Obi202 interference/static on PH1/PH2

[BGood]

I am trying to diagnose the source of loud audio distortion/interference which recently seems to be originating from my OBI202.  The interference sounds somewhat like an AC hum, but also like the induced static you would hear from brushes on a fan motor.  I will try to attach an mp3 audio recording of dialtone and interference I am hearing from SP1 on PH1 and SP2 on PH2, along with the .jpg waveform illustrating both audio samples, and would appreciate suggestions for eliminating the interference.  As illustrated by the accompanying audio and waveform samples, the interference is much more pronounced on the Google Voice SP2/PH2 channel, rendering this channel pretty much unusable.

System configuration: 

•   Obi202 with 2 DID lines and 2 providers (Anveo on SP1 and Google Voice on SP2)
•   House wiring is disconnected from public TelCo at the network interface
•   PH1 connected through house wiring to V-Tech Dect 6.0 wireless phone system
•   PH2 connected through house wiring to RCA analog 2-line wall phone

So far, I have done the following diagnostic steps with no perceptible reduction in interference:

•   Disconnected the V-Tech wireless phone system to test if it was the source of the interference
•   All of the following steps were performed after disconnecting the in-house telephone wiring from the Obi202 and testing with an Ideal Belt Set probe connected directly to the Obi202
•   Found no obvious local sources of AC interference from the surge protection, cable modem, router, and switch sharing the Obi202 AC outlet
•   Reset the Obi202 several times and changed routing/plug orientation of the AC adapter
•   Swapped the Obi202 AC adapter with the AC adapter from another old Obi202

When testing with the Ideal probe connected directly to the Obi202, I noticed that reducing the probe's monitor volume would eliminate most of the interference.  Unfortunately, lowering the volume on the endpoint phones simply makes the audio inaudible.

Following up on the signal improvement observed when the probe monitor volume was reduced, I wonder if the audio output of the Obi202 might be internally overdriving some element of its internal PH1/PH2 channel or the externally-connected phones?

In the Obi202's ObiExpert configuration, I see that PH1 and PH2 are configured with ChannelTxGain of -5 and ChannelRxGain of -4.  Would modifying either of these settings impact the interference I am hearing?  Looking at the waveform, I notice that the amplitude of the Google Voice SP2/PH2 channel (where the interference is worst) is much higher than on the SP1/PH1 channel.

Thanks in advance for any suggestions.

[drgeoff]

I've looked at the spectrum of your audio file.

The "GV dial tone" has 350Hz -34.2dB, 439Hz at -20.7 dB, 2336 Hz at -32.4 dB as the 3 strongest components.

The "Anveo dial tone" has 350Hz -37.4dB, 440Hz -22.1dB, 2336Hz -51.6dB as the 3 strongest components.

The default OBi dial tone script is programmed to produce 350Hz -18dB, 440Hz -18dB.

Dial tone is generated within the OBi itself.  It does NOT come from the ITSP.  If you have the same gain settings for PH1 and PH2 and the same tone selections you should be getting identical dial tone on both ports.  Of course you need something to take a phone port off-hook to make the OBi send dial tone to the port so if using different phones to go off-hook you could expect differences.

You don't need a network connection for dial tone.  Have you tried unplugging the ethernet cable and listening to the dial tones?

[BGood]

Hi drgeoff,

In response to your suggestion, I unplugged the ethernet cable and listened to the dialtone on PH1 and PH2 from my hardwired desk phone.  I have not made a new recording yet, or tried to listen through the Ideal probe, but the interference and amplitudes of the two channels sound identical to those I originally posted.

Would it be helpful to your analysis if I made a new recording of the two channels with the ethernet cable disconnected?

Thanks for your response and suggestion.

[BGood]

drgeoff:

Following up on your spectral analysis, is the 2336Hz component the suspected interference?  If the culprit was introduced via the AC supply, would wrapping the power supply line around a toroid filter help?  Or am I grasping at straws?