Need RF sinusoidal signal with DC component

[STOIKOV]

for testing purposes, I need an RF signal mounted in a DC component in order to apply it at the input of an amplifier.
Is it possible to connect in series the RF generator (about 900MHz) with a DC power supply (2.5V) to obtain the desired wave form ? no matter what the frequency is or a impedances coupling issues ?
RF generators can provide 2.5V as offset ?

 

[rfmw]

"Is it possible to connect in series the RF generator (about 900MHz) with a DC power supply (2.5V) to obtain the desired wave form ?"

I hope that you meant it differently from what you've written...


If you only have an AC-coupled signal generator, you need a bias-tee to superimpose the RF signal on the DC signal. For 900MHz you can easy make a bias-tee, since there are many chokes appropriate for this frequency range, or you can even use just a resistor instead of choke.

 

[STOIKOV]

Thanks for your reply.
How is a bias-tee, chokes, or resistor implemented for this purpose ?
The circuit I'm working on is an integrated mixer, so in simulation I set a sinusoidal source with DC component

VIN 1 0 SIN (2.5 0.05 900MEG)

where 2.5 is a DC value and 0.05 is the amplitude; amd want to implement this wave form in laboratory.
(see picture of input signals in amplifiers)

 

[rfmw]

Look at these commercial bias-tees:

**broken link removed**

and bellow is a photo of a wideband home-made bias-tee (500kHz - 3GHz). Sorry about the bad photo, but there is a FR-4 PCB with 50 Ohm microstrip line with 100pF||100nF in series and 3u3 choke with another choke (with parallel resistor) in series. Note the flexible coax cables soldered to the bottom groundplane of the microstrip with the help of soldered brass pieces...

 

[STOIKOV]

Thanks, I saw the Bias-Tees file and understood an AC and DC signal can be combined.
Do you know if these Bias-Tees can be used with microprobes for on-wafer measurements ??

 

[rfmw]

I don't see why not. Perhaps others, more experienced with that kind of measurement setups, would explain the reasons not doing so. What comes to mind is that bias-tee affects the S11 and S21, so this should be taken into account with the measured results...