6

I've recently been given a project to lay out some PCBs which will contain a few Mini-Circuits RF amplifiers, and their associated biasing components. However, since I need to have several stages of amplification, I need inter-stage AC coupling to prevent the bias from each amplifier affecting the other amplifiers in the chain.

That said, I basically therefore need to chose series capacitors for coupling signals in the ~1.420 GHz range, and I do not really have the experience to know what I should specifically be looking for in terms of specifications.

I know I need capacitors with high Q, and probably X7R or NP0, ideally low insertion losses (S21), but I'm really feeling my lack of experience and knowledge in RF when digging through datasheets.

What are the relevant considerations for choosing parts for this kind of application?

JYelton
  • 32,302
  • 33
  • 134
  • 249
Connor Wolf
  • 31,938
  • 6
  • 77
  • 137
  • 1
    You don't need this for 1.4 GHz, but if you really want to go crazy you can try [these](http://www.atceramics.com/series.aspx?sid=40&series=550L-Series) – The Photon Jun 25 '14 at 01:10
  • @ThePhoton - While this project isn't too high frequency, there are several project that *do* use extremely high frequencies around in the lab where I work (8, 10 Ghz microwave telescopes). Those are quite interesting! And digikey even carries them! – Connor Wolf Jul 02 '14 at 00:30
  • The documentation for those capacitors is also fascinating. They discuss silver epoxy, and using indium solder. – Connor Wolf Jul 02 '14 at 00:31
  • if your application is classic narrowband rf you probably don't need the huge capacitance values, so you still don't need those fancy parts. If you're doing 10-25 Gb/s baseband digital, they're very useful. – The Photon Jul 02 '14 at 01:34
  • @ThePhoton - In this case, I'd be making scientific instruments. We'll make maybe 2 or 3 complete systems, so the marginal cost increase isn't worth worrying about. – Connor Wolf Jul 02 '14 at 01:48
  • See also: https://electronics.stackexchange.com/a/568035/256265 – KJ7LNW Jun 08 '21 at 22:44

2 Answers2

4

You should be looking for capacitors that exhibit series tuning resonance well above the frequency of interest - the self inductance of the capacitor twinned with its capacitance cause a sharp series resonance. Take a look at this graph from AVX for their MLCC NP0 range in 0805 size: -

enter image description here

More than likely the 10pF will exhibit resonance around the 2 GHz mark. I'd say you either make use of this (because, as an AC coupler it'll still block DC just fine) or find a better grade but for sure you are going to find a lot of caps resonant in the GHz region you want to operate at.

Neil_UK
  • 158,152
  • 3
  • 173
  • 387
Andy aka
  • 434,556
  • 28
  • 351
  • 777
  • 1
    The main thing I'd add to this for future readers is, generally, smaller packages have lower parasitic inductance, so an 0402 will have higher resonant frequency than an 0805. – The Photon Jun 24 '14 at 20:38
  • @ThePhoton very true – Andy aka Jun 24 '14 at 21:04
  • Some manufacturers (including Murata and AVX) also make low-inductance "reversed" packages such as 0204 (rather than 0204). These packages have the electrodes on the long side. This makes the current path shorter and fatter. – Evan Nov 21 '16 at 18:55
0

For the best in microwave capacitors try single layer capacitors from dilabs http://www.knowlescapacitors.com/dilabs/en/gn/products/single-layer-capacitors

Well you did ask about microwave capacitors !

N.G. near
  • 435
  • 2
  • 10
  • Is that intended as a serious answer? It flags may spam sensor, but I don't think it is. I don't think it provides any information though. – Sean Houlihane Sep 24 '16 at 12:05