2 GHz 180 degrees phase shifter circuit

[kurtulmehtap]

Dear All,
We have to design a 2 GHz 180 degrees phase shifter circuit. Are there any transistor circuits? Any pointers?
Thanks in advance

 

[jiripolivka]

I can see you have no knowledge about the problem.
2 GHz corresponds to 15 cm wave length, so your device should achieve 180 degrees phase shift by adding or removing 7.5 cm to the transmission line.
You can do it mechanically by making a coaxial line extender, or, by switching line sections . You have not specified if your shifter should work in steps and how many steps you need to achieve 7.5 cm.
Typically electronic phase shifters use short line sections switched by PIN diodes and their drivers can use transistors or ICs.
Check e.g. MiniCircuits products, they show commercial products as well as driver schematics needed.

 

[kurtulmehtap]

I can see you have no knowledge about the problem.
2 GHz corresponds to 15 cm wave length, so your device should achieve 180 degrees phase shift by adding or removing 7.5 cm to the transmission line.
You can do it mechanically by making a coaxial line extender, or, by switching line sections . You have not specified if your shifter should work in steps and how many steps you need to achieve 7.5 cm.
Typically electronic phase shifters use short line sections switched by PIN diodes and their drivers can use transistors or ICs.
Check e.g. MiniCircuits products, they show commercial products as well as driver schematics needed.

Thanks for the answer, we are aware of adding/removing 7.5 cm transmission line to get the required phase shift.
The problem is 7.5 cm is a lot on the PCB and the error in the desired phase cannot be minimized by this method.
Is there an alternative method? We checked out the Minicircuit device but it has only variable phase shift, not constant.
We would prefer a transistor based solution so that we can also port the circuite into an ASIC.

 

[thylacine1975]

If you're targeting an ASIC, are you seeking a logic device or an analog solution?
(Here I use the word "logic" to mean you care about bi-phase clocks, setup/hold times, bus timing etc., and "analog" to mean radios, mixers, LNAs, filters etc.)

If it's logic you're after, then will something along the lines of an inverter realised from a single (ASIC-able) FET meet your needs? Perhaps the complementary output of your clock buffer/driver? Maybe via a logic family featuring differential signalling (PECL, LVDS etc) driving a single-ended buffer?

If it's analog, then transmission lines, miniature transformers, hybrid/branchline/ratrace couplers, lumped LC/RC/RL techniques would be your friends.

 

[vfone]

Same length of various transmission lines like: stripline, microstrip, coaxial, waveguide, twisted-pair, etc, would give totally different phase shift.
On top of this, the same length transmission line made on the same approach (stripline, microstrip, coaxial, etc) would give different phase shift if they use different dielectric materials.

 

[volker@muehlhaus]

We would prefer a transistor based solution so that we can also port the circuite into an ASIC.

For RFIC, circuits are often differential. Then, 180° phaseshift = reverse polarity is easily available.

 

[kurtulmehtap]

For RFIC, circuits are often differential. Then, 180° phaseshift = reverse polarity is easily available.

We Investigated some differential clock drivers, buffers. Their problem was the duty cycle of the inverted output. For a %50 duty cyle, the output can vary between %45 to %55 which is an unwanted feature.

 

[volker@muehlhaus]

We Investigated some differential clock drivers, buffers. Their problem was the duty cycle of the inverted output. For a %50 duty cyle, the output can vary between %45 to %55 which is an unwanted feature.

I was talking about analog RFIC (full custom design). It will help if you can specify your application/requirements/restrictions in more detail ...

 

[biff44]

a center tapped transformer will give you 0 and 180 degree bits, then use a non-reflective spdt switch to choose the output. Choose the windings so that it is 50 ohm at all three terminals.

other than that, you can do a shifman phase shifter in microstrip

a double balanced mixer with a dc coupled IF port will give you 0 / 180 phase shift as you bias the IF port +/- 5 mA, but its going to have 6 dB of loss