Microwave / RF behavior of SMT component hold down adhesives ?

[EE-digger]

** sorry for dupe, cannot fix at the moment **

To replace an inadequate interconnect method between two tiny assemblies, I need to place a U.FL connector on an existing PCB, it will have an excellent gold surround and signal to solder to, but no footprint since it's an afterthought.  Solder paste is auto deposited and pick and place no problem, but I need to restrict movement during reflow.  Frequencies are all under 3GHz.  Signal is a 50 ohm path.

Does anyone have experience with the high frequency / dielectric / loss tan characteristics of the adhesives used to hold components down (when needed) for reflow?  Any experience with particular brand or product?

Thanks !

[Andrey_irk]

The best way to get the loss tanget of a material is to ask the manufacturer. But, you may not need it.
What laminate do you use for your PCB? If it is FR4, then the adhesive will hardly make any differrence.
The lack of a proper footprint can harm your VSWR much more than the presence of the adhesive. And, what parameters do you want to get from that interconnect?
Still, my rough estimate will be Eps = 3 ... 4 and tanD about 0.05 (maybe more, but not muchl). Based on my experience.

[tszaboo]

The best way to get the loss tanget of a material is to ask the manufacturer. But, you may not need it.
What laminate do you use for your PCB? If it is FR4, then the adhesive will hardly make any differrence.
The lack of a proper footprint can harm your VSWR much more than the presence of the adhesive. And, what parameters do you want to get from that interconnect?
Still, my rough estimate will be Eps = 3 ... 4 and tanD about 0.05 (maybe more, but not muchl). Based on my experience.

They will not have this info, they never do. When you ask them for dielectric constant, you get 50Hz, or if they are really scientific, 10KHz.

[EE-digger]

The best way to get the loss tanget of a material is to ask the manufacturer. But, you may not need it.
What laminate do you use for your PCB? If it is FR4, then the adhesive will hardly make any differrence.
The lack of a proper footprint can harm your VSWR much more than the presence of the adhesive. And, what parameters do you want to get from that interconnect?
Still, my rough estimate will be Eps = 3 ... 4 and tanD about 0.05 (maybe more, but not muchl). Based on my experience.

They will not have this info, they never do. When you ask them for dielectric constant, you get 50Hz, or if they are really scientific, 10KHz.

Material is FR4.  The plan is to eliminate 3 very bad serial inductances caused by cantilever spring contacts.

True, they seldom have this info, always 50/60Hz related.  Got lucky once while potting a BLE antenna for a medical device.  Mfgr still had no numbers but a long successful track record with their compounds.  Took a few extra iterations but worked out well.

Looking at the jacks some more, I think two diagonal placed dots can grab it without touching the signal pin.  Plus I have low impedance of the signal on my side.

ATEX, heh?  I've designed two small products to meet it.  Can be a challenge when you need a 0402 size part and the ATEX rated version is a half inch/cm cube    What can be more of a challenge is labeling it properly.

[Andrey_irk]

Can't say about 50/60Hz, but the dielectric constant value measured at 1MHz (which is pretty much standard) is usually very close to that at 3GHz. If it isn't, then the loss at 3GHz will be over the roof (in other words, huge loss affects epsilon value).
FR4 itself is very lossy, so you will unlikely notice any difference with the adhesive.