ProSoundWeb Community

Please login or register.

Login with username, password and session length
Advanced search  

Pages: [1]   Go Down

Author Topic: X32 buzz with lost clock - UPDATED  (Read 1302 times)

Brian Jojade

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 3391
    • HappyMac Digital Electronics
X32 buzz with lost clock - UPDATED
« on: July 20, 2021, 03:35:25 PM »

With a setup I've done many times before, I've run into a strange scenario that is a bit perplexing to me.

The configuration is X32 at FOH, shielded ethercon run to amp rack left side into an X32 rack, then shielded ethercon running to right side x32 rack.

Everything works great, and I can tap the outputs as needed.

However, if I power down the FOH mixer, the RIGHT side set of speakers (at the end of the chain) buzzes.  Sounds like a ground loop issue.  But, the ONLY feed into the system is AES50.

When I power up the system, I power up the FOH rack first.  Then I power up the amp racks together.  When the amps turn on, you'll get a buzz for about 7 seconds, then a pause of silence, then buzz for 3-4 seconds, then life is good.

This is only happening on the 2nd rack at the end of the line.  World clock comes from FOH.

It's not a huge deal because when everything is powered up, it works fine.  I'm just perplexed at why the system would BUZZ when the AES clock isn't there.
« Last Edit: July 26, 2021, 01:18:23 PM by Brian Jojade »
Logged
Brian Jojade

Bob Faulkner

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1497
  • Raleigh, NC
Re: X32 buzz with lost clock
« Reply #1 on: July 21, 2021, 06:10:46 PM »

You referenced the setup has been done many times before.  For the issue you are having, Is the setup in the same place, or a different place?

Logged

Brian Jojade

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 3391
    • HappyMac Digital Electronics
Re: X32 buzz with lost clock
« Reply #2 on: July 22, 2021, 07:55:11 AM »

You referenced the setup has been done many times before.  For the issue you are having, Is the setup in the same place, or a different place?

Same venue, same location. Same wires.

Only difference I can think of is firmware is version 4 on the racks.
Logged
Brian Jojade

Kevin Maxwell

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1792
  • USA SW CT 46miles from MidTown Manhattan ATCF
Re: X32 buzz with lost clock
« Reply #3 on: July 22, 2021, 08:53:36 AM »

Are you familiar with the old joke – a man goes to the doctor and swinging his arm around wildly he says it hurts when I do this and the doctor replies then DON’T DO THAT.

This is a situation where you shouldn’t be doing what you are doing. DON’T DO THAT. The amps to the speakers or powered speakers should ONLY be turned on after everything else is turned on and completed its startup sequence. Then you can turn on the speakers and you also must you do the opposite on system turn off. Shutdown the amps or powered speakers first giving them time to turn off completely. As we used to call it let the Caps (capacitors in the power supply) discharge but with the modern amps the Caps don’t have that kind of reserve power in them they are no longer that big. We used to shut down the amps and listen to them slowly discharge with some sound going thru them. Then we knew they were off. But with the DSP built into many of the powered speakers that old trick is probably not a good idea anymore. 

The noise that digital devices will make when losing the clock signal can be annoying or so bad as to potentially be destructively loud. It isn’t intended to be run without the clock, so don’t do that. 

How some systems are designed is not always clear. One would think that they would mute all sound until everything is up and running and some devices do that. But maybe they realize that if you follow the proper turn on and shutdown sequence then you will never have this problem. So why make the devices more costly when if used properly it isn’t a problem. Also this is why in an install I always push for a properly configured power sequencing system. They aren’t always willing to spend the money and rely on the users to know how to turn things on and off in the right sequence.
Logged

Dan Mortensen

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 1072
Re: X32 buzz with lost clock
« Reply #4 on: July 22, 2021, 02:32:03 PM »

Are you familiar with the old joke – a man goes to the doctor and swinging his arm around wildly he says it hurts when I do this and the doctor replies then DON’T DO THAT.


This, and everything else Kevin wrote.

Weird things happen when devices are used incorrectly, which is why they should always be used correctly.
Logged

Russell Ault

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 2499
  • Edmonton, AB
Re: X32 buzz with lost clock
« Reply #5 on: July 22, 2021, 05:58:47 PM »

{...} The noise that digital devices will make when losing the clock signal can be annoying or so bad as to potentially be destructively loud. {...}

I think part of the challenge here is that we've gotten so accustomed to network-based audio transports that "fail to silent" that it's easy to assume that all digital transports do that. In protocols like AES67, each device is responsible for generating its own clock, so a network failure will only interrupt the signal stream while device's clock (and therefore its audio engine) will keep on ticking away in silence. Conversely, devices connected via AES50 all share a single clock, so separating a device from its AES50 clock master will effectively crash its audio engine.

{...} Sounds like a ground loop issue.  But, the ONLY feed into the system is AES50. {...}

FWIW, AES50 (or, rather, the X/M32 iteration, with its shielded cabling requirement) is perfectly capable of causing ground loops.

{...} When I power up the system, I power up the FOH rack first.  Then I power up the amp racks together.  When the amps turn on, you'll get a buzz for about 7 seconds, then a pause of silence, then buzz for 3-4 seconds, then life is good. {...}

Here's my WAG: given the above quote from Kevin, I'd be surprised if the X32 didn't feature some mechanism to mute the audio outputs when the audio engine loses its clock. Building on this, I'm wondering if this muting mechanism does weird things to the impedance present at the outputs or, more specifically, does different weird things to the impedance of the "hot" and "cold" signal lines on the outputs. This could break down the balanced/differential nature of those outputs and drastically reduce the connections' CMRRs, which would explain why the buzzing only happens when the X32 Rack isn't clocking properly.

Given that the buzzing rack is separated from the clock by an intermediary, turning both racks on at the same time means that the Right-side X32 Rack won't have access to a stable clock until after the Left-side X32 Rack has figured its own clocking out, and this is probably what you're hearing on startup. As others have mentioned, the only "correct" way to avoid this is to make sure that each X32 is stable and happy before switching on its associated amps (and vice-versa).

-Russ
Logged

Brian Jojade

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 3391
    • HappyMac Digital Electronics
Re: X32 buzz with lost clock
« Reply #6 on: July 22, 2021, 07:56:24 PM »

Yes, I know, don't do that. I get it. I'm very aware of turning things on and off in the proper order.  The reason I found this issue though, is we lost power FOH and that's when the buzz appeared.  I didn't have time for extensive troubleshooting to figure out why it happened.

But, what's very strange to me is that it's only the ONE unit in ONE position that's buzzing with no clock.  The other X32 at at the stage didn't buzz when clock was lost.  It wasn't the X32, because they ended up getting swapped out and the replacements had the very same behavior.

Hopefully next week I'll have time to rebuild the rig in the shop and test the scenario to see if I can repeat the issue.  The only thing I can think of right now is that something in firmware 4 does something strange when you pass AES clock through 2 devices causing the one on the end of the chain to get funky.  This is the first time I've used firmware 4 in that configuration.  Never an issue with devices connected with a single cable to the master clock.
Logged
Brian Jojade

Brian Jojade

  • Hero Member
  • *****
  • Offline Offline
  • Posts: 3391
    • HappyMac Digital Electronics
Re: X32 buzz with lost clock - UPDATED
« Reply #7 on: July 26, 2021, 01:53:18 PM »

Update.

Got to the shop, did a test run. Yes, same behavior shows up on the problem X32 position.  No clock = noisy signal.  Did a factory reset of the unit and reloaded the scene file, now it seems to work as expected.  Very strange. Apparently something in the software update process confused the unit.  It's weird that it happened to 2 units identically, but the other unit didn't have the problem.  In any case, there's something to be said about doing a factory reset after a firmware update, I guess.
Logged
Brian Jojade

ProSoundWeb Community

Re: X32 buzz with lost clock - UPDATED
« Reply #7 on: July 26, 2021, 01:53:18 PM »


Pages: [1]   Go Up
 



Site Hosted By Ashdown Technologies, Inc.

Page created in 0.031 seconds with 22 queries.