[Asterisk-Users] Terrible crackling on analogue line and X100Pcard

[Stuart Ford]

Rich Adamson ...

> Okay. There are two fairly small integrated circuits on the board.
> Can you post the part numbers on those chips?

Not at the moment, unfortunately. The card is in a datacentre 200 miles 
away. However, and while I appreciate that you can't read anything from 
this, but this is the card:

http://i24.ebayimg.com/01/i/02/c3/8b/43_1.JPG

(the same Ebay vendor is selling dozens of them)

> NMI = non maskable interrupt (or somthing like that). Those messages
> would suggest there is a problem with that card and the wcfxo driver.

Right.

> The "mode is FCC" is saying the zaptel drivers are assuming a
> card that matches US telco standards. Again, without the chip set
> numbers, I can't tell if that card will work correctly in the UK.
> If it does not support UK standards, not likely you'll ever get
> the echo to go away.

Sounds like I've bought a hooky card then :(

> Okay, then there is about a 90% chance the card's chip set was
> designed for the US telco standards. I'll be able to tell more
> once you post those part numbers.

It's going to be a couple of weeks before I can do that, unfortunately.

> Are you using a "opermode=UK" or anything like that in /etc/zaptel.conf?

No. Should I try that or do you not think it'll make that much difference in 
light of the findings above?

> It would appear you have several interrupts that aren't being used.
> Have you tried looking at the bios setup to see if you can disable
> any unused interrupts (like 3 for com1 port)?

Yes I tried all that. Even freeing up other interrupts caused the system to 
fail to boot. I didn't understand it.

> If there is nothing in the bios relative to configuring interrupts, then
> you only choice is to move the card to other slots in hope of
> finding one that assigns a different interrupt.

Now that I didn't try.

> There is at least a better then 50% chance sharing the interrupt
> between the wcfxo driver and the raid controller (#11) is causing
> at least some of the crackling noise. You might try establishing
> a call and do a large file copy (to exercise the disk) to see if
> disk activity causes the noise.

I tried this and it made no noticeable difference. The crackling's always 
quite bad to start with, and since RAID controllers are basically small 
computers themselves it probably stands to reason that activity will be 
pretty much constant.

[zttest]
>> Is that good?
>
> Yes, that looks good.

That's something at least!

> You might try going back to an earlier version (or cvs head) to
> see if that has any impact on the noise.

I think that in light of what you've advised above I don't think this is the 
problem. I also completely disagree with regressing to former versions of 
software to hack something into working, since I might actually need a 
feature or bugfix that's provided by the later version in order to have 
something else working!

So, here's what I think I'll do:

1) I'm going to start with buying a genuine X100P, to eliminate the 
possibility of cheapness affecting things. However, as I said, I'm having 
tremendous trouble sourcing one. Does anyone know of any outlets in the UK 
who supply them? I've seen them for sale on several US websites, but I'm 
going to assume that they'll only work with US telcos?

2) Once I've got one, I'll have another stab at the interrupt mess again. 
I'll try to move the card between PCI slots and have another, more 
structured, session with the BIOS.

Sound like a plan?

Thanks

Stuart