On 1/19/2017 11:57 AM, Tim Wescott wrote:
> On Thu, 19 Jan 2017 10:32:30 -0600, Tim Wescott wrote:
>
>> On Thu, 19 Jan 2017 02:08:33 -0500, rickman wrote:
>>
>>> It has been a long time since I designed any telephony equipment. I
>>> know the E&M circuits are simple conceptually, but there are any number
>>> of real world issues that impact the design, which I can't remember
>>> much of.
>>>
>>> The battery voltage is approximately 48 volts and the circuits are
>>> current limited. Control is done with relay contacts in the old days
>>> and sensing was done with a relay coil. Now I'm sure that has all been
>>> replaced with a few transistors. I tried looking for ICs that would
>>> handle this, but came up empty handed. Is there anything out there
>>> that is designed for this? I know there are chips for the SLIC and
>>> such, but I don't see anything for E&M.
>>>
>>> This is not for a circuit to a CO, so there won't be exposure to
>>> lightning or AC wire crosses. This will be between a PBX and
>>> networking equipment (circuit to packet). I'm not sure it even needs
>>> 48 volts since the interface is really current driven. Anyone know if
>>> E&M typically uses a lower voltage these days like 12 volts?
>>
>> It's been 30 years since I've done any circuit design for telephony
>> applications. Weigh my responses accordingly.
>>
>> Is the audio channel on E&M wet? Or are you talking about the signaling
>> wires?
Signalling.
>> I'm surprised there's modern telco systems out there that would prefer
>> to run a bunch of E&M trunks someplace rather than just a fiber -- I
>> have this vision of fiber run to a little beige box just outside the
>> building,
>> then 20 feet of trunk lines into a PBX that has no excuse not to have a
>> fiber connection.
That's what this is, a connection from the PBX to a network that will
carry the calls over Ethernet by whatever physical medium you have.
My boards go in the existing products this company offers, but they are
changing their business model from a hardware/software company to just a
software company. They are handing the hardware business (for this
product line) to another company and will only sell the software.
Strange but true! So my boards will no longer be used.
My contact with the company suggested I might be able to get a little
consulting business with the hardware company since they know nothing
about telephony. Unfortunately I don't remember much myself.
>> I suspect that most people use solid-state relays, or perhaps they roll
>> their own SSR circuits. I'd like to say just use a MOSFET, but that
>> assumes that every telco gets the polarity right: it's probably a good
>> idea not to count on that.
I think they get the polarity right... or you can just use a diode
bridge to connect the current sensor. In some configurations power
comes from the PBX only, in others both ends source power. They don't
mix up the polarity, but it will change depending on the configuration,
type 1, 2, 3, etc.
> Oh -- and I'd design for at least 48V, just to be on the safe side. If
> there's one thing I remember about working with telephony, it's that
> almost nothing you can do will break those expensive wires stretched from
> hither to yon, but getting good consistent operation out of the thing
> depends on sticking to the BelCor standards (or whatever is used today).
Yeah, thanks.
--
Rick C