I have a rf signal generator (50 ohm, 3.2mw max power, going to be transmitting 13.56mhz. It's a HackRF SDR) , which I want to connect to a homebrew RF power amp. As a newbie to RF design, I know that whatever I do, I'm going to let some magic smoke out on my first try building this amp. In this case, I'd rather that the smoke not be emitted from my $400 RF gen. Thus, I'd like to somehow isolate the output of my RF source from the input of my rf amp, so that I can tinker around with the amp (adjust base currents to the amp transistors, for instance) without being nervous. Keeping the same impedance would be a plus. Is there some way to do this?
2\$\begingroup\$ Signal generators in general don't like DC so connecting the signal to your amplifier via a capacitor (100 nF ceramic type for example) to block any DC from getting back to the generator would be a good start I guess. \$\endgroup\$– BimpelrekkieJul 16, 2015 at 12:46
\$\begingroup\$ @IC_designer_Rimpelbekkie Ok, I'll keep that in mind, thanks! \$\endgroup\$– 0xDBFB7Jul 16, 2015 at 12:50
\$\begingroup\$ Capacitive coupling with a simple FET follower behind should be high-enough impedance to match on the source side of the capacitor. You can also match a HF 1:1 signal transformer to 50Ohm load using maths and capacitors. \$\endgroup\$– AsmyldofJul 16, 2015 at 12:52
There are a couple of things that you can do.
1) insert a low-power buffer amp between the output of the signal generator and the power amp. Quite frankly, you are going to need multiple amplifier stages anyway. So make the first amp stage function as a buffer.
Have a look at Mini Circuits for really easy-to-use single-chip solutions.
2) Add a couple of big-ass diodes back-to-back across the output of the signal generator. 3.2 mW is about 0.4V P-P, so a pair of silicon diodes connected back-to-back from the signal pin to ground should protect the generator should something bad come back towards it.
Be sure to choose diodes that have low capacitance. I know that it can be hard to find the combination of low capacitance and high-current capability but I know they exist. Unfortunately, I don't recall where I saw them.