Using microstrip "rectangular spiral" distributed inductors.

[Darktrax]

For when the lumped component SMD inductor will no longer be good enough (at X-Band).

Having to use the microstrip spiral inductor becomes necessary because of the self-resonance and other characteristics that SMD components have. I am working with 0.508mm (20 mils) thick PTFE/ceramic/glass type PCB for X-Band project. Passive components are mostly 0402 size.

Implementing the "square spiral" requires a track to escape from the middle, either using a "wire bridge" or similar conductor tab, or by using two vias, and a short track on the underside, which seems the obvious way, but somewhat inconvenient because of the need to machine a clearance into the case beneath. There may be other effects I have not yet seen.

Is there available a SMD type component to be the "bridge", maybe obtainable in various lengths, that can be used on the topside, to jump over the tracks of the printed inductor, while keeping the conductor suitably separated from the tracks underneath?
I have looked, but not seen any so far. Maybe I am using the wrong search keywords.

 

[thylacine1975]

Heya DarkTrax - here's some RF crossovers I've contemplated (but never ended up needing) before... they top out at 6GHz though (so not really much help at X band - perhaps they'll help start a more fruitful search...).

https://www.anaren.com/products/rf-crossovers
https://www.richardsonrfpd.com/Pages/Product-End-Category.aspx?productCategory=10179

Are spiral inductors useful as high as X band? I'd always resorted to narrowband transmission line structures long before then - luckily I've never needed to try anything broadband and attempt anything new!

 

[Darktrax]

Are spiral inductors useful as high as X band? I'd always resorted to narrowband transmission line structures long before then - luckily I've never needed to try anything broadband and attempt anything new!

Thanks much thylacine.
I would not burden you with my complications, but will explain the motive..
All the microstrip for actual wanted band is done using the usual stubs, coupled lines , etc.

Unfortunately, devices useful at X-Band usually have gain that gets enormous at lower frequencies, right down to DC! Stability at unwanted frequencies is the problem. Unstable behaviour at frequencies far higher out of band also to be dealt with.

The feedback circuits needed to deal with this, while at the same time becoming "non-components" as we get to frequencies where they cannot work, require higher values of inductance. It may be that a lumped inductor will work here, dealing with unwanted gain at audio and HF, and cease to operate at X-Band, but the model and the parasitics are complicated and awkward.

That is why I thought a printed spiral, (or two, or three) might be a way to keep it predictable, and hence the need for bits I have never used before.

 

[johnjoe]

Depending on your inductors value, Murata has some with self resonance frequency up to 20GHz.

 

[Darktrax]

Depending on your inductors value, Murata has some with self resonance frequency up to 20GHz.

Thanks johnjoe - I will be checking out the vendors.

Regarding self-resonance. Should we actually regard it as "self roll-off" instead?
"Self-resonance" kind of implies that once you get above the self-resonance, all is well again, when actually, the frequency response curves usually show the inductance falling away.. getting less, as frequency increases beyond the self-resonance.

I just assume that one has to use parts where the self-resonance is well above the design frequency.

 

[johnjoe]

What do you mean with "self roll-off"? Beyond the self resonance the inductor "acts" like a capacitor.