Hissing/buzzing noise on DT700 IP phones on SV8500

[phadobas]

I've been getting some reports from some DT700 users that they have noticable background hissing noise on their phones. The persons who started complaining say it's a change and it hasn't always been like this.

What I've got is a quiet office, making a test call to any outside party (even if it's just across a CCIS link to our old NEC), and I get a completely usual phone call. But when I have the other person mute his phone and I turn my receiver volume all the way up, some background hissing noise starts kicking in. If I breath into the mouthpiece (not blow, just breath), this hissing intensifies, and may stay intensified even if I stop breathing. On top of that, there is a definite, constant, steady buzz on the line under these circumstances. I would have never noticed it in normal conversations, only under these extreme conditions.

While this may sound paranoid, I believe this buzz contribute to the generation of the hissing sound.

Then when I make a long distance call, I definitely hear the hiss in the background, even if I don't turn up my receiver's volume. I've gotten used to it, but now I start questioning its validity.

From what I've been able to re-produce, IP-to-IP phones sound clean. It's only IP to non-IP calls. I started suspecting the IP-PAD cards, but I have one of those MGC PRI boxes (the 1U rackmount thing with a PRI blade), and made a long distance test call via it. It bypasses the PAD cards since the conversion from IP to non-IP happens inside THIS box, and I still heard the buzz. Steady, constant buzz. Only when all parties are quiet, and I turn my volume up high. It's definitely there.

I'm at loss as to where to even start debugging this.

It may have always been there, and nobody really cared to report (and I never noticed). The only reason I even started experimenting with this is because of the original report of the presence of hissing sound.

 

[3racefan3]

Login into the phones and in the maint. option or somewhere ( i cant remember where exactly right now) you will see an option for adjust, change sidetone noise to 0 from 4 and enable noise cancellation.

 

[phadobas]

I've tried noise suppression before making this post, to no much improvement. I should have put that in my original post. I've now also tried the side-tone change, but still didn't get improvement.

The problem became worse in the last 2 weeks or so. I've looked at the grounding of the network switches, even the grounding of the phone system itself but they are all good.
I've many subnets, many vlans, but I have QOS enabled on the switches as well as turned on in the phone system (ALOCL + AIVCL)

 

[phadobas]

Saga continues. This buzz is like when you have an old analog phone and you move too close to a fluorescent light. Obviously it's very mild, but it's there if you turn up your volume on the phone.

I've checked for electrical noise sources in and around the switch room. Nothing. Checked all grounding cables, but they are all connected securely.
I've used CMOD to switch CPU-s to the sby, and also switched the TDSW, but the buzz remained.

It is true that reducing the side-tone volume to 0, and enabling noise suppression on the DT700 phone reduces this buzz by about 50%, but I'm facing a pbx-wide situation, so I want to find out where it this noise coming from, instead of just suppressing it on the phones.

The only other time-coincident change I can think of, is that all phones on all subnets on this system got a LOC ID assigned by ALOCL. And then AIVCL was used to assign G711 with 20ms payload and echo on between all combination of locations.

What I've just noticed in this setup is that per the manual, the telephony server is represented by 0.0.0.0 in ALOCL, and gets a LOC ID=0. Yet, at the same time, the person who did this assignment, took the IP subnet of the LAN 1 interface and assigned LOC ID=2. Shouldn't that also be 0?

And any other ideas?

 

[phadobas]

So I've made some more tests.

I've gotten an electrical company check all the grounding and they didn't find any problem. I even had all grounding completely disconnected from the system and the buzz remained unchanged.

So the buzz doesn't occur between IP stations.

It occurs when the PAD cards are involved. BUT! Not always. If I call from an IP phone to an analog phone within the same system, there is NO buzz even though the PAD cards are involved. The buzz only occurs, when a PAD card is involved AND a trunk card is also involved in the call.
I've confirmed the buzz on all calls that touched an incoming or outgoing trunk (route). They are all some sort of PRI connections. I also have a CCIS to a 2400ICS system, and the buzz occurs there too.

So it somehow appears to be related to an IP phone-PAD card-PRT card speech path.

Does this mean anything to anybody?

 

[belevedere]

Do you have redundant power supplies? If so, switch over and remove (unplug) the off line PSU. Your buzz may be a PSU breaking down.

 

[phadobas]

Thank you for the advice, belevedere.

I have dual power supplies on all 3 shelves. So I tried removing them all one-by-one, and unfortunately it didn't make a difference.

Only now one of the power supplies didn't came back up

 

[phadobas]

I've been doing some research about the reported phenomenon, which is the background hissing noise. It almost sounds like white noise. I've been targeting the buzz as the source of it, but for all I know, we always had this buzz...

So now onto hissing noise. An article on cisco website talks about VAD (voice activity detection), which generates comfort noise for a user, so during moments of silence, you don't think you got disconnected. This seems like what's happening here, but this comfort noise is being too much. I've never heard of the equivalent of this in NEC. Is there such a thing? Somehow I need to turn this down.

 

[phadobas]

It appears to have been resolved.
In command AIPDL, under Advanced, there is a PAD setting for tie lines for R_PAD. I've changed that from 0 to 10 (total arbitrary), and it completely eliminated the buzz and the hissing noise.

I went through the MAT logs for the last 18 months and found nothing that would have created this problem. In fact, somewhere in Jan 2014, I've looked at these settings and they were the same as yesterday. So not sure what caused it to go hay-wire, but the solution lies in the APDIL command.

 

[phadobas]

Well, looks like I have to "re-open" this thread. The solution I described earlier brings about an un-desired reduction in volume.

The new information I can add is this:
When a call to/from a DT700 phone is happening, often when the 2 party talks at the same time, the phone system can't figure out who is louder (I think) so the two voices collide and result in digitized garbled mess.

This is still on calls that involve the IP PAD cards AND some CCIS or PRI trunk card.
On long distance calls, when the far is tends to not speak loud and clear, my server room's background noise is liable to completely suppress what the other side is saying and I can't hear anything of what he says.

So far anything I've tried with APDIL, (to increase volume of their speech or my speech) just didn't improve the problem or made it worse. Increasing my volume suppresses theirs even more, and increasing theirs also increases a serious background hissing noise which I think could be called "comfort-noise" but it is un-naturally strong.

So I just don't know if my APDIL settings are completely out of whack, or what?

On CCIS and PRI routes, the ARTD CDN30 is set to 4, per the manual. Although I tried to set them to 7 (a tech recommended), it didn't improve.

Anybody has a good working system? Can you share ARTD for your PRI-s?
And the APDIL settings for IP PAD and IP stations (they are both under the Advanced tabb).

 

[phadobas]

And maybe the switch settings on the PRI card and on the IP PAD cards (DPSW command)?