Using flux on Rogers substrate - bad idea?

[tec5c]

I am trying to find useful information on whether using flux to assemble a Rogers RO3003 board for microwave applications is a bad idea or not.

My initial feeling is to avoid it, though I thought I would get some other opinions. 

[2N3055]

From Rogers own fabrication note for RO3000:

.."
Most final finishes (HASL, Sn, Ag, Ni/Au, OSP, etc...) have been applied to RO3000 materials without issue or
special concern. A rinse/bake regimen, if not done as part of a solder mask process, should be done prior to
HASL or refl ow exposures. When flux is needed, acid fluxes are recommended over solvent fluxes. The HASL
or reflow exposure should be performed as soon as possible after the flux has been applied. ENIG should be
considered for higher Er materials, especially RO3010 laminate, only when necessary. Background plating of
etched substrate surfaces can occur during ENIG plating.
"..

http://www.rogerscorp.com/documents/634/acs/Fabrication-Guidelines-for-RO3000-and-RO3200-Series-High-Frequency-Circuit-Materials.pdf
Hope that helps..

[G0HZU]

Most of my experience is with Rogers 4003C or 4350 and I can add some info wrt cleaning solvents. I know you were asking about flux but here is some info about cleaning solvents anyway...

I've used IPA on (milled) dev boards to clean them up and make them look presentable but there is a downside to this. The IPA gets into the laminate and it can take a very long time to fully dry out again. It can look dry on the surface but the insertion loss up in the GHz region can be affected by 0.5dB to 1dB. If you want to (accurately) measure insertion loss of a low loss filter or RF switch built on Rogers 4003 or 4350 then don't clean the board unless you can afford to wait a long time for the material to fully dry out.

It can take an hour or two even on a tiny test PCB for the insertion loss to recover to where it was before the board was cleaned with IPA. I normally work with test boards milled on a router and maybe this makes it easier for the IPA to get into the laminate as the surface has been grazed by the milling tools but cleaning with IPA can easily cause confusion wrt sudden changes in insertion loss!

[tec5c]

Thanks for the replies.

I think I may "go in dry" with this one 

I had previously read the article from rogerscorp, though it seemed to conflict with a few IEEE papers I have read on the same matter which evidently, seem to coincide with what G0HZU has said. So I thought I'd get some more opinions. 

[IconicPCB]

I have milled some rogers for a client in fact built 3D structures by laminating multilayer milled designs.

Major problem is keeping loses down due to contamination  and in fact ensuring lamination process works as required WITHOUT MECHANICAL TEXTURING OF COPPER to facilitate bonding as this alters path lengths.