keywords: ip pbx voip gateway gsm gateway

×

Notice

The forum is in read only mode.
  1. Forum
  3. End of Life Products Support
  5. D115 Series
  7. Remote hybrid echo not being canceled by hardware ec
× Questions about D-110P 1 Span E1/T1/J1 Interface Card

Remote hybrid echo not being canceled by hardware ec

13 years 2 months ago #6543 by lordvadr
Greetings,

I’m a seasoned asterisk administrator but I’m new to OpenVox and the T1/E1/J1 PRI cards.
I have two installations, one with a Digium TE121B, one with an OpenVox D115E, both with their relative hardware echo canceller modules.
We do not appear (have had no complaints) to have abnormal echo problems with the Digium card.

When this problem first cropped up, we tried various things.
One evening, I unloaded all the DAHDI modules and reloaded them with the MG2 software echo canceller disabled (unneeded and not used anyway) which, at that time, appeared to fix the problem.
It now appears intermittent.
I have no idea if this actually did anything useful.

The phone numbers we were reliably able to reproduce and now can intermittently reproduce the echo on are all POTS lines in relatively remote places, so it’s clear that this is far end hybrid echo—something the echo-can should be able to take care of.
I can hear the transmitted audio in the received audio stream.
Its amplitude is quite a bit lower than the transmit channel audio.
By the time the delay inherent in VoIP is added, it is a noticeable echo and annoying.

I have properly gained the card using telco milliwatt stations.

1) I can’t get any pre-echo-canceller audio with dahdi_monitor (-R/-T/-S options) from the OpenVox card, while I can with the Digium card.
Is this a known bug or is there something I’m doing wrong.
I am converting the audio dumps with ‘sox -r 8000 -b -c 1 -t raw -U rx.raw rx.wav’.
All I get is a single tone, and it’s different for each dump.
This all works with the Digium card.
What I’m attempting to do is determine the relative amplitude of the received echo before and after the echo canceller.

2) What is the best way to go about trouble shooting the echo canceller?
/proc/dahdi/1 shows the echo canceller on the channels, but at best, it’s not working well enough.
I can’t say whether it’s working at all or not because I am unable to get at the pre-ec audio.

Any help you can give us would be greatly appreciated.
I have followed the information in every last how-to I can find, including Digium’s and voip-info’s.
We are at a loss and will decommission these cards if we cannot get them to work reliably.
I understand echo is a difficult problem, which is why we spent the extra money on the echo cancellers.

We are running asterisk 1.6.2.11, dahdi 2.2.0/2.2.0.
Application is a T1 with a 23 channel ISDN pri.

Please don’t suggest upgrading unless there’s a known issue.
2.4.0 didn’t want to compile with either 1.6.2.11 or 1.8.anything.
We are running the only dsp firmware available on the website.
This problem should be taken care of in hardware since from my view-point, it’s end-to-end digital.
The system is in production and I can’t afford to take it down for hours to upgrade on a hunch.

One thing I do notice is the dahdi_voicebus module that speaks of the digium echo canceller.
Should this be removed?

Thank you again for your time and I look forward to a response.
13 years 2 months ago #6544 by Jessie.liao
HI,
The MG2 software echo canceller must be disabled.
Please show your dmesg information.
What's the output of "cat /proc/interrupts"?
If possible,could you send your SSH to my email:[email protected]? I'll check that for you.
Please add my MSN/G-talk:[email protected] to contact with me.
13 years 2 months ago #6545 by james.zhu
hello:
you can refer the read me for debug the card:
debug (most modules)::
Sets debug mode / debug level. With most modules 'debug' can be either
disabled (0, the default value) or enabled (any other value).
+
+
wctdm and wcte1xp print several extra debugging messages if the value
of debug is more than 1.
+
+
Some modules have "debugging flags" bits - the value of debug is a
bitmask and several messages are printed if some bits are set:
- ztdummy:
* 1: DEBUG_GENERAL - general error messages.
* 2: DEBUG_TICKS - Show that the module is alive :-)
- wctdm24xxp:
* 1: DEBUG_CARD
* 2: DEBUG_ECHOCAN
- wct4xxp:
* 1: DEBUG_MAIN
* 2: DEBUG_DTMF
* 4: DEBUG_REGS
* 8: DEBUG_TSI
* 16: DEBUG_ECHOCAN
* 32: DEBUG_RBS
* 64: DEBUG_FRAMER
- xpp: Previously (before 1.2.26 / 1.4.11) it was called "print_dbg".
See also README.Astribank:
* 1: GENERAL - General debug comments.
* 2: PCM - PCM-related messages. Tend to flood logs.
* 4: LEDS - Anything related to the LEDs status control. The driver
produces a lot of messages when the option is enabled.
* 8: SYNC - Synchronization related messages.
* 16: SIGNAL - Zaptel signalling related messages.
* 32: PROC - Messages related to the procfs interface.
* 64: REGS - Reading and writing to chip registers. Tends to flood
logs.
* 128: DEVICES - Device instantiation, destruction and such.
* 256 - COMMANDS - Protocol commands. Tends to flood logs.
==================================================
when loading the driver by: modprobe wct4xxp debug=XXXX
13 years 2 months ago #6546 by lordvadr
Jessie,

I have added you to my msn contact list. I'll keep an eye out for you. I'm going to guess you're in UTC+8. I'm in UTC-6 so it might be tough to get a hold of each other. Unfortunately I can't give you the ssh login just yet, but I have a request into mgmt.

I will look into enabling debug on the card this evening. Here are the answers to your questions:

MG2 is disabled.
Here is the dmesg from the most recent module load. I can see the earlier load up in the file that references the MG2 ec, but that has been disabled...
Stopped D115P/D115E, Turned off DMA
D115P/D115E: Disabling interrupts since there are no active spans
Found D115P/D115E at base address fc4ffc00, remapped to ffffc20000014c00
D115P/D115E version c01a0000, burst OFF
FALC version: 00000005, Board ID: 00
Reg 0: 0x663b0400
Reg 1: 0x663b0000
Reg 2: 0xfc4ffc08
Reg 3: 0x00000000
Reg 4: 0x00000000
Reg 5: 0xfc4ffc14
Reg 6: 0xc01a0000
Reg 7: 0x00001000
Reg 8: 0x00000000
Reg 9: 0x00000080
Reg 10: 0xfc4ffc28
Found a Card: OpenVox D115P/D115E
D115P/D115E: Launching card: 0
D115P/D115E: Setting up global serial parameters
Stopped D115P/D115E, Turned off DMA
Found D115P/D115E at base address fc4ffc00, remapped to ffffc20000014c00
D115P/D115E version c01a0000, burst OFF
FALC version: 00000005, Board ID: 00
Reg 0: 0x677c4400
Reg 1: 0x677c4000
Reg 2: 0xfc4ffc08
Reg 3: 0x00000000
Reg 4: 0x00000000
Reg 5: 0xfc4ffc14
Reg 6: 0xc01a0000
Reg 7: 0x00001000
Reg 8: 0x00000000
Reg 9: 0x00000080
Reg 10: 0xfc4ffc28
Found a Card: OpenVox D115P/D115E
D115P/D115E: Launching card: 0
D115P/D115E: Setting up global serial parameters
About to enter spanconfig!
Done with spanconfig!
About to enter startup!
Not prepped yet!
D115P/D115E: Span 1 configured for ESF/B8ZS
Not prepped yet!
Not prepped yet!
Not prepped yet!
Opvxd115: Setting yellow alarm on span 1
SPAN 1: Primary Sync Source
timing source auto card 0!
VPM400: Not Present
timing source auto card 0!

VPM450: echo cancellation for 32 channels
Opvxd115: Clearing yellow alarm on span 1
VPM450: hardware DTMF disabled.
VPM450: Present and operational servicing 1 span(s)
Completed startup!
device eth1 entered promiscuous mode
device eth1 left promiscuous mode
Opvxd115: Setting yellow alarm on span 1
timing source auto card 0!
timing source auto card 0!
timing source auto card 0!
timing source auto card 0!
timing source auto card 0!
timing source auto card 0!
Opvxd115: Clearing yellow alarm on span 1

/proc/interrupts
CPU0 CPU1 CPU2 CPU3
0: 3907271269 0 0 0 IO-APIC-edge timer
1: 3 0 0 0 IO-APIC-edge i8042
2: 0 0 0 0 XT-PIC cascade
3: 13 0 0 0 IO-APIC-edge serial
4: 375 163 0 0 IO-APIC-edge serial
8: 0 0 0 0 IO-APIC-edge rtc
12: 4 0 0 0 IO-APIC-edge i8042
14: 1664624 71751301 0 0 IO-APIC-edge ide0
98: 1105162 0 193945359 0 PCI-MSI eth1
129: 172898 647860 2649217236 272381848 IO-APIC-level opvxd115
161: 0 0 0 0 IO-APIC-level uhci_hcd:usb3
169: 23 0 0 0 IO-APIC-level ehci_hcd:usb1, uhci_hcd:usb2, uhci_hcd:usb4
201: 289697 12655657 8725 1205 IO-APIC-level megasas
NMI: 138300 43503 175397 43165
LOC: 3723234469 3743187669 3723240859 3743185157
ERR: 0
MIS: 0

Additionally, here is /proc/dahdi/1 during a call:
Span 1: D115P/D115E/0/1 "D115P/D115E (PCI/PCI-E) Card 0 Span 1" (MASTER) B8ZS/ESF ClockSource

1 D115/0/1/1 Clear (In use) (EC: VPM450M)
2 D115/0/1/2 Clear (In use)
3 D115/0/1/3 Clear (In use)
4 D115/0/1/4 Clear (In use)
5 D115/0/1/5 Clear (In use)
6 D115/0/1/6 Clear (In use)
7 D115/0/1/7 Clear (In use)
8 D115/0/1/8 Clear (In use)
9 D115/0/1/9 Clear (In use)
10 D115/0/1/10 Clear (In use)
11 D115/0/1/11 Clear (In use)
12 D115/0/1/12 Clear (In use)
13 D115/0/1/13 Clear (In use)
14 D115/0/1/14 Clear (In use)
15 D115/0/1/15 Clear (In use)
16 D115/0/1/16 Clear (In use)
17 D115/0/1/17 Clear (In use)
18 D115/0/1/18 Clear (In use)
19 D115/0/1/19 Clear (In use)
20 D115/0/1/20 Clear (In use)
21 D115/0/1/21 Clear (In use)
22 D115/0/1/22 Clear (In use)
23 D115/0/1/23 Clear (In use)
24 D115/0/1/24 HDLCFCS (In use)
13 years 2 months ago #6547 by lordvadr
I have more info on this. The echo canceler works, but only on channel 1. I verified this by configuring my other asterisk system to answer a call and immediately run the Echo command. On channel 1, there is no echo (actually there's a cancellation artifact but I'm assuming only because the "echo" is so loud). On every other channel, the echo is there in full amplitude, even though in /proc/dahdi/1, it indicates the operation of the echo canceller.
/etc/dahdi/system.conf:

# Span 1: D115P/D115E/0/1 "D115P/D115E (PCI/PCI-E) Card 0 Span 1" (MASTER)
span=1,1,0,esf,b8zs
bchan=1-23
dchan=24
loadzone = us
defaultzone = us


/etc/asterisk/dahdi-channels.conf:

group=0
txgain=0
rxgain=0
context=from-pstn
switchtype = national
signalling = pri_cpe
channel => 1-23


Anything else you need me to post. Tonight, I'm going to confirm that this behavior happens on all channels besides 1, and that I get echo on channel 1 with echo-can disabled. I will also turn on the echocan debugging and see what happens. If you have any ideas on this though, please let me know.

-CJO-
13 years 2 months ago #6548 by Jessie.liao
HI CJO,
The EC works only when calling.
If your calling via channel 1,the EC status of channel 1 detects ON. And the other no calling channels will be OFF unless you make calls through them.
  1. Forum
  3. End of Life Products Support
  5. D115 Series
  7. Remote hybrid echo not being canceled by hardware ec
Time to create page: 0.047 seconds
Powered by Kunena Forum