Hacker News Reader

Comment by TaylorAlexander | Hacker News Reader

Hah, thanks for the edit. Yeah we've already gotten the range. I've even checked the output of our boards with a radio expert (google Earl McCune, he's an advisor), and using his many expensive Agilent scopes we've experimentally verified that performance of my design is pretty much exactly on par with the reference designs, taking into account that we are using a custom designed RF balun [1] instead of the discretes they use on the reference design. Our RF Expert confirmed I had made a good choice with that, as the losses are tiny compared to ease of manufacturing it provides.

[1] (I think it's this one, I'd have to check the part number, they have a couple similar ones). http://www.johansontechnology.com/datasheets/chipset-specifi...

View on HN

Are you sure you're able to legally transmit at 1km. Had thought antenna gain + transmit power had to be around 0 dBm, at least for 434MHz.

Hadn't seen the rules for 900MHz though.

I'm pretty sure that to get 1 Watt you have to use direct sequence or frequency-hopping spread spectrum under FCC part 15.247... Just sitting on a frequency as a 15.249 device or general ISM device you aren't allowed to use much power at all.

See FCC part 15.247: http://www.gpo.gov/fdsys/pkg/CFR-2001-title47-vol1/pdf/CFR-2...

I work with UHF Gen2 RFID readers all day which do 1 Watt/30 dBm into 6 dBi = 4 Watts EIRP, but they are definitely required to do hopping.

EDIT: and after looking at the 50 mW 900 MHz XBee units, they are doing hopping as well according to their datasheet.

So it looks like if you want to keep the range, you may want to start looking at hopping algorithms...

From my reading of the rules, you are limited to broadcasting +36 dBm EIRP (30 dBm [1W] into a 6 dBi antenna -> 4W EIRP). This can be a single frequency. In the 900 MHz band, your bandwidth is limited to 902-928 MHz (US limitations, different in other global regions).

You can happily beam out a 915 MHz single-tone CW carrier at +36dBm EIRP all day and that is fine.

The reason your gen2 reader is frequency hopping is not FCC mandated but protocol mandated. The UHF RFID gen2 tags have a high manufacturing variation and the RF frequency response varies greatly between tags. Some tags may be resonant closer to 902 MHz and some tags may be resonant closer to 928 MHz. Hopping over various frequencies allows the reader to address all tags in its view.

The only reason the readers are "required" to do hopping is to conform to the gen 2 protocol. In fact, from what I remember you can use the LLRP (low-level reader protocol) to stuff the frequency hopping table with a single frequency so that it will stop frequency hopping.

(My PhD was on high-data rate (up to 100 Mbps) rfid/backscatter communication.)

This is what I initially thought as well, having built an ISM band transceiver around the 350/400 MHz bands a few years ago. I recalled much lower power limits.

(I had posted something along the lines of "there's no way in hell the FCC would let you transmit at half a watt" but have since edited).

Apparently the 900 MHz bands are different, and the FCC has rules governing output power going into the antenna (1 watt), as well as an effective radiated power (which essentially places limits on your antenna gain) (4 watts).

[1]: http://www.afar.net/tutorials/fcc-rules/

Wow that is very much radiated power!

I remember range testing a 434MHz product and being hundreds of meters away (line-of-site). The product passed FCC testing.

So a watt going into a 6dBi antenna... Flutter's measurements make sense now!

First thing that crossed my mind too. It's definitely forbidden in Europe.

Yup, but we can 868MHz in Europe, and with all the support from international backers I think we may end up doing that as a stretch goal! Just have to talk it our with our RF expert, which we're doing!

The Xbee Pro 900 is 50mW, for reference.

Reply on news.ycombinator.com