under unlicensed 2.4GHz FCC part 15 for digital modulation
system(non-spread spectrum), it says that the
1. maximum conducted power is 1W
2. the spectral density be not greater than 8dB for every 3khz band
during continous transmission
3. the 6-dB bandwith is at least 500khz

question:
1. what is spectral density in this sense and its significance?
2. how can i measure my spectral density?
3. how does 1, 2 and 3 relate to each other?

I'm not going to bother trying to look it up--but is that not supposed
to be something like 8dBm/3kHz BW, or some such? 8dB without a
reference doesn't mean much.

If it's 8dBm/3kHz, then the meaning should be obvious: in any 3kHz
band there may not be more than +8dBm power (I guess around 7
milliwatts). Spectral density is commonly measured with a spectrum
analyzer; many modern ones have band markers that will let the
instrument perform a band power measurement.

If you spread 1 watt uniformly over 500kHz, you'd have 1W*3kHz/500kHz,
or 6mW, in each 3kHz bandwidth. So the +8dBm limit would mean that the
power should be spread very nearly uniformly over that band. (Why did
they use units of watts in one place and dBm in another??)


Cheers,
Tom

dBm = watts = 0.001 watt as a reference

K7ITM wrote:

I'm not going to bother trying to look it up--but is that not supposed
to be something like 8dBm/3kHz BW, or some such? 8dB without a
reference doesn't mean much.

If it's 8dBm/3kHz, then the meaning should be obvious: in any 3kHz
band there may not be more than +8dBm power (I guess around 7
milliwatts). Spectral density is commonly measured with a spectrum
analyzer; many modern ones have band markers that will let the
instrument perform a band power measurement.

If you spread 1 watt uniformly over 500kHz, you'd have 1W*3kHz/500kHz,
or 6mW, in each 3kHz bandwidth. So the +8dBm limit would mean that the
power should be spread very nearly uniformly over that band. (Why did
they use units of watts in one place and dBm in another??)


Cheers,
Tom

hmmm..thanks

all i though from the first rule that i can transmit at 1W continously
for all applicable frequencies...you cleared that nicely....thanks

in other words, when i have a transmit modulation bandwith of 500kbps
my peak power level must be 8 dBm...right? what if i use 6khz or 5khz?
thanks

If your modulation spreads the power evenly over 500kHz, if I
understand the rules right, you can transmit 1 watt. If your
modulation spreads the power evenly over 50kHz, you could transmit 0.1
watts (+20dBm). If your modulation spreads the power evenly over 5kHz,
you could transmit 0.01 watts (+10dBm). It _appears_ that you could
transmit +8dBm (6.something milliwatts) however you want in that band:
as an unmodulated carrier, or spread however you want. I don't suppose
they talk about the time period over which you measure the spectral
power density...but they probably wouldn't like it if you transmitte 1
watt that swept linearly and slowly across the 500kHz, say once a
minute. Over a long enough period, that would be uniform power per
unit bandwidth, but a one second snapshot would show 1 watt in
500kHz/60 = 8.3kHz bandwidth.

Hope that makes sense. Beyond that--see your lawyer! ;-)

Cheers,
Tom

in other words, when i have a transmit modulation bandwith of 500kbps
my peak power level must be 8 dBm...right? what if i use 6khz or
5khz?does proportionality hold here or just say in any bandwith not
greater than 500khz, the peak power level should not always exceed
8dBm?..

thanks

On 2 May 2006 01:09:53 -0700, "electro"
wrote:

in other words, when i have a transmit modulation bandwith of 500kbps
my peak power level must be 8 dBm...right?

No - that is not correct, in my opinion. My interpretation is that the
spectral power density in a 500 Hz bandwidth must not exceed the limit
of 2 milliwatts per Hz. (Simply spread the 1 watt over a width of 500
Hz.
In my original post I thought your message said 500Hz. Now I see you
are saying 500kHz. Which is correct?

what if i use 6khz or
5khz?does proportionality hold here or just say in any bandwith not
greater than 500khz, the peak power level should not always exceed
8dBm?..

The 8 dBm is a red-herring. I took your value of 8 dB as Gospel. If
someone says 8 dB they don't mean 8 dBm and vice versa.

Can you make sure of the accuracy of your excerpts, because I see
nothing in the Part 15 regs that correlate with your numbers.

In other words, I'm working with nothing but your excerpts and common
sense electrical engineering.

On 1 May 2006 06:47:45 -0700, "electro"
wrote:

under unlicensed 2.4GHz FCC part 15 for digital modulation
system(non-spread spectrum), it says that the
1. maximum conducted power is 1W
2. the spectral density be not greater than 8dB for every 3khz band
during continous transmission
3. the 6-dB bandwith is at least 500khz

question:
1. what is spectral density in this sense and its significance?
2. how can i measure my spectral density?
3. how does 1, 2 and 3 relate to each other?

This is just a guess, but let's see what we can do to make some sense
out of this.

If you were hoping that the 1 watt of power would be evenly
distributed across the 3 kHz bandwidth, that would be for the purpose
of making sure that no one would be putting all of his 1 watt into a
single frequency, which would be quite a strong concentration of power
in comparison to having it spread out over the 3 kHz. The idea being
that the power when concentrated is very noticeable and might cause
QRM, whereas the same wattage spread out evenly over 3 kHz would
barely exceed the background noise and would not be noticeable.

So, you might want to require that signals not be very "peaky", or
concentrated. In order to put limitations on "peakiness", you might
require it to be "at least so wide" and "no higher than x" in units of
spectral density.

Assume that in the above requirement that the 500 Hz bandwidth
requirement is to be interpreted as their requirement that avoids the
signal being concentrated in a narrow band of frequencies. Also assume
that the maximum power density in that 500 Hz bandwidth should be
commensurate with the entire 1 watt being distributed over only 500 Hz
instead of 3 kHz. Notice how we have traded off the requirements so
that they are being applied to 1/6th of the 3 kHz. If we are allowed
to concentrate all of our power in 1/6th of the bandwidth, then it
follows that we would be required to keep the power density in that
bandwidth at no higher than 6X the level it would have when spread out
over the entire 3 kHz.

I find that it is probably not a coincidence that the 6:1 ratio is
almost exactly 8 dB in power ratio.

thanks!