Tuesday, September 8

Tim O'Neill: I got your email, but the return address was wrong. Please write to me again.

More about radio interference

Last week's discussion of radio interference produced a flood of similar stories and some suggestions for overcoming interference. The problem isn't just with modems, either. Phil Rolfe wrote:

I just wanted to add some more feedback to your lively discussion of radio interference. I just moved away from a house that was just 50 yards or so from a radio tower. I can tell you that there were times when you could hear the music on your phone and even over your speakers (when you had it on but in cassette play mode you could listen to it as if you really were tuned into the particular station).

Charlie Han had a similar problem with his computer speakers:

This is in response to your article about radio interference with modems and such. I don't have those problems described, but I could hear an AM radio station if I turn my speakers up to max without any audio coming out (i.e., in DOS with nothing running but speakers are on max volume) and I could hear a AM Radio talk show... but the computer in my other room would not have that problem.

Other readers reported problems with picking up radio broadcasts on their modem, as discussed last week. Russell May:

This is in response to your request for experiences with interference from nearby AM radio stations. I live about 3/4 mile from a pretty strong radio tower(5 Kw, 610 KHz). A few years ago I had a 14.4 Kbps VIVA internal modem which picked up the radio station very well. I could hear country music on the modem speaker, connections were unreliable, busy signals would sometimes be recognized which were not there. I bought an RFI filter for $18.00 from a local AT&T phone store. Its label says "Z100B". Inside it has a 7 MHz choke. I had no problems with the modem after connecting the RFI choke.

Solutions

Like Ken Lotts last week, Russell May solved his problem using an RFI filter.

Jeff McClure pointed me to an article on the American Radio Relay League, Telephone EMI (electromagnetic interference, which includes radio waves):

This is in response to the Friday (4 Sept) article concerning radio interference to telephones. There are any number of things that could cause it, but I thought I'd point you to a web page that could be helpful. It was published by an organization called the American Radio Relay League. The ARRL is made up of amateur radio operators, and the article is geared toward helping hams find and correct the cause of our own signals interfering with phone equipment. Since the info is for hams, some of it won't be applicable, but the troubleshooting techniques are exactly the same as you'd use to try to eliminate commercial RFI in your own telephone lines. Take a look.

Keep in mind that the cause of this kind of RFI is almost always either defective equipment/wiring or bad design in the telephone device. Many times, one device on the line (possibly like the modem mentioned in your news article) will have a component that acts like a diode detector. It takes the RF energy, efficiently converts it to audio, and then injects it into the phone line (where every device on the line will hear it). Sometimes the only solution is to replace the offending device with one that doesn't have this problem.

For anyone who is interested in further info, the ARRL offers a free "RFI kit" that explains the legal, technical, and even social details surrounding this issue.

Jeff Clark notes that your premises wiring can also serve as a radio antenna if it's miswired:

Your contributor who is picking up radio on the phone lines probably has a wiring problem. The audio lines are supposed to be a twisted-pair which tends to reject external interference. It is likely to be in the house phone wiring. Use the trick that you posted earlier of plugging in your modem where your phone service enters the house. If the problem shows up there, call the phone company to fix their wiring between your house and their switch. If the problem goes away you need to fix your house wiring.

Two common ways to compromise your twisted-pair wiring are:

1. Missing the pair.

When people wire phones within houses they sometimes don't pick paired wires for the audio signals. You don't get common-mode signal rejection if your wires are twisted with other signals rather than as a pair. I don't have the color code with me, but at least check that the colors match from the house entry to all the phone jacks. When in doubt pay the phone company to do it. Inside the house it's your problem. If the miswire is outside of your house the phone tech should figure it out and the fix should be at their expense. Be sure to say that it is a sound problem rather than a modem problem. They don't fix modem problems.

2. Miss-matched wire lengths.

When people wire or rewire phones in houses they sometimes leave wires going to unused jacks. If the jack isn't going to be used or if the jack is for another phone line they sometimes only partly connect the wires. You can end up with one wire of your audio pair connected out to this jack and another that doesn't run out to the jack. The result is that one wire of your pair has an antenna hung on it, the run out to the remote jack. To fix this open all of your phone jack boxes and look for inconsistent wiring. If you know certain jacks aren't being used try and disconnect them completely, it will help in many ways. If you have a multimeter their is a quick electrical go/no-go test. Unplug all phones, modems, etc. from phone jacks. Unplug the main jack where phone service enters the house. Use your voltmeter to check that the lines are dead. To test for unmatched wire lengths you test for unmatched wire capacitance. Using high-resistance setting on your ohmmeter test how long it takes to read infinite resistance between each audio line and ground. When you first connect the ohmmeter to a wire current will flow into the wire's capacitance and you will get a resistance reading that runs up the scale as the capacitance charges. If there is much of a difference between the charge times of the two lines you have a stub or detour on the line that takes the longest to charge.

In perhaps related news, Pacific Business News reports that contractors are now installing better quality phone cables to keep up with the demands of faster modems.

---

Friday, September 11

The U.S. Patent and Trademark Office granted Patent Number 5,801,695 to 3Com and Brent Townshend for a technology underlying today's 56K modems, according to News.com. Rockwell and Townshend are still engaged in a lawsuit, and Lucent claims it holds competing patents.

The Washington Post reports that ISPs are upset about what they consider anti-competitive practices on the part of Bell Atlantic. Bell Atlantic charges customers a $523 signup fee for high speed DSL if they use another ISP, but charges less than $200 if the customer uses Bell Atlantic for their DSL ISP. The telephone companies have such strong control over DSL provisioning that anti-competitive practices and the appearance of anti-competitive practices are inevitable.

Computer Shopper and Network Computing examine progress towards ADSL standards.

Jonathan Yuen found the trick for getting a Macintosh-compatible V.90 upgrade for his Hayes Accura. The tip was published on Ric Ford's Macintouch web site, where it was submitted by Gabriel Sterritt:

To get the real v.90 update for Macs, you need to email customer.service@hayes.com with the number "5011" in the subject. You will receive an autoresponse message that includes the v.90 firmware update file.

However, Jonathan adds a word of caution:

I installed the upgrade and it refused to correctly connect with my school's 33.6 modems (always a dropped carrier right after connection). So, I'm a bit wary of the upgrade. (The upgrade also includes the flasher to downgrade back to k56flex).

Son of more radio interface

Following up on recent reports, Jesda Gulati offers another example of the value of an RFI filter in dealing with radio interference:

Infinital Networks is selling RFI filters for $12 plus shipping. RFI filters drown out radio interference almost entirely from phone lines. Here in Spokane, Washington, this section of town is home to almost all local radio and TV stations. I live within at least a mile or a mile and a half of them. On a 28.8 Fujitsu PCMCIA modem, performance was boosted from 4800bps to 26400bps to Infinital Internet. On a Global Village 33.3 Platinum, a jump was made from 9600 to 32000. Normally $18 or $20, these sell for merely $12. To obtain one, visit http://infinital.ml.org/.

Previous week