[asterisk-users] TDM400P won't ring GM phone of mere 0.1B

[Yuan LIU]

>From: John Novack <jnovack at stromberg-carlson.org>
>
>yliu11 at hotmail.com wrote:
>>This is rather bizarre: My TDM11 (one FXS) rings a $10 passive phone with 
>>REN of 1.0B, a cheap speaker phone of 0.3B, and a cordless phone with 
>>marked REN of 0.0B.  But it couldn't properly ring this 27935GE3-B (FCC ID 
>>G9H2-7930) cordless phone rated at merely 0.1B.  Rarely, the phone will 
>>crack out an occasional weak and abrupt beap, but never a normal ring.  
>>Otherwise Asterisk and TDM400P works with this phone, dialing, voice, 
>>callerID and all that. (In fact, when the FXS "rings", the display lights 
>>up as it should.)
>>
>>This is a dual-line 2.4 GHz cordless phone.  It rings normally when 
>>plugged into wall jack.
>>
>>Any idea?  Though not described in the manual specifications, Table B1 
>>suggests that an FXS module (S110M?) supports at least REN 3.
>>
>>Yuan Liu_______________________________________________
>>
>Have you tried toe boost ring voltage option then recompile Zaptel?
>It is normally set to a fairly low voltage
>
>John Novack

Thank you so much!  I googled a bit about how to change ring voltage and 
only found an old and suspended feature request from last year that 
concerned wcfxs.c, which is now superceded by wctdm.c.  Yet the same method 
applies.  So I changed the value of RING_X from 0x0160 to 0x023A as 
suggested for European countries (peak of 85V according to that document) 
but left frequency (RING_OSC) unchanged, like the following:

--- wctdm.c.1.2.10      2006-07-07 11:02:39.000000000 -0700
+++ wctdm.c     2006-12-13 22:04:28.862053256 -0800
@@ -81,7 +81,8 @@
{18,5,"OSC2Y",0x0000},
{19,6,"RING_V_OFF",0x0000},
{20,7,"RING_OSC",0x7EF0},
-{21,8,"RING_X",0x0160},
+// {21,8,"RING_X",0x0160},
+{21,8,"RING_X",0x023A},                        // ring voltage set higher
{22,9,"RING_Y",0x0000},
{23,255,"PULSE_ENVEL",0x2000},
{24,255,"PULSE_X",0x2000},

Now the GM phone actually rings, though still a little strangely.  Guess I 
just have to experiment a little to completely address this - but that 
feature request 0004542 should really be revived.

A configuration string "boostringer" was mentioned in several messages, 
including one concerning TDM400P, all without indicating the applicable 
configuration file.  This has no apparent effect on TDM400P wherever I 
tried.

BTW, I made some interesting tests - I'm relatively new to this, so bare 
with my learning curve.

Without an oscilloscope, I used a DT-830B multimeter to test three "lines": 
the Digium FXS, a Linksys WRTP54G FXS (yes, that's Vonage), and a PSTN land 
line (SBC).  This meter has a strange behaviour: it allows DC voltage to 
pass through when in AC mode if polarity is favourable.  Guess it's too 
cheap to contain a decent capacitor.  But this gives me an opportunity to 
observe the difference between SLIC in a CO and one in a home appliance.

With Digium and Linksys, if I change polarity of the meter to allow DC 
passthrough, idle AC would appear to be 108V, which drops to 60V during ring 
phase; but with SBC, the apparent "AC" would appear boosted to 161V.  Is 
this because SLIC in a home appliance outputs square waves instead of sine 
waves?

Linksys' AC output (without DC passthrough) measures about 61V, about the 
same as SBC, while Digium's measured about 45V.  Now Digium also measures 
about 61V, but there's still some strangeness.  Maybe an oscilloscope is 
really needed:-(