Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!tut.cis.ohio-state.edu!rutgers!sunybcs!kitty!larry From: larry@kitty.UUCP (Larry Lippman) Newsgroups: sci.electronics Subject: Re: Using "Free" telephone power Summary: Limit to power draw from telephone line... Message-ID: <3128@kitty.UUCP> Date: 29 Apr 89 17:10:29 GMT References: <636@serene.UUCP> Organization: Recognition Research Corp., Clarence, NY Lines: 110 In article <636@serene.UUCP>, gbell@pnet12.cts.com (Greg Bell) writes: > This is in response to the person who was talking about lighting an LED from > the 50 VDC phone line voltage as an indicator for hook status: > > That's the easy project! But, how 'bout making a circuit that will light an > LED when the line voltage DROPS (ie. the phone goes off hook). I tried it > once, and got it to work moderately well. I used a combination of voltage > dividers to keep a transistor off if the voltage is 50 or so. Worked alright > until the phone rang... the circuit answered the phone! Got a little > annoying. No kidding! :-) That's what I'm been trying to tell sci.electronics readers for quite some time. > I'm still amazed at how phones run entirely off the phone line power. They > have lit dials, and audio amplifiers... and I had trouble lighting an LED! > (trouble means the phone company equipment gets loaded down enough that it > shuts off your line for a while!). > > Can anybody offer insite on the secrets of running audio amps and chips off > the meagre phone line power? How 'bout the ammount of current the phone line > can source? Again, here are some assorted comments about telephone loop voltages and currents: 1. The on-hook (i.e., idle) voltage of the vast majority of central office telephone loops ranges between 48 to 52 volts, with 50 volts being a pretty typical value. However, this is NOT ALWAYS the case, since on-hook voltages of lines supplied from various subscriber line concentrators, subscriber line carrier apparatus and subscriber lines on long loops with loop extenders can vary from a low of 6 volts (that's right, 6 volts using Continental FDM AML subscriber line carrier) to a high of 96 volts (using Cook Electric or equivalent loop extenders). A few electronic PABX's may have 24 volts instead of 48. There are two morals to the above: (a) Only a 500-type telephone set with passive network will work on all telephone loops at all times; anything else will fail under some circumstances. (b) If in doubt, measure the on-hook voltage of the telephone loop. 2. In general, good telephone circuit design practice dictates that no less than 100,000 ohms should be bridged across a telephone line in an on-hook state. That means you can draw a maximum of 0.5 milliampere of current with _assurance_ of not causing trouble with your telephone and not causing telephone company automatic line insulation test (ALIT) apparatus to print false trouble reports. 0.5 mA is, by the way, sufficient current to keep a well-designed reperatory dialer memory or clock circuit alive. 3. In general, good telephone circuit design practice dictates that no current path to ground should ever exist between tip and/or ring of the telephone loop and ground. From a practical standpoint, a value of at least 100,000 ohms leakage resistance can be used. The only exception to the above is equipment for use on ground-start PBX loops and for test purposes. An additional exception to the above involves party line circuits, but most telephone company tariffs prohibit the use of ANY apparatus on a party line circuit that is not directly obtained from the operating telephone company. 4. From a practical standpoint, bridging less than 20,000 ohms across a telephone line (i.e., a current flow of more than 2.5 mA) in an on-hook state is an immediate invitation to trouble, especially with respect to false ring-tripping as described in the referenced article. 5. In the off-hook state power to operate amplifiers and other circuit elements may be obtained from the voltage drop across a series resistor (i.e., in series with the talk circuit). From a practical standpoint, any such series resistance must be limited such that no less than 30 mA of loop current flows in an off-hook state. Taking an example value of a 250 ohm resistor on a 30 mA loop, we have 7.5 volts available with a DC power of 225 mW. 225 mW is actually a decent amount of energy to play with (in this day and age of CMOS circuits) - but you will ONLY have it available when the telephone is off-hook. If you attempt to obtain a larger voltage drop through a higher series resistor, such that the off-hook loop current drops below 30 mA you run the risk of: (a) having the central office apparatus fail to detect an off-hook condition; (b) being unable to trip ringing; (c) being unable to send rotary dial pulsing signals to the central office; and (d) having insufficent current available to operate telephone set DTMF dialing and talk circuits. 6. The "series resistor" described in (5) above may be effectively replaced with solid-state circuit elements, such as a constant current regulator set for a minimum of 30 mA. 7. Any series element for the purpose of obtaining DC power as described above should be bridged with a non-polarized capacitor to minimize the voice-frequency insertion loss of such element. It should be obvious from the above guidelines that any BRIDGED connection of an LED will draw excessive on-hook current, but that a series connection (i.e., "in use" indicator) is okay. <> Larry Lippman @ Recognition Research Corp. - Uniquex Corp. - Viatran Corp. <> UUCP: {allegra|ames|boulder|decvax|rutgers|watmath}!sunybcs!kitty!larry <> VOICE: 716/688-1231, 716/773-1700 {att|hplabs|utzoo}!/ <> FAX: 716/741-9635, 716/773-2488 "Have you hugged your cat today?"