Disadvantages of 56K

• Introduction

• Only one A/D conversion allowed

• 40 to 50K more likely

• 56K for downloads only

• 53K limit in the U.S.

• Requires digital lines at the ISP

• Won't work with some PBXes

53K limit in the U.S.

Because of FCC limitations, x2 and V.90 are effectively limited to 53K in the U.S. K56flex is not affected. This isn't a huge problem, as most people probably wouldn't connect at the full 56K anyway, due to the phone line impediments that prevent many people from connecting at 28.8K, and that prevent most people from connecting at 33.6K. And, really, it's just a 6% difference. It's a PR loss because doubting Thomases have latched onto it as a reason to criticize 56K technology.

There is quite a bit of mis-understanding on this subject. Here's some clarification:

• Only x2 V.90 modems are affected. K56flex modems are not affected. See below.
• The FCC's rules apply only in the U.S. Users in other countries have reported full 56K connections.
• The FCC does not limit modem speeds per se. Instead, it limits the amount of power that can be sent through the phone line in order to prevent interference with other electrical devices.
• It's possible that the FCC may grant a waiver in the future that would allow speeds of 56K or higher in the U.S.
• The 53K limitation is not built into the modems you buy. Instead, it is enforced in the server equipment at the online service or Internet service provider you're dialing into. A software change at the ISP could remove the limitation, and server equipment shipped overseas does not have the limitation.

September 18 update: FCC moves to remove limit

The FCC has proposed increasing the power limit to eliminate the effective 53K ceiling. CNN and InfoWorld have coverage.

What do you mean K56flex isn't affected?

I was first alerted to this by a Communications Week article. In the article, Mike Ziehl, Communications Product Marketing Director for Rockwell, is quoted as saying:

Our product allows 56K connection while living within the FCC power limitation," Ziehl says. "It's just the general noise interference in the local loop that makes it unlikely that you'll achieve 56K.

I contacted Rockwell for confirmation, and received this reply from Eileen Algaze of Rockwell public relations:

On the 56K/FCC issue: Mike Ziehl's statement was that RSS product was designed to be able to get 56K while living within the FCC power limitation, BUT that other phone network impairments like noise on the local loop etc would still prevent 56K from happening in most cases. We do not "get around or bypass FCC" we are fully compliant with the FCC power limitation. We just utilize different coding schemes than x2 that allow for the average power over a 3 second interval to remain less than -12dbm as the FCC specifies.

Perry Ho, Product Marketing Manager for Lucent Technologies' Modem and Multimedia Group, also confirmed that K56flex is not limited to 53K.

In a comp.dcom.modems thread on this subject, Dan Moore of Diamond Multimedia offered more information on the subject of the 53K limit:

The FCC/PTT power limits apply to all modems, X2, K56flex, and V.90. The -12dBm limit on transmitter power for digitally attached equipment in the US limits the initial version of X2 to 53K. I do not know what the limit is for the revised version of X2 that 3Com was (is?) testing. K56flex uses a different data phase mapping scheme than X2, this allows it to achieve 56K data rates with the -12dBm limit. The V.90 mapping scheme, modulus conversion, is the same as the X2 method. V.90 allows different PCM code sets in the 6 time slots in a data frame. This gives it a some flexibility in controlling the transmitted power by using different constellations (PCM code sets) in each phase, this may allow it to achieve 56K with a -12dBm power limit. (I honestly don't know, I haven't seen any simulations of the power level of a 56Kbps V.90 constellation.)

I've seen several messages asking what, if anything, is being done to change the power limit currently set by the FCC. TIA TR30.1 (the US standard committee for modulations) has sent requests to the appropriate committees asking for an increase in the allowed power level on digitally attached CPE in the US. It appears that power levels as high as -6dBm would not cause any problems in the current US network. It will take many months (or more likely years) for this change to occur, there are a LOT of groups involved and all must agree or at least not object. (The RBOCS are involved, the LD carriers are involved, several TIA and/or ANSI committees are involved, the FCC is involved. Unfortunately this isn't a simple or quick thing to change.)