Getting matching transformer from telephone

According to Roy, you want a ground loop.

He is such an electrical genius!

NOT!

bwuahahahahahahahahahahahahaha!

Interesting thread (not the childish abuse, the technical stuff!). I didn't know that loading coils were so common in the US, for example.

My twopenn'orth, as someone who has designed line interface circuits in the past:

This is true - someone (not sure who, sorry) commented earlier that this doesn't matter because the speed of light is so high, and that's true for local calls but not for long distance ones. These days there is packet delay to worry about as well. There are echo cancellers but they're not perfect. Mismatches also affect loudness.

It's true that isolation is not important for a well-insulated telephone. It's also true that differential amps are a cheap alternative to transformers for the hybrid part. But I have definitely seen phones with transformers in them in the past (the 80s). And I have designed interfaces myself (for modems) that used transformers in preference to op amps because of their superior common mode rejection, isolation and (electrical) robustness.

I have, from the days before transistors - they were specially constructed to tolerate DC current without saturating the core or overheating.

--
Phil McKerracher
www.mckerracher.net

A 10kV arc from a 2MV lightning strike *COULD* make it all the way into the handset, and OUT of the perforations in the handset, through the earpiece or mouthpiece, and hit the user. The microphone and the earpiece transducer both use metal cans, making the distance to the user a mere 1/4" through air. Not good.

This is ONE of the many reasons that isolation elements are incorporated at VARIOUS locations in the system. One of which is at the CABLE connection to the phone itself, which is why isolation elements can be found at these positions. This is a standard element of device design where human contact is present, and has nothing to do with it being in a plastic case. It isn't your Dad's AC fed two wire drill motor with an un-phased power cord and metal case. It is, however, in close (electrical) proximity (potentially) with lightning events, and that is why arresting elements have been incorporated.

Most incorporations are overkill, as it were, but I am happy that our scientists and engineers of decades past were concerned about such things.

Idiots today seem to think everything is low voltage and harmless.

:On Fri, 2 Jan 2009 02:25:48 -0000, "Phil McKerracher" : wrote: : :>

:>It's true that isolation is not important for a well-insulated telephone. : : : A 10kV arc from a 2MV lightning strike *COULD* make it all the way into :the handset, and OUT of the perforations in the handset, through the :earpiece or mouthpiece, and hit the user. The microphone and the :earpiece transducer both use metal cans, making the distance to the user :a mere 1/4" through air. Not good.

Let's get one thing straight about the old POTS telephone. It has never been designed to include an "isolation transformer" - where "isolation" means galvanic isolation....

A plain old POTS telephone is basically sacrosanct in terms of its lack of requirement to include galvanic isolation to anything. All it has to withstand is the normal voltage potentials found on the line itself.

However, ANY Customer Equipment (CE) which is powered from or has any connection to the mains supply, and which interfaces to a telephone line, MUST have galvanic isolation. The POTS telephone itself DOES NOT.

CE which is connected to the mains supply MUST have an approved mains transformer (or SMPS) with the required galvanic isolation, and separation of the telephone line side from the mains powered side using an approved line interface transformer (600:600 usually). Also the spacing between any conductors on the mains powered side and the telephone line side of the circuitry must be at least 6mm (iirc). This includes any opto isolators used to provide signal interface from the mains side through to the telephone line side. These devices must have physically wide separation between terminals on the the input and output sides, and they must a high voltage breakdown rating (eg. 7.5kV). In this way 3 levels of galvanic isolation are incorporated in mains powered CE to pride galvanic isolation "from the mains powered side".

: : This is ONE of the many reasons that isolation elements are :incorporated at VARIOUS locations in the system. One of which is at the :CABLE connection to the phone itself, which is why isolation elements can :be found at these positions. This is a standard element of device design :where human contact is present, and has nothing to do with it being in a :plastic case. It isn't your Dad's AC fed two wire drill motor with an :un-phased power cord and metal case. It is, however, in close :(electrical) proximity (potentially) with lightning events, and that is :why arresting elements have been incorporated.

In the old days of magneto, central battery manual and even central battery auto equipment there was no isolation in the terms you mention above. Yes, they did have "protection" devices on the telephone line to limit voltage surges and high current ingress onto the line, but the telephone line today is essentially a straight connection from the line interface equipment (which these days may have VDR's and that's it) through to the telephone at the other end. The types of protection which were once used on the POTS line were; fuses to isolate the line in the event of a heavy current surge, heat coils to break the circuit if an aerial line came into contact with a foreign potential which did not produce sufficient current to rupture the line fuses, and lightning arresters (spark gap or gas discharge). These devices were provided at both the exchange end (on the MDF) and the customer end of the line (eg. protector #1). After the 1960's these items were rapidly disappearing from the telephone environment due to the majority of line plant being underground where it had almost zero chance of coming into contact with foreign potentials.

In modern days where most cabling is underground, these devices are dispensed with - no heat coils (which were only really required for pole mounted telephone lines), no fuses and no lightning arresters. Even today a 3 stage surge protector (gas discharge tubes, fusible resistors and VDR's) is only required to be connected to a telephone line at the either the customer end or the exchange end where part of the construction is aerial, and/or the local area is prone to constant lightning activity. Due to the requirement for customer premises cabling and CE to conform to standards which provide for adequate separation from foreign voltage sources there should be little chance of a breakdown in galvanic isolation from these sources to the telephone line.

On Fri, 02 Jan 2009 00:32:00 GMT, Paul B wrote:

: :Hello Tony and everyone, I am the OP. : :I've been away for a few days and I see there's so many posts that :now I'm trying to get through them all! : : :MY OBJECTIVE : :My aim is to take voice recordings made on various equipment and save :them to a PC. Some of the voice recordings are of telephone :conversations made onto tape. I would prefer to have fed the phone :signal direct to the PC but I get a lot of noise. : :I want to preserve as much quality as possible because it will :probably be necessary for a third party to identify the person :speaking. : :------ : :Secondly and quite separately from the above.... : :I didn't raise this problem in my first post. I am getting hum and :noise when I record using a purpose build connector (Retell model 156 :~ see link below) to a hand-held battery-powered flash-memory :recorder even when the phone is on hook. I can't see where the hum :is coming from unless it is on the phone line because there can't be :a ground loop this time. : :

: :I do know my landlines don't have all the hum and noise so they must :be doing something which I want for my recorder! I thought may be a :transformer to better terminate the Virgin Media phone line might :help but I am out of my depth here and line termination may be the :wrong idea altogther. : : :DEFINITIONS : :I guess my use of the word "matching" is not a very good electrical :description. I'm not seeking to match impedances and I get the :feeling that in electrical engineering, "matching" is often shorthand :for impedence matching. So apologies for any confusion I have :caused. : :I want to minimise any ground loop to reduce hum and other spuriae so :perhaps I should have said "isolating" transformer. : :Retell have a model (the 157) which connects direct to a PC and I :believe it is identical to the 156 except it has the additional :transformer I am asking about.

Get an approved line isolation unit such as this

Since your PC is mains powered and it may not have the required isolation between the mains side and the sound card input you can do your own thing using an approved 600:600 transformer with 3kV isolation rating to interface the telephone line to the sound card input.

Actually, the Delta LIU is a bit of an overkill for what you want to do so here is a much cheaper solution

Basically this will just be a 600:600 line transformer without the DC signal pass-thru capability in the Delta unit.

:Actually, the Delta LIU is a bit of an overkill for what you want to do so here :is a much cheaper solution

: :Basically this will just be a 600:600 line transformer without the DC signal :pass-thru capability in the Delta unit.

If you have access to Farnell then here's a suitably rated 600:600 line isolation transformer

Actually EC's damned near are perfect!

Years ago, when Data Set Termination cards first started using echo cancellation we had a field technician install one and then have a fit trying to test it! The loopback unit on the customer side of the DST appeared to go into loopback, but the testboard technician did not see a tone coming back. So out came the test equipment, and sure enough the tone was looped back, but it didn't get through this DST. Yet when a tone generator was applied, that tone worked fine. Upon investigating the new type of DST it was discovered that it had EC... and it was literally reducing the level of the looped tone to the point were the test equipment at the testboard didn't even see it.

Whatever, mismatches don't affect loudness so much as the frequency response and transient response. It will make a call sound "tinny", and that is perceived at needing more loudness to understand.

It's true that hearing the echo as an echo requires significant time delay, but hearing the echo as an annoying distortion requires very little delay.

Isolation on a tel line is not related to insulating the user from the device.

Long after the transistor became ubiquitous telephones virtually all used transformers. It wasn't until extremely good, extremely cheap op amps were developed that *any* telephone design went without a transformer. Even today, a lot of telephones have a transformer though, but it is encapsulated and not easily recognized.

Indeed, the difference between other transformers and the hybrid network in a telephone is being able to handle at least 120 mA of current without saturating (I've never heard of one overheating...).

A number of cheap substitutes at various times have tried to get by with less able designs, and many modems have been marketed with transformers that saturate at something far less than is required. It often results in poor data error performance when used on lines with more than 30 mA or so of loop current (which do happen to be very common).

--
Floyd L. Davidson 
Ukpeagvik (Barrow, Alaska)              floyd@apaflo.com

Noises that sound like a hot frying pan with sizzling oil in it, are caused by leakage to ground on one side of the telephone line. Noises that are 50-60Hz hum are caused by an imbalance on the telephone line. You can easily have a combination of both too.

That is almost guaranteed to be a problem with balance. I'd expect your PC input is an unbalanced circuit, while the telephone is a balanced circuit. Connecting them together causes the tel line to be unbalanced, which picks up noise.

Probably as above, mixing balanced and unbalanced circuits. Given that they don't exactly tell us what is in any of those products, it isn't possible to be sure when and where any of them can be used.

I doubt you have a ground loop, and an isolation transformer probably isn't going to help (though there is a remote chance that it would... but it is very remote).

That appears to be true.

Consider some of the characteristics you need in your equipment. You'll want a device that "bridges" the line. That will prevent it from loading the line down and reducing the volume to the person using the telephone set. Typically a bridging device has about

10x the impedance of the circuit being bridged. (That will cause only 0.3 dB reduction in volume due to the extra load. However, it also cause about a 30 dB reduction in signal to the bridging device! Probably not significant for recording, but it does mean that transmitting a recording back to the telephone line requires significant power output.)

You also want the device to have no DC connection, so that it will not interfere with the supervision of the telephone line (on/off hook status). Use a non-polarized capacitor, somewhere from 1 to 4 mfd, to provide that isolation. Ideally that would be placed at the mid-point of a split winding transformer primary; or one capacitor on each side in series; but could be just one capacitor in series on one side.

If you connect such a device directly across the telephone line, you will get all the clicks from the hook switch, you will get the ring current full force, and you will hear the local user very loud in comparison to the distant user. Hence you probably would want to connect it internally in the telephone set, across the receiver (where the telephone design will have reduced the effects listed above).

The secondary of the transformer should be as appropriate for what ever device you are connecting it to. That may or may not be high or low impedance and may or may not be balance or unbalanced. The balance issue, however, doesn't actually make any difference because of the isolation provided by the transformer. The impedance is a bit tricky, because if the recording device is high impedance then you'll want a 1:1 transformer, but if it is 600 Ohms you'll want one with a 1:10 impedance ratio.

--
Floyd L. Davidson 
Ukpeagvik (Barrow, Alaska)              floyd@apaflo.com

He DOES NOT need a 600 ohm transformer since the input impedance of the sound card is not 600 ohms !

It's more likely to be in the tens of kilohms.

What he really wants is something like a 10k:10k 'line bridging' transformer. In practice, using a '600 ohm' transformer will probably be ok, but being incorrectly loaded will degrade the sound quality (freq resonse may be peaky etc).

Graham

What kind of screwy cables are those ? 1kHz has NOTHING to do with their characteristic impedance.

Here we use CW1308.

And why the heck do you think ADSL is based around an assumption of typical 100 ohm charactistic cable / line impedance ?

CRETIN !

Graham

Try running ADSL down a line with loading coils ! There's no need for them anymore. These days you just adjust the line gain.

Graham

I'm afraid I wasn't around in 1935. 73 yr old technology is somewhat irrelevant.

These days a phone is a chip and a few passives (that *aren't* big inductors).

Graham

OLD (and no longer used) is all you seem to know about.

Which go via optical fibre or microwave link, NOT cable any more.

Graham

In the UK the 'master socket' contains the 'spark gap'.

Graham

using

card

In

incorrectly

An 'audio expert' would put a 600 ohm resistor across the secondary to provide proper loading but no expects you to know that.

--
http://improve-usenet.org/index.html

aioe.org, Goggle Groups, and Web TV users must request to be white
listed, or I will not see your messages.

If you have broadband, your ISP may have a NNTP news server included in
your account: http://www.usenettools.net/ISP.htm


There are two kinds of people on this earth:
The crazy, and the insane.
The first sign of insanity is denying that you\'re crazy.

Some classsic audio howlers in there !

High-fidelity Microphone

" Common uses include matching the relatively low 2K ohm output impedance of a microphone to an amplifier?s much higher line input impedance of

10K ohms. Studios commonly use the three terminal ?XLR? type of connector which is a balanced connection method with a terminal for a center tap. A separate ground terminal, tied to the XLR connector?s case is almost always present. The center tap may be used to phantom-feed a small amount of current for powering a pre-amp or active ?condenser? microphone."

LMAO !

The rest looks useful though.

Graham

Not irrelavant when we were discussing the history of why 600 ohms became the standard impedance for audio distribution.

Modern phones don't use carbon mics either yet you claim

Perhaps you'd prefer the inside of the ubiquitous BT 746 which was introduced in 1967 and was around for a very long time - we didn't replace ours till well into the 90s

This is a 1978 version.

In 1985 it was fitted with the new style plug and became the 8746 but still had the "Induction coil"

However, regardless of what you know or don't know, this thread was started by someone who wanted to know whether he could re-use the transformers in telephones for another purpose so clearly he has some!

--
Stuart Winsor

For Barn dances and folk evenings in the Coventry and Warwickshire area
See: http://www.barndance.org.uk