1
\$\begingroup\$

I am trying to find/understand how to invert a RF signal at around 5 GHz, i.e. go from a 5 V to 0 V signal to a 0 V to (-5 V). I know this is possible to do with an op-amp on slower circuits, but all the amplifiers I have found have all just been straight power amps, not an inverting op-amp. Is there any good way to do this, or some simple way I have not understood?

(PS the exact signal change I want is 0 V to 0.8 V SDI-3G signal to -1.5 to 0 V signal if that helps.)

\$\endgroup\$
5
  • \$\begingroup\$ You will get a better answer if you explain your application. Is this signal a logic signal or an analog signal? If the latter what is the shape ( sinusoid, square, ec.) . Do you need a true inversion of the signal or just a DC level shift? \$\endgroup\$ Commented Jun 26, 2025 at 22:29
  • 7
    \$\begingroup\$ That is NOT an inversion but a level shift. If you think differently then please explain why. \$\endgroup\$ Commented Jun 26, 2025 at 22:36
  • \$\begingroup\$ you can still get opamps that work fine at 5GHz, just expect to pay $100 for them... \$\endgroup\$ Commented Jun 27, 2025 at 2:23
  • 1
    \$\begingroup\$ The title is very misleading, as can be seen from the answers. There's no RF signal, it's not 5 GHz, you're not inverting it. \$\endgroup\$ Commented Jun 27, 2025 at 7:41
  • 2
    \$\begingroup\$ More importantly, what is the spectral content of the signal (is DC included?), what is the bit rate, and what is the impedance? Concentrating on voltage is not the way to go here: the voltage might be higher or lower, it's the impedance that determines whether that requires an amplifier (or other active circuitry), or a passive solution (matching network, transformer, bias tee, resistors). What you're doing is most important: what is your signal source (SDI-3G apparently; including a spec sheet will be helpful), and what is its destination (receiver)? There is probably a standard solution. \$\endgroup\$ Commented Jun 27, 2025 at 10:27

2 Answers 2

5
\$\begingroup\$

IF all you need to do is level shift a 0 to 5V signal down to -5 to 0V AND (very important) your signal truly IS RF (in that it has no significant DC component), then you may be able to use an AC coupling capacitor and terminate your load to -5V instead of the more common 0V like in the image below. This is commonly used in a lot of high speed digital stuff (PCIe, USB3 etc.) so that the TX and RX ends can sit at their preferred bias voltages.

AC coupled load referenced to -5V

You can test out this circuit using the Falstad circuit simulator

N.B. to get this to work reliably, you WILL need some decent RF-rated caps in 0402 packages or smaller (ideally 0201) to avoid notable parasitic inductance problems, along with good PCB transmission line and reference plane design (unless it's all coax, in which case you can just buy a DC block module off the shelf).

P.S. ”inversion” of a signal normally means "flip the HIGH and LOW states without changing the voltage range of the signal" - i.e. it's still swinging between 0V and +5V

\$\endgroup\$
0
\$\begingroup\$

At 5 GHz, a transformer is a few loops of wire. Diode-clamp or bias the secondary so it doesn't go above ground, and that's a solution. Source and load impedance matching is another problem.

\$\endgroup\$

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.