[Asterisk-Users] TDM400P Revision question.

[Robert Webb]

On Thu, 14 Apr 2005 10:59:11 -0600
  Rich Adamson  <radamson at routers.com> wrote:
>> >> > > I was following a discussion on this list about 
>>the 
>> >>TDM400P 
>> >> > revisions. 
>> >> > >It is my understanding that the current revision 
>>that 
>> >>one 
>> >> > should have 
>> >> > >is the Rev. H and not the E/F. I have not yet been 
>> >>able to 
>> >> > verify the 
>> >> > >rev stamped on the board, but zaptel is reporting 
>> >>that I 
>> >> > have the Rev. 
>> >> > >E/F. I just bought this card in January direct 
>>from 
>> >>Digium and was 
>> >> > >wondering if I got the wrong Rev.
>> >> > >somehow?? I have been having some intermittent 
>> >>problems but only 
>> >> > >thought it was my setup.
>> >> > > 
>> >> 
>> >> 
>> >> 
>> >> I did some more testing today. I called Digium on 
>>4/12 
>> >>and they
>> >> suggested some things to try, like different 
>> >>motherboard, switching pci
>> >> slots, etc.. I did everything they asked, except for 
>>the 
>> >>mother switch
>> >> as I do not have a different one to put in the system 
>>at 
>> >>this time.
>> >> 
>> >> So, after all that, my ringing issue still persists. 
>>Too 
>> >>some
>> >> measurements from bot the card and my POTS line in 
>>both 
>> >>the on-hook
>> >> state and ringing state. I uses a digital multi-meter 
>>to 
>> >>make the
>> >> measurements on both. Here are the results....
>> >> 
>> >> TDM400P
>> >> 
>> >> Before slot change:
>> >> 
>> >> On hook idle:
>> >> 
>> >> 43.8 Volts DC
>> >> 0    Volts AC
>> >> 
>> >> Ringing:
>> >> 
>> >> 0    Volts DC
>> >> 56.4 Volts AC
>> >> 
>> >> After slot change:
>> >> 
>> >> On hook idle:
>> >> 
>> >> 48.7 Volts DC
>> >> 0    Volts AC
>> >> 
>> >> Ringing:
>> >> 
>> >> 0    Volts DC
>> >> 65.5 Volts AC
>> > 
>> > We can only assume the above represents a fxs module 
>>on 
>> >the
>> > card. Correct?
>> > 
>> > I would find it hard to believe that changing slots 
>> >would cause
>> > the on hook DC voltage to change from 43v to 48v. That 
>> >smells like
>> > a funcky voltmeter. Slots should have nothing to do 
>>with 
>> >DC
>> > voltage unless the module is simply bad. The AC 
>> >(ringing) voltage
>> > is reasonable, but again it should not have changed 
>> >simply
>> > because of a slot change; again, questionable 
>>voltmeter.
>> > 
>> >> On my POTS line:
>> >> 
>> >> On hook idle:
>> >> 
>> >> 43.8 Volts DC
>> >> .013 Volts AC
>> >> 
>> >> Ringing:
>> >> 
>> >> 50.5 Volts DC
>> >> 93.9 Volts AC
>> > 
>> > The on hook DC voltage from all US telco's will 
>> >factually be
>> > in the 48v to 52v range. If their central office 
>> >equipment produced
>> > 43 volts, they would have alarms going off all over 
>>the 
>> >place.
>> > Their alarms would trigger somewhere in the 46 to 48 
>> >volt range.
>> > So, that measurement implies the voltmeter is not 
>> >accurate.
>> > The AC (ringing) voltage is well within acceptable 
>>telco 
>> >limits
>> > and can range from about 70v to upwards of 105v.
>> > 
>> >> Could it bee that from the phone company they retain 
>>the 
>> >>DC offset
>> >> voltage while applying a ring frequency and as it 
>> >>appears on the TDM it
>> >> shuts off the DC offset when ringing starts. Could 
>>this 
>> >>be the issue
>> >> with those of us in the U.S. having ringing issues 
>>with 
>> >>the TDM's??
>> > 
>> > Doubtful that is an issue. The reason for saying that 
>>is 
>> >the chipset
>> > used on the fxo & fxs modules was manufactured by 
>> >Silicon Labs, and
>> > those same chipsets are used in other telephony 
>> >equipment worldwide.
>> > Silicon Labs is known for good to excellent products. 
>>If 
>> >their chipsets
>> > didn't function correctly, there would have been a 
>>large 
>> >uprising a
>> > couple of years ago when those chips were first 
>> >produced. That
>> > hasn't happened, and they don't have a lengthy chip 
>> >revision history.
>> > 
>> > Asterisk code does not have any control over 
>> >adding/removing the DC
>> > component during ringing, so that's not an issue 
>>either. 
>> >Doubtful
>> > that adding/removal the DC component would have any 
>> >impact on 
>> > normal telephone sets, however there certainly could 
>>be 
>> >funcky sets
>> > that don't like that DC removal.
>> > 
>> > Given the number of postings relative to the TDM card 
>> >lately, I don't
>> > remember exactly what your ringing issue was. Could 
>>you 
>> >remind us
>> > without deleting the significant parts of the above?
>> > 
>> > 
>> 
>> Even though it is long, I will leave everything intact.
>> 
>> I have had a few issues with dropped calls when using 
>>the 
>> FXS to FXO connection. Not sure what the issue is with 
>> that. THe main issue I have is with the ringing on the 
>>FXS 
>> card. I have three differnt brands of phones and all 
>>three 
>> do the same thing. I might get two or three calls in 
>>where 
>> everything works fine. But then the next one will cause 
>> intermittent ringing one all phones and no data for 
>>caller 
>> id.
>> 
>> I have tried every combination of the phones I have that 
>> is possible. From only one of each type hooked directly 
>>to 
>> the FXS card to hooking the card to my internal house 
>> wiring and using various combos of the phones connected.
>> 
>> It almost acts like the phones are requiring just a hair 
>> more ring voltage to work properly. That is why I was 
>> testing the voltage levels. I will try and grab a 
>> different meter to test with.
> 
> There is a compile-time option to increase the ring 
>voltage.
> I don't recall the specifics, but its likely in wctdm.c 
>or
> an associated header file. (As you probably can tell, I 
>don't
> use the fxs modules on my TDM card.)
> 


Yes, I do know about the compile time option and it is 
enabled. Well, at least the ZAPTEL driver is saying it is. 
I will reload the driver without the option and see what 
the voltage does.

>> The system is a PIII 933MHZ, VIA chipset and has a 500 
>> watt power supply in it. So I don not think it is a 
>>power 
>> issue from the computer itself.
> 
> That shouldn't be a problem at all.
> 
>> The reason I asked about the DC offset during ringing, 
>>is 
>> that on the telco side, I noticed that the offset 
>>remained 
>> even when ringing voltage was applied. On the TDM, it 
>>does 
>> not. In the manual for the chipset that someone sent me, 
>> there is the option to apply a DC offset voltage during 
>> ringing. Additionally, the telco side gives the 93 Volts 
>> AC when ringing where the TDM is only supplying the 65 
>>AC 
>> as according to my meter. These tests were conducted 
>> several times to get average readings.
> 
> I think you can safely assume that Mark (and/or digium) 
>tested
> the fxs modules with at least a couple of phones, so its 
>unlikely
> there is a software issue such as the offset parameter 
>lurking.
> 
>> As far as the issue with DC voltage on the POTS line 
>>only 
>> being  43.8 DC, my guess was that is just an issue with 
>> voltage drop on the line because of distance between me 
>> and the CO.
> 
> No possible way. If everything is truly on hook, there 
>isn't
> any current draw and therefore no way for a voltage drop 
>to
> occur. Basic ohm's law.
> 


Thank you for the reminder. Been a while since I have 
really used any of this.

>> If I need to answer anything else, please let me know.
>> 
>> 
>> P.S. - I do realize that that the chipset is a good 
>> shipset. I am just wondering if everything is setup 
>> correctly in the zaptel driver for the proper 
>>programming 
>> of the registers in the chipset to make it function 
>>within 
>> specs.
> 
> I don't know about the fxs module, but I know for a fact 
>that
> the fxo module has more capability then what has been 
>implemented
> in the current driver.
> 


Would you by any chance have the manual for the module on 
the FXO card?? I have the one for the FXS. In fact, I 
think you sent it to me.