Microfilters - surge protectors?

A friend of mine says that microfilters not only separate the digital
signal from the voice signal (but everyone knows that already), they
also act as a pseudo surge protector to stop your ADSL modem getting
fried, which is especially important for me because I have a PCI modem.

Is this true or is it tripe? I take what he says about computers with a
pinch of salt, after all, he swears by AOL (yes I said "by", not "at"!)

Anton

Anton Gysen wrote:
> A friend of mine says that microfilters not only separate the
> digital signal from the voice signal (but everyone knows that
> already), they also act as a pseudo surge protector to stop your
> ADSL modem getting fried, which is especially important for me
> because I have a PCI modem.
>
> Is this true or is it tripe? I take what he says about computers
> with a pinch of salt, after all, he swears by AOL (yes I said "by",
> not "at"!)
>
> Anton

Well i suppose it has got a capacitor inside, but personally I wouldn't like
to depend on it....

In article <buh5ij$rm1$(E-Mail Removed)>, Anton Gysen <anton@REMOVE
THECAPITALLETTERSantongijsen.com> writes
>A friend of mine says that microfilters not only separate the digital
>signal from the voice signal (but everyone knows that already), they
>also act as a pseudo surge protector to stop your ADSL modem getting
>fried, which is especially important for me because I have a PCI modem.
>
>Is this true or is it tripe? I take what he says about computers with a
>pinch of salt, after all, he swears by AOL (yes I said "by", not "at"!)
>
The one whose lid I removed had filter components on the phone side,
whereas the ADSL output was connected straight to the input. So if
there is any surge protection, it can only be for the phone.
--
Tom

On Mon, 19 Jan 2004 17:55:46 +0000, Anton Gysen
<(E-Mail Removed)> wrote:

>Is this true or is it tripe?

tripe, the ADSL port is pass through

see the bottom of http://www.adslnation.com/support/filters.php for a
circuit diagram

Phil

Phil Thompson wrote:
> On Mon, 19 Jan 2004 17:55:46 +0000, Anton Gysen
> <(E-Mail Removed)> wrote:
>
>
>>Is this true or is it tripe?
>
>
> tripe, the ADSL port is pass through

Thought so, he does come up with random "facts" a lot. I don't even have
a filter connected to my ADSL modem, it goes straight into the phone
line via a standard 2-way splitter, the only filter I use goes on the
phone which is connected to the other connector on the splitter (FWIW
it's the BT MF50 one that's supplied with the AOL Broadband start-up
kits and it seems to do the job OK for filtering voice calls. I would
like to point out that I do *not* use AOL! I use a proper ISP (NDO).

Anyway, is there any point in using a better microfilter with my setup?

Anton Gysen wrote:
>
> Anyway, is there any point in using a better microfilter with my setup?

As the saying goes, if it isn't broke don't fix it. If your filters work
fine then there is no reason to change them.

As for surge protection.
Your standard NTE5 BT master socket has a Zener Diode in that will
regulate the voltage from the line but will do little to prevent a
lightning strike.

Some filters such as our XF-1e also have a Zener Diode across the A&B
terminals to regulate the voltage further. Although there are no
components isolating the modem as you can see from the diagrams the A&B
terminals are connected straight through. Although there are no Zener
Diodes on the two low spec filters that have schematics on that page
some of the others do have them.

The purpose of the Zener Diode in our XF-1e is primarily there to
regulate the voltage to our transistor circuit but it also has another
purpose in that it will offer limited protection to the equipment
attached to the filter. Electricity will take the path of least
resistance. If you connect a lamp to a battery electricity will pass
through the filament to illuminate the lamp. If however you then place a
wire across the terminals of the battery the lamp will receive no power
as the electricity all flows through the wire.

Now the Zener Diode is connected across the incoming telephone terminals
and if the voltage increases the resistance of the Zener diode reduces
to absorb the excess power (turns it in to heat). If the voltage goes
high enough the Diode will effectively become a short across the
terminal. Basically this reduces the amount of power reaching your
modem. As long as the surge does not stay on the line long enough to
destroy the diode (Thermal Breakdown) your equipment will be safe.

It is not uncommon for a filter to take the hit from a surge destroying
the filter but protecting the modem. So while surge protection is not a
feature of a filter it can be a desirable side effect. If you are
concerned about lightning however the only protection is a full
lightning arrester, but even then they are not very effective, that's
why you have insurance with them.

On Mon, 19 Jan 2004 20:17:24 +0000, Anton Gysen
<(E-Mail Removed)> wrote:

>Anyway, is there any point in using a better microfilter with my setup?

only if you are having problems with voice quality or modem dropouts.
Your setup is correct, unfiltered ADSL, filtered phone.

Phil

On Tue, 20 Jan 2004 02:39:09 +0000, Darren Grant
<(E-Mail Removed)> wrote:


>As for surge protection.
>Your standard NTE5 BT master socket has a Zener Diode in that will
>regulate the voltage from the line but will do little to prevent a
>lightning strike.

There are no Zener Diodes in a NTE5 - there is a gas discharge spark
gap surge arrestor across the incoming pair.

>The purpose of the Zener Diode in our XF-1e is primarily there to
>regulate the voltage to our transistor circuit but it also has another
>purpose in that it will offer limited protection to the equipment
>attached to the filter.

Limited being the operative word. Microscopic would be more correct.

>Now the Zener Diode is connected across the incoming telephone terminals
>and if the voltage increases the resistance of the Zener diode reduces
>to absorb the excess power (turns it in to heat).

You are describing a Transorb (Metal Oxide Varistor), not a Zener
Diode. A Zener Diode has a defined reverse breakdown voltage above
which it switches very rapidly from non-conducting to conducting. A
Zener diode does not absorb power to any significant extent - it
would depend upon an associated load resistor for this.

The most effective cheap protection is gained from a combination of
Transorbs, which are quick reacting but have poor power handling
characteristics and Spark Gaps which are slow reacting but have
excellent power handling capabilities.

>Basically this reduces the amount of power reaching your
>modem. As long as the surge does not stay on the line long enough to
>destroy the diode (Thermal Breakdown) your equipment will be safe.

Except that such simple devices, even when they work, protect only
from voltage differentials between the A & B lines. With lightning
induced surge the problem and the damage in almost every case is not
the voltage difference between A&B - that doesn't change much, but
the difference between the voltage on the lines and earth.

The only way of protecting against this is to have a filter which is
earthed separately from the phone lines. Effective common mode
filters are both expensive and big.

>It is not uncommon for a filter to take the hit from a surge destroying
>the filter but protecting the modem.

If you are talking about your own ADSL filters this is more likely to
be down to them being designed and built with lower tolerances than
most telephony devices and simply failing earlier rather than them
having any protective effect. No ADSL filter will have any
measurable protective effect.

>If you are
>concerned about lightning however the only protection is a full
>lightning arrester, but even then they are not very effective,

Proper lightning protection, against even a direct strike, can be
most effective. You won't find it in most houses though (and it
costs more than most houses).

> that's
>why you have insurance with them.

The cheapest protection is to buy a combined mains/telco device such
as the Belkin ones which have insurance included. They don't work
any better than the others (ie are equally ineffective) but Belkin
are aware the risk of lightning damage is small and the are playing
an actuarial game. They do minimise damage from remote strikes.

--
Peter Parry.
http://www.wpp.ltd.uk/

Peter Parry wrote:
> There are no Zener Diodes in a NTE5 - there is a gas discharge spark
> gap surge arrestor across the incoming pair.
>

Traditionally yes they did, but that is no longer the case. Pressac are
the manufacturer of the current NTE 5 and it has a Zener diode in to
regulate incomming line voltage. At least that is what is inside the one
they sent me.

> You are describing a Transorb (Metal Oxide Varistor), not a Zener
> Diode. A Zener Diode has a defined reverse breakdown voltage above
> which it switches very rapidly from non-conducting to conducting. A
> Zener diode does not absorb power to any significant extent - it
> would depend upon an associated load resistor for this.
>
> The most effective cheap protection is gained from a combination of
> Transorbs, which are quick reacting but have poor power handling
> characteristics and Spark Gaps which are slow reacting but have
> excellent power handling capabilities.

Yes you are correct, allthough there is a knee voltage respose on the
reverse breakdown. The Zener in the NTE 5 is classic Zener Regulator.
With my limted electronics knowledge (I am not an electronics engineer)
the resistance is found in the wire from the exchange. More information
can be found at http://www.btwebworld.com/sinet/ in SIN document 351.
As I understand it the line provides a 70v supply and the job of the
Zener is to prevent it exceeding 100v.


>
> Except that such simple devices, even when they work, protect only
> from voltage differentials between the A & B lines. With lightning
> induced surge the problem and the damage in almost every case is not
> the voltage difference between A&B - that doesn't change much, but
> the difference between the voltage on the lines and earth.
>
> The only way of protecting against this is to have a filter which is
> earthed separately from the phone lines. Effective common mode
> filters are both expensive and big.

Exactly what I said, this can not be considered protection from a
lightning strike only over voltage from the BT Exchange.

>
> If you are talking about your own ADSL filters this is more likely to
> be down to them being designed and built with lower tolerances than
> most telephony devices and simply failing earlier rather than them
> having any protective effect. No ADSL filter will have any
> measurable protective effect.
>
I am talking about filters in general, mainly failed Excelsus ones from
customers who have needed to replace them. It would seem that the
breakdown of the components in the filter has resulted in less impact on
the connected equipment. Not by design but just one of those things that
the components in the filter seem to take the hit.

>
> The cheapest protection is to buy a combined mains/telco device such
> as the Belkin ones which have insurance included. They don't work
> any better than the others (ie are equally ineffective) but Belkin
> are aware the risk of lightning damage is small and the are playing
> an actuarial game. They do minimise damage from remote strikes.
>
That again is exactly what I was saying, lightning protection to most
means consumer products such as the belkin ones. They can offer some
protection but not a direct hit, hence the insurance.

On Tue, 20 Jan 2004 14:43:38 +0000, Darren Grant
<(E-Mail Removed)> wrote:


>Traditionally yes they did, but that is no longer the case. Pressac are
>the manufacturer of the current NTE 5 and it has a Zener diode in to
>regulate incomming line voltage. At least that is what is inside the one
>they sent me.

I'm not sure you understand the purpose of a primary protective
device such as that in the NTE5. They are not designed to protect
equipment but to protect the cabling within the house by limiting
voltages and currents to levels below which the wiring might ignite.

Look at the different characteristics of the components and you will
realise either what you are looking at is either not a Zener or not
an NTE5. It cannot be a single Zener Diode connected across A&B.

A typical gas discharge device used in an NTE 5 will trigger at
between 150/250V and is capable of a surge discharge current of 5 to
ten thousand Amps for 20microseconds. It fails safe to a permanent
short circuit.

A typical Zener diode (of size capable of fitting into an NTE5
circuit board) will have a transient capability at 150V limiting of
about 13Amps for 10 microseconds. This is woefully inadequate. It
will often fail unsafe (open circuit). Zener diodes are designed for
voltage regulation, not for protection against surge impulses.

Moreover a single Zener diode is a normal diode below the Zener
voltage - it always conducts in one direction which would now make
the NTE5 line polarity dependent.

Transient Voltage Suppressors (TVS) are bipolar avalanche
semiconductor devices similar in packaging to a Zener diode which are
designed to provide secondary protection against voltage and current
transients. The silicon TVS is designed to operate in the avalanche
mode and uses a relatively large junction area to absorb large
transient currents. Even so its capacity is nothing like that of a
spark gap and it would mainly be used in equipment to remove the
spikes a spark gap will pass (and cause) before triggering. TVS's
are usually designed to fail short circuited (safe).

>> You are describing a Transorb (Metal Oxide Varistor), not a Zener
>> Diode. A Zener Diode has a defined reverse breakdown voltage above
>> which it switches very rapidly from non-conducting to conducting. A
>> Zener diode does not absorb power to any significant extent - it
>> would depend upon an associated load resistor for this.

>Yes you are correct, allthough there is a knee voltage respose on the
>reverse breakdown.

I'm not at all sure what you are taking about here.

>The Zener in the NTE 5 is classic Zener Regulator.

There is no need for a regulator on POTS circuits.

>With my limted electronics knowledge (I am not an electronics engineer)
>the resistance is found in the wire from the exchange.

The protective device is there to prevent damage from transients on
the line which can occur at any point - you cannot rely upon wiring
resistance to the exchange.

>As I understand it the line provides a 70v supply and the job of the
>Zener is to prevent it exceeding 100v.

The job of the exchange is to limit the line voltage. The job of the
primary protection device is to limit transients induced between the
exchange and the subscriber to a point where they will not pose a
danger to the wiring. The job of a secondary protective device is to
protect the equipment.

>> The only way of protecting against this is to have a filter which is
>> earthed separately from the phone lines. Effective common mode
>> filters are both expensive and big.
>
>Exactly what I said, this can not be considered protection from a
>lightning strike only over voltage from the BT Exchange.

You don't get overvoltages from BT exchanges.

>> No ADSL filter will have any measurable protective effect.
>>
>I am talking about filters in general, mainly failed Excelsus ones from
>customers who have needed to replace them.

That is an ADSL filter is it not?

>It would seem that the
>breakdown of the components in the filter has resulted in less impact on
>the connected equipment. Not by design but just one of those things that
>the components in the filter seem to take the hit.

The fact that a filter has failed does not mean it protected anything
other than the users imagination. As the circuit configuration
provides no protection it is difficult to see how you could justify
this hypothesis.

>That again is exactly what I was saying, lightning protection to most
>means consumer products such as the belkin ones. They can offer some
>protection but not a direct hit, hence the insurance.

Direct hits are not the problem (in that no consumer device will do
anything other than to melt). It is ground strikes some distance
away which produce a damaging voltage gradient.
http://www.st.com/stonline/books/pdf/docs/3690.pdf covers the
subject.

If you want to see UK lightning activity it can be viewed at
http://irishwan.org/modules.php?name...showpage&pid=8 or
http://www.meteorologica.co.uk/freedata_lightning.asp



--
Peter Parry.
http://www.wpp.ltd.uk/