BPL/Main.net Observations

In July, 2004, Rochester Public Utilities (RPU), Rocheser, MN, activated a small BPL test area of approximately 10 subscribers. Main.net equipment (Generation One Technology using DSSS) was used. Radio frequency noise measurements at the RPU BPL trial site were made beginning in late May by RARC club members Mel KCØP and Niel WØVLZ. Measurements were taken on multiple days using an ICOM 706 and Hustler mobile antenna system to establish baseline RF noise levels, noise levels during initial BPL system operation and noise levels after amateur radio band notching had been implemented. Measurements were taken on all amateur bands from 160M through 6M. A tabulation of these measurements has been forwarded to the Ed Hare, W1RFI.

Most of the monitoring was done parked near a pole that had a Main.net BPL repeater mounted on it. RPU provided access to a computer attached to the BPL network so that we could increase network loading ourselves. Main.net informed RPU when all HF ham bands were notched on this network.

Summary of Observations

• We heard no noise from the Main.net system below about 3MHz or above 23MHz.
• When no transfers were active we heard sparse clicks scattered across the ham bands. We are concerned that these (heartbeat?) clicks will intensify as the number of subscribers increases.
• The most disturbing interference was noted when we downloaded large files over the 'net (a 600 MByte Linux download is great for this). During the downloads we noted increased noise on 80 through 30 mtrs while 20 mtrs and up remained relatively quiet.
• Transmitting on the 17 mtr band took TCP/Internet throughput to zero. (As measured using the Windows XP Performance Monitor in Control panel -> Administrative Tools). See http://www.io.com/~nielw/BPL/RF_susceptibility.htm for additional information about RF susceptibility.
• We had to move two blocks out of the test area to clearly receive the "15 MHz WWV away from BPL test area" signal
• Background noise on the HF bands varies somewhat day-to-day. Main.net interference might be confused with normal background noise or with that caused by other interference generators. At the same time the design of the main.net equipment does not allow easy shutoff and restore of service. This makes it difficult to confirm without any doubt that a particular interference is BPL related. During this initial test consisting of less than 10 subscribers we were fortunate to monitor at times when we were the only users and could control BPL activity.

The following audio files illustrate the type of interference/signals observed on 3.924 MHz and 15 MHz WWV with and without BPL transfers active. These recordings were made on September 7, 2004 with notching active on the ham bands including 3924KHz. WWV at 15MHz was not protected by notching. They reflect only a single active user/download.

• 3.924 MHz with no BPL transfers active (325KBytes)
• 3.924 MHz with BPL transfers active (352KBytes)
• 15 MHz WWV in BPL test area (345KBytes)
• 15 MHz WWV 2 blocks away from BPL test area (340KBytes)