SMD component fixture for antenna matching prototyping

I am currently designing an antenna that exhibits a slight capacitative reactance (54 ohm -44j)
Since I have crude prototyping tools (a VNA and nothing more than the toner transfer method to make boards at home), I was researching a way to test and replace SMD components on the matching PCB on the fly

The matching network uses a differential output. I need to be able to replace two capacitors and one inductor.
I wondered if, by using vias with pins soldered into them, I could squeeze the SMD components between the two pins to test and reuse the board for other projects, (standard 50 ohm impedance and always the same f0 for all my projects)

I don't think homemade prototype board strips would suffer many soldering cycles before breaking down.

I would use rather big SMDs (2012 or 2520 metric)
Is it a viable option ? or does any fixtures exist to clip SMD components in ?
What is the standard approach ?

Any help appreciated.

I would not recommend pressure-fit components, simply because there is a high chance of them coming loose leading to frustration.

I always solder to ensure proper contact.

Instead of perf boards etc, I suggest you use full clad copper with simplistic scratch-out isolation. FR4 board of course. This should be rugged enough. Your SMD components are likely to break long before the pcb does!

- - - Updated - - -

I would highly recommend using thru-hole ceramic caps instead of SMDs, simply because the end tags of smd caps tend to come off quite easily.

- - - Updated - - -

What frequency range are you working with?

rodv92 said:

What is the standard approach ?

Click to expand...

I just solder the SMD matching components. That's fast and easy.

Mainly 1.8 GHz and 2.4 GHz. I was thinking that SMD is the way to go for higher frequencies because of the parasitic effect of the through hole leads, is there a way to mitigate this ? like making a pit and not going "through the pcb" ?
thanks.

You may try wooden toothpicks with some glue. Component is glued to the toothpick and then placed to required points on PCB with slight pressure. Solder plated PCBs often have uneven surfaces, especially at the bottom side. Important PCB area and component pads may be polished with sandpapered and then covered with some flux and thin layer of solder using solder wick.

There exist antenna-tuning (matching) RFICs, I saw this going
on at the last "straight job" employer (although it was way
outside my activities & interests). Can't help you with selection
or even "who's who", but maybe the tip puts you on a path to
getting a user-tweakable in-situ match capability which you
can either leave in place, or substitute a fixed element when
you are happy. Believe most such critters will use a serial link
and a 3-5V supply (RF CMOS), so you might have to get more
hardware on the bench, but cut-and-try would sure benefit
from bits replacing soldering at least until you're feeling ready
for a deeper commitment.

The stuff I saw going past, was for low and high band cell
phones and noises were made about integrating into RF
switch IC products and front end modules.

dick_freebird said:

There exist antenna-tuning (matching) RFICs, I saw this going
on at the last "straight job" employer (although it was way
outside my activities & interests). Can't help you with selection
or even "who's who", but maybe the tip puts you on a path to
getting a user-tweakable in-situ match capability which you
can either leave in place, or substitute a fixed element when
you are happy. Believe most such critters will use a serial link
and a 3-5V supply (RF CMOS), so you might have to get more
hardware on the bench, but cut-and-try would sure benefit
from bits replacing soldering at least until you're feeling ready
for a deeper commitment.

The stuff I saw going past, was for low and high band cell
phones and noises were made about integrating into RF
switch IC products and front end modules.

Click to expand...

This is an excellent suggestion !! ! I wish they had these available when we were young experimenters.

@dick_freebird ... pleas PLEASe refer some mfgs and typical components !!

From what I have seen in the past, these antenna tuner chips (e.g. Qualcomm) are essentially switched capacitor banks. So this doesn't help with the element where component value might differ most from from the precalculated ideal values: the inductor.

As mentioned above, I don't really see the benefit of the fixture approach. After measuring the input impedance, you can do a calculation of the matching circuit, and with proper layout technique the values for C will be close. For L, you might need to go a bit smaller in nominal value due to increase of value towards SRF, but if you use manufacturer S-parameter models you can usually capture that effect well. So overall, there is not much need for swapping components.

With a fixture, you might just introduce parasitics from the fixture, which might again change things. I think that's not a smart approach at GHz frequencies.