2 x rme HDSPe / Multiface II / Octamic II
  • From: Houston, TX
  • Registered: 2007-05-11
  • Posts: 1,198

Re: Can different wordclock cable lentghs cause sync issues?

At both ends of the BNC chain there is an Octamic with termination switched on

Can you elaborate on this setup and your BNC routing through each device?  Since you mentioned that your "Sync" status is solid across all 3 Muiltifaces - it would seem that your clocking is correct - but your description above sounds odd to me if the MF #1 is your system WC master.

Are all 3 MF's in the same PC?  Are you getting any DPC Latency spikes when utilizing the 3rd card's I/O (or any combo of I/O)?  Do you have any other PCI devices stepping on the shared PCI bus?

And lastly, is this crackling audible if monitored through Totalmix into Headphones (or whatever) to eliminate ASIO and the PC itself as an issue?

Trying the combo of SPDIF/ADAT clocking is also an option as mentioned above to rule out your BNC chain.


PS - Woo-hoo - I got props from MC.  I feel "big time" now! :-P

cool

MADIface-XT+ARC / 3x HDSP MADI / ADI648
2x SSL Alphalink MADI AX
2x Multiface / 2x Digiface /2x ADI8
  • From: Frankfurt, Germany
  • Registered: 2011-03-23
  • Posts: 56

Re: Can different wordclock cable lentghs cause sync issues?

Matthias,

apparently, the problem was caused by the rme drivers.

I read the read-me of the new beta driver and noted "optimized call-backs for multiple cards on multi-processor machines". I have a single-cpu multi-core system with 3 cards so I gave it a try.

It looks like the problem went away with the install of the new beta drivers. I just tested the system with 3 cards, channels on all 3 cards in use, 32 samples buffer size, with no clicks.

Thankx.

Regards, Mikael

2 DAWs, each C 7/64, W7/64, Dell Prec. T5500 (2 x X5690), 2 x rme HDSPe / Multiface II / Octamic II

7 Reply by MC

  • MC
  • MC
  • Administrator
  • Offline
  • From: Germany/Thailand
  • Registered: 2007-04-06
  • Posts: 25,779

Re: Can different wordclock cable lentghs cause sync issues?

Well, that is - in the end - good news!

Regards
Matthias Carstens
RME

MC's Website