Re: [Discuss-gnuradio] power amplifiers on TX

[Daniel Pocock]

On 30/12/15 22:24, Marcus Müller wrote:
> Hi Daniel,
> Cannot stress this enough:
> Don't try to do everything to the max right from the start. Sure, 100mW
> is a lot less than what can do in the licensed bands, but then again,
> not coming from an amateur background, 120W right out scare me. Please

Maybe some clarification is in order:

- need to check if the power amp's output is proportionate to the input
power

- does the USRP and GNU Radio provide a way to manage output power
(power into the amp), or is it always constant at 100 mW?

- it is legal - e.g. 400W is the limit in the UK and regs are similar in
other countries for fully licensed hams:
http://www.ofcom.org.uk/static/archive/ra/publication/ra_info/br68r11/br68.htm

> make sure no more than -15dBm are fed into the USRP RX. So at an output
> power of 120 W ~= 51dBm, you need isolation of at least 66dB between
> your TX antenna and the RX port of your USRP for the TX frequency. Also,
> considering you're buying a device that can potentially cover 70MHz to
> 6GHz, spending lots of money on a powerful amplifier that can but
> operate on a few MHz really sounds like an unbalanced investment. Maybe
> reduce the output power (unless you want to do moon bounce, maybe), and
> get separate filters, just to keep the option of not operating in
> 144-147MHz; your whole operational range is much smaller than the amount
> of spectrum you can control with the USRP at once.
> 

Agreed - many people would be satisfied with something up to 50W,
similar to a mobile rig, this was just the first thing Google found when
I included "100 mW" in my search query.

I have had SAREX contacts with 5W from a handheld device and J-pole, but
that was in Australia where the population is very thinly distributed
and I could have been the only person transmitting.  There were no
nearby mountains (unlike my current QTH) and no tall buildings.  In a
densely populated region like Europe, more power may appeal to some people.

> Of you're not going to use two highly directive antennas for RX and TX
> to achieve isolation:
> You will either need an RX/TX switch (that you should definitely control
> using the USRPs GPIO pins, which can be programmed to certain states
> when transmitting, receiving or doing full duplex) to isolate RX from TX
> when transmitting, or an extremely expensive circulator-based device.
> 
> To be honest, you seem to be throwing money at your problem, that partly
> including having a bit of uncertainty what you want to do. I'd rather
> start small, buy the USRP, and maybe a preselective filter, experiment a
> bit with "short links", then invest in antennas, filters, switches and
> or amplifiers as you build up wishes and experience.
> My feeling is that you'd probably want to first RX only on different
> bands with different antennas, see what's possible with and without
> preselection, then decide on one or two bands, get more specific filters
> for these, switches, and then an amp.
> Don't start with something that you can use to defrost your antenna, and
> then try to figure out where you fried which parts of your rig.

Thanks for this feedback, I wasn't going to go out and buy this first, I
just wanted to start putting together a vision of what the big picture
looks like and what all the pieces will cost.

Buying an amp that can fry an egg is a poor substitute for having the
right antenna.  On the receive side, this fact is obviously even more
unavoidable.

Regards,

Daniel