kif.rocks ist einer von vielen unabhängigen Mastodon-Servern, mit dem du dich im Fediverse beteiligen kannst.

Serverstatistik:

153
aktive Profile

blue_led

Does anyone have open-source layouts for 2-port series and shunt component test PCBs using SMA connectors? Plus points if it's made for one of the JLCPCB stackups.
Something like the SV1AFN VNA test PCBs but as an open-source design that I can tweak a bit for different components. Also the SV1AFN design is shunt only.

Of course I could design my own and that's a great exercise but
a) there are many different aspects I'm still learning about and having a known-good design let's me concentrate on some of them
b) quite a bit of tweaking can go into getting the impedance flat and I'd prefer not to waste PCBs for things others have done before. I do that enough for other aspects I'm trying to learn about. 😅

@electronic_eel @blue_led I don't have any for SMT chip passives, only other devices.

The de-embedding is nontrivial especially if you want to remove the footprint effects so you can generalize component footprints to other stackups.

Getting S-params of the component and footprint, de-embedded to the edge of the footprint, is straightforward with a 2x thru standard and the IEEE 370 reference code (or Amphenol's GUI based on it).

@azonenberg @electronic_eel
I don't *think* I need the measurement result to be transferable to other stackups. But maybe I should try two different stackups just to get a feeling for how much difference it will make.
Thanks for the pointer to IEEE P370! I was looking at scikit-rf 2x-thru de-embedding Split* / *Cancel before. The IEEE P370 support looks straightforward to use and there's exactly one obvious choice for my case (IEEEP370_SE_NZC_2xthru) so I'll go with that instead.

@electronic_eel @azonenberg
Thanks for the pointer. FWIW the 4-layer OSH park stackup is quite different from those offered by JLCPCB (different prepreg/core thickness, dk 3.61 vs. 4.05/4.25/4.6) so I guess I'd need to order from OSH park or adjust all dimensions. (TBH I didn't expect their stackups to be *that* much different).

@blue_led @electronic_eel Yeah I like the FR408HR a lot. It's a near exact drop in equivalent for RO4350B except for the higher Df, so you can prototype RF stuff then move to Rogers and get the same layout working the same way but less loss.