Can someone help me? I'd like to come up with a freq doubler that will
do 1 GHz to 2 GHz (from a 500MHz to 1Ghz source). I thought of using a
mixer and simply tieing the RF and LO and getting my out on the IF, but
all the spec's for all the parts I've seen are assuming downconversion,
and the RF & IF is in the multi-GHz ranges, while the IF is below 1.5
GHz at best. I also seem to remember reading somewhere, that you can
also use the IF port as the RF port, and get the result from the
(original) RF port. Will this work? I'm working with about a 1-5 dbm
signal (I guess I can through in an mmic if I need more power, but I'd
like to keep the circuit small).
I haven't quite been down the road of tripling yet, as I'd like to
keep the error multiplication down (2 is better than 3).

-mikal

In article . com,
Mikal wrote:

Can someone help me? I'd like to come up with a freq doubler that will
do 1 GHz to 2 GHz (from a 500MHz to 1Ghz source). I thought of using a
mixer and simply tieing the RF and LO and getting my out on the IF, but
all the spec's for all the parts I've seen are assuming downconversion,
and the RF & IF is in the multi-GHz ranges, while the IF is below 1.5
GHz at best. I also seem to remember reading somewhere, that you can
also use the IF port as the RF port, and get the result from the
(original) RF port. Will this work? I'm working with about a 1-5 dbm
signal (I guess I can through in an mmic if I need more power, but I'd
like to keep the circuit small).
I haven't quite been down the road of tripling yet, as I'd like to
keep the error multiplication down (2 is better than 3).

-mikal


http://www.minicircuits.com/dg03-204.pdf

Here's their line of doublers. All involve 12-15dB of conversion loss,
and want input signals on the order of 10-15 dBm input levels, so
you're going to have to boost the signal before the multiplier as well
as after (as well as doing some filtering to remove unwanted f1, f3,
and f4 products).

The AK-3000 and RK-3000 doublers are thru-hole packages with 70-1500
MHz in, 140-3000 MHz out. They want 12-15dBm input power, and have a
10-16dB conversion loss depending on frequency, phase of the moon, and
other characteristics. They run $60.

The MK-5 is 10-1000 in, 20-2000 out with 10-20 dBm input power and
13-17 dB conversion loss. It's in a fancy little box with SMA input
and output, yours for only $77.

You can find a lot of mini-circuits stuff on eBay, but I don't see the
doublers there very often.

--
Namaste--

it appears as though I can use the solderable (what I was looking for)
KC2-11. But, that's ALOT of loss during the conversion. anyone else
out there know of a better 'wideband' solution? Should I triple?

Can you make (or buy) a broadband transformer with a center-tapped
secondary, and use a pair of Schottky diodes arranged just like a
full-wave rectifier? At your low power level, I suppose you'll want at
least a 1:1:1 transformer ratio, and 1:2:2 likely will be better. Some
experiments or careful Spice simulations with good models should help
optimize the conversion efficiency, which won't be wonderful at those
low levels because of the drop in the diodes. Be sure to provide a DC
return path for the diodes or it's not going to work.

Since that balanced push-push circuit suppresses the fundamental and
third harmonics, you can get _nearly_ an octave of coverage with a
single set of good filters, but note that 2*500MHz = 1000MHz, and
4*500MHz will also be generated, yielding 2000MHz output at the same
time, which would be passed by your filter. Will that be a problem to
you?

Cheers,
Tom